From be5a9604b753a39db23d54da71a13a288fb0d1eb Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 15:47:44 -0400
Subject: [PATCH 1/8] ci: path-filter Python workflows; add Rust workflow

Python CI/CD (unix, microsoft) now skips runs when only Rust-port files
change (crates/, Cargo manifests, docs/, Justfile). New rust.yml runs
fmt + clippy -D warnings + the full test suite (including golden parity)
on Linux/macOS/Windows with autocrlf disabled so byte-level goldens hold.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 .github/workflows/microsoft.yml | 14 +++++++++++
 .github/workflows/rust.yml      | 44 +++++++++++++++++++++++++++++++++
 .github/workflows/unix.yml      | 14 +++++++++++
 3 files changed, 72 insertions(+)
 create mode 100644 .github/workflows/rust.yml

diff --git a/.github/workflows/microsoft.yml b/.github/workflows/microsoft.yml
index 378e766..731a22a 100644
--- a/.github/workflows/microsoft.yml
+++ b/.github/workflows/microsoft.yml
@@ -3,8 +3,22 @@ name: Microsoft
 on:
   push:
     branches: [ main ]
+    paths-ignore:
+      - 'crates/**'
+      - 'Cargo.toml'
+      - 'Cargo.lock'
+      - 'Justfile'
+      - 'docs/**'
+      - '.github/workflows/rust.yml'
   pull_request:
     branches: [ main ]
+    paths-ignore:
+      - 'crates/**'
+      - 'Cargo.toml'
+      - 'Cargo.lock'
+      - 'Justfile'
+      - 'docs/**'
+      - '.github/workflows/rust.yml'
 
 jobs:
   build:
diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml
new file mode 100644
index 0000000..1368f6a
--- /dev/null
+++ b/.github/workflows/rust.yml
@@ -0,0 +1,44 @@
+name: Rust
+
+on:
+  push:
+    branches: [ main, rust-port ]
+    paths:
+      - 'crates/**'
+      - 'Cargo.toml'
+      - 'Cargo.lock'
+      - 'tests/golden/**'
+      - 'tests/more_languages/**'
+      - 'tests/path_to_test/**'
+      - '.github/workflows/rust.yml'
+  pull_request:
+    branches: [ main ]
+    paths:
+      - 'crates/**'
+      - 'Cargo.toml'
+      - 'Cargo.lock'
+      - 'tests/golden/**'
+      - 'tests/more_languages/**'
+      - 'tests/path_to_test/**'
+      - '.github/workflows/rust.yml'
+
+jobs:
+  test:
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, macos-latest, windows-latest]
+    steps:
+      # golden parity tests compare bytes; CRLF translation would break them
+      - name: Disable git autocrlf
+        run: git config --global core.autocrlf false
+      - uses: actions/checkout@v4
+      - uses: dtolnay/rust-toolchain@stable
+      - uses: Swatinem/rust-cache@v2
+      - name: Format
+        run: cargo fmt --all -- --check
+      - name: Clippy
+        run: cargo clippy --all-targets --all-features -- -D warnings
+      - name: Test
+        run: cargo test --workspace --all-features
diff --git a/.github/workflows/unix.yml b/.github/workflows/unix.yml
index e919220..d831be4 100644
--- a/.github/workflows/unix.yml
+++ b/.github/workflows/unix.yml
@@ -3,8 +3,22 @@ name: Linux & MacOS
 on:
   push:
     branches: [ main ]
+    paths-ignore:
+      - 'crates/**'
+      - 'Cargo.toml'
+      - 'Cargo.lock'
+      - 'Justfile'
+      - 'docs/**'
+      - '.github/workflows/rust.yml'
   pull_request:
     branches: [ main ]
+    paths-ignore:
+      - 'crates/**'
+      - 'Cargo.toml'
+      - 'Cargo.lock'
+      - 'Justfile'
+      - 'docs/**'
+      - '.github/workflows/rust.yml'
 
 jobs:
   # CI

From 3535e28bb554d7a9f690cf99f0e326edb544fdaf Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 15:48:02 -0400
Subject: [PATCH 2/8] test: add legacy golden harness and captured goldens

generate_legacy_goldens.py captures the Python implementation's outputs:
per-fixture components, token/line counts, full tree renders, and
v1-scope tree renders (deferred-language extractors stubbed). These are
the behavioral contract for the Rust port's parity suite.
diff_components.py is the dev-loop diff tool.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 tests/golden/diff_components.py               |   59 +
 tests/golden/generate_legacy_goldens.py       |  156 ++
 .../tests__dot_dot__my_test_file.py.json      |    6 +
 ...tests__dot_dot__nested_dir__.env.test.json |    6 +
 ...ests__dot_dot__nested_dir__pytest.ini.json |    4 +
 ...ot_dot__nested_dir__test_tp_dotdot.py.json |    7 +
 ...nguages__group1__CUSTOMER-INVOICE.CBL.json |   40 +
 ...more_languages__group1__JavaTest.java.json |   29 +
 ..._more_languages__group1__JuliaTest.jl.json |   17 +
 ...more_languages__group1__KotlinTest.kt.json |   52 +
 ...__more_languages__group1__LuaTest.lua.json |    8 +
 ...e_languages__group1__ObjectiveCTest.m.json |   10 +
 ..._more_languages__group1__OcamlTest.ml.json |    9 +
 ..._more_languages__group1__addamt.cobol.json |   20 +
 ...s__more_languages__group1__lesson.cbl.json |   45 +
 ...ests__more_languages__group1__test.js.json |   40 +
 ...ests__more_languages__group1__test.ts.json |   41 +
 ...__more_languages__group2__PerlTest.pl.json |    9 +
 ...__more_languages__group2__PhpTest.php.json |   10 +
 ...languages__group2__PowershellTest.ps1.json |   27 +
 ...re_languages__group2__ScalaTest.scala.json |   16 +
 ..._more_languages__group2__apl_test.apl.json |    8 +
 ...sts__more_languages__group2__c_test.c.json |   39 +
 ...s__more_languages__group2__go_test.go.json |   12 +
 ...sts__more_languages__group2__test.csv.json |   10 +
 ..._more_languages__group3__bash_test.sh.json |   11 +
 ..._more_languages__group3__cpp_test.cpp.json |   74 +
 ...ore_languages__group3__csharp_test.cs.json |   48 +
 ..._languages__group3__hallucination.tex.json |   30 +
 ..._more_languages__group3__ruby_test.rb.json |   14 +
 ...e_languages__group3__swift_test.swift.json |   23 +
 ...s__more_languages__group3__test.capnp.json |   23 +
 ..._more_languages__group3__test.graphql.json |   19 +
 ...ts__more_languages__group3__test.lean.json |   15 +
 ...s__more_languages__group3__test.proto.json |   23 +
 ...__more_languages__group3__test.sqlite.json |   13 +
 ...re_languages__group3__test_Cargo.toml.json |   12 +
 ...uages__group3__test_json_rpc_2_0.json.json |   11 +
 ..._languages__group3__test_openapi.yaml.json |   16 +
 ..._languages__group3__test_openrpc.json.json |   14 +
 ...anguages__group3__test_pyproject.toml.json |   16 +
 ...ests__more_languages__group4__RTest.R.json |    9 +
 ..._more_languages__group4__erl_test.erl.json |   20 +
 ...re_languages__group4__haskell_test.hs.json |    8 +
 ...anguages__group4__mathematica_test.nb.json |    8 +
 ...more_languages__group4__matlab_test.m.json |    7 +
 ..._more_languages__group4__rust_test.rs.json |   50 +
 ...sts__more_languages__group4__test.zig.json |   11 +
 ...ore_languages__group4__test_fsharp.fs.json |   11 +
 ...re_languages__group4__test_tcl_tk.tcl.json |    8 +
 ...s__more_languages__group4__tf_test.tf.json |   12 +
 ...sts__more_languages__group5__Makefile.json |   14 +
 ...e_languages__group5__ansible_test.yml.json |    8 +
 ...guages__group5__app-routing.module.ts.json |    7 +
 ...guages__group5__app.component.spec.ts.json |   10 +
 ...e_languages__group5__app.component.ts.json |   11 +
 ...more_languages__group5__app.module.ts.json |    7 +
 ...e_languages__group5__checkbox_test.md.json |   23 +
 ..._languages__group5__checkbox_test.txt.json |   13 +
 ...anguages__group5__environment.test.ts.json |   10 +
 ...re_languages__group5__hello_world.pyi.json |    7 +
 ...more_languages__group5__k8s_test.yaml.json |    8 +
 ...guages__group5__requirements_test.txt.json |   14 +
 ..._languages__group5__rust_todo_test.rs.json |   13 +
 ..._more_languages__group5__sql_test.sql.json |   27 +
 ...roup5__standard-app-routing.module.ts.json |    6 +
 ...sts__more_languages__group5__test.env.json |   25 +
 ...anguages__group5__testJsonSchema.json.json |    9 +
 ...e_languages__group5__testPackage.json.json |   13 +
 ...nguages__group5__tickets.component.ts.json |   54 +
 ...up6__Microsoft.PowerShell_profile.ps1.json |   35 +
 ...ore_languages__group6__catastrophic.c.json |  196 ++
 ...nguages__group6__cpp_examples_impl.cc.json |    7 +
 ...nguages__group6__cpp_examples_impl.cu.json |    7 +
 ...anguages__group6__cpp_examples_impl.h.json |    7 +
 ...more_languages__group6__edge_case.hpp.json |    4 +
 ...__more_languages__group6__fractal.thy.json |   30 +
 ...ages__group6__python_complex_class.py.json |    6 +
 ...guages__group6__ramda__cloneRegExp.js.json |    6 +
 ...more_languages__group6__ramda_prop.js.json |    9 +
 ...languages__group6__tensorflow_flags.h.json |  102 +
 ...tests__more_languages__group6__test.f.json |   11 +
 ...ts__more_languages__group6__torch.rst.json |    7 +
 ...ests__more_languages__group6__yc.html.json |    4 +
 ...anguages__group7__absurdly_huge.jsonl.json |   14 +
 ...re_languages__group7__angular_crud.ts.json |   18 +
 ..._more_languages__group7__structure.py.json |   35 +
 ...s__more_languages__group7__test.metal.json |   11 +
 ...ts__more_languages__group7__test.wgsl.json |   20 +
 ..._languages__group_lisp__LispTest.lisp.json |    7 +
 ...nguages__group_lisp__clojure_test.clj.json |   13 +
 ...guages__group_lisp__racket_struct.rkt.json |    6 +
 ...anguages__group_lisp__test_scheme.scm.json |   11 +
 ...guages__group_todo__AAPLShaders.metal.json |   30 +
 ...anguages__group_todo__crystal_test.cr.json |    4 +
 ...languages__group_todo__dart_test.dart.json |    4 +
 ...anguages__group_todo__elixir_test.exs.json |    4 +
 ...e_languages__group_todo__forward.frag.json |    4 +
 ...e_languages__group_todo__forward.vert.json |    4 +
 ...e_languages__group_todo__nodemon.json.json |    4 +
 ...e_languages__group_todo__sas_test.sas.json |    4 +
 ...nguages__group_todo__testTypings.d.ts.json |    4 +
 ...uages__group_todo__test_setup_py.test.json |    4 +
 ...e_languages__group_todo__vba_test.bas.json |    4 +
 ...languages__group_todo__wgsl_test.wgsl.json |    8 +
 ...s__path_to_test__class_method_type.py.json |   33 +
 .../tests__path_to_test__empty.py.json        |    4 +
 .../tests__path_to_test__file.md.json         |    6 +
 .../tests__path_to_test__file.py.json         |    6 +
 .../tests__path_to_test__file.txt.json        |    4 +
 .../tests__path_to_test__version.py.json      |    6 +
 .../tests__dot_dot__my_test_file.py.json      |    1 +
 ...tests__dot_dot__nested_dir__.env.test.json |    1 +
 ...ests__dot_dot__nested_dir__pytest.ini.json |    1 +
 ...ot_dot__nested_dir__test_tp_dotdot.py.json |    1 +
 ...nguages__group1__CUSTOMER-INVOICE.CBL.json |    1 +
 ...more_languages__group1__JavaTest.java.json |    1 +
 ..._more_languages__group1__JuliaTest.jl.json |    1 +
 ...more_languages__group1__KotlinTest.kt.json |    1 +
 ...__more_languages__group1__LuaTest.lua.json |    1 +
 ...e_languages__group1__ObjectiveCTest.m.json |    1 +
 ..._more_languages__group1__OcamlTest.ml.json |    1 +
 ..._more_languages__group1__addamt.cobol.json |    1 +
 ...s__more_languages__group1__lesson.cbl.json |    1 +
 ...ests__more_languages__group1__test.js.json |    1 +
 ...ests__more_languages__group1__test.ts.json |    1 +
 ...__more_languages__group2__PerlTest.pl.json |    1 +
 ...__more_languages__group2__PhpTest.php.json |    1 +
 ...languages__group2__PowershellTest.ps1.json |    1 +
 ...re_languages__group2__ScalaTest.scala.json |    1 +
 ..._more_languages__group2__apl_test.apl.json |    1 +
 ...sts__more_languages__group2__c_test.c.json |    1 +
 ...s__more_languages__group2__go_test.go.json |    1 +
 ...sts__more_languages__group2__test.csv.json |    1 +
 ..._more_languages__group3__bash_test.sh.json |    1 +
 ..._more_languages__group3__cpp_test.cpp.json |    1 +
 ...ore_languages__group3__csharp_test.cs.json |    1 +
 ..._languages__group3__hallucination.tex.json |    1 +
 ..._more_languages__group3__ruby_test.rb.json |    1 +
 ...e_languages__group3__swift_test.swift.json |    1 +
 ...s__more_languages__group3__test.capnp.json |    1 +
 ..._more_languages__group3__test.graphql.json |    1 +
 ...ts__more_languages__group3__test.lean.json |    1 +
 ...s__more_languages__group3__test.proto.json |    1 +
 ...__more_languages__group3__test.sqlite.json |    1 +
 ...re_languages__group3__test_Cargo.toml.json |    1 +
 ...uages__group3__test_json_rpc_2_0.json.json |    1 +
 ..._languages__group3__test_openapi.yaml.json |    1 +
 ..._languages__group3__test_openrpc.json.json |    1 +
 ...anguages__group3__test_pyproject.toml.json |    1 +
 ...ests__more_languages__group4__RTest.R.json |    1 +
 ..._more_languages__group4__erl_test.erl.json |    1 +
 ...re_languages__group4__haskell_test.hs.json |    1 +
 ...anguages__group4__mathematica_test.nb.json |    1 +
 ...more_languages__group4__matlab_test.m.json |    1 +
 ..._more_languages__group4__rust_test.rs.json |    1 +
 ...sts__more_languages__group4__test.zig.json |    1 +
 ...ore_languages__group4__test_fsharp.fs.json |    1 +
 ...re_languages__group4__test_tcl_tk.tcl.json |    1 +
 ...s__more_languages__group4__tf_test.tf.json |    1 +
 ...sts__more_languages__group5__Makefile.json |    1 +
 ...e_languages__group5__ansible_test.yml.json |    1 +
 ...guages__group5__app-routing.module.ts.json |    1 +
 ...guages__group5__app.component.spec.ts.json |    1 +
 ...e_languages__group5__app.component.ts.json |    1 +
 ...more_languages__group5__app.module.ts.json |    1 +
 ...e_languages__group5__checkbox_test.md.json |    1 +
 ..._languages__group5__checkbox_test.txt.json |    1 +
 ...anguages__group5__environment.test.ts.json |    1 +
 ...re_languages__group5__hello_world.pyi.json |    1 +
 ...more_languages__group5__k8s_test.yaml.json |    1 +
 ...guages__group5__requirements_test.txt.json |    1 +
 ..._languages__group5__rust_todo_test.rs.json |    1 +
 ..._more_languages__group5__sql_test.sql.json |    1 +
 ...roup5__standard-app-routing.module.ts.json |    1 +
 ...sts__more_languages__group5__test.env.json |    1 +
 ...anguages__group5__testJsonSchema.json.json |    1 +
 ...e_languages__group5__testPackage.json.json |    1 +
 ...nguages__group5__tickets.component.ts.json |    1 +
 ...up6__Microsoft.PowerShell_profile.ps1.json |    1 +
 ...ore_languages__group6__catastrophic.c.json |    1 +
 ...nguages__group6__cpp_examples_impl.cc.json |    1 +
 ...nguages__group6__cpp_examples_impl.cu.json |    1 +
 ...anguages__group6__cpp_examples_impl.h.json |    1 +
 ...more_languages__group6__edge_case.hpp.json |    1 +
 ...__more_languages__group6__fractal.thy.json |    1 +
 ...ages__group6__python_complex_class.py.json |    1 +
 ...guages__group6__ramda__cloneRegExp.js.json |    1 +
 ...more_languages__group6__ramda_prop.js.json |    1 +
 ...languages__group6__tensorflow_flags.h.json |    1 +
 ...tests__more_languages__group6__test.f.json |    1 +
 ...ts__more_languages__group6__torch.rst.json |    1 +
 ...ests__more_languages__group6__yc.html.json |    1 +
 ...anguages__group7__absurdly_huge.jsonl.json |    1 +
 ...re_languages__group7__angular_crud.ts.json |    1 +
 ..._more_languages__group7__structure.py.json |    1 +
 ...s__more_languages__group7__test.metal.json |    1 +
 ...ts__more_languages__group7__test.wgsl.json |    1 +
 ..._languages__group_lisp__LispTest.lisp.json |    1 +
 ...nguages__group_lisp__clojure_test.clj.json |    1 +
 ...guages__group_lisp__racket_struct.rkt.json |    1 +
 ...anguages__group_lisp__test_scheme.scm.json |    1 +
 ...guages__group_todo__AAPLShaders.metal.json |    1 +
 ...anguages__group_todo__crystal_test.cr.json |    1 +
 ...languages__group_todo__dart_test.dart.json |    1 +
 ...anguages__group_todo__elixir_test.exs.json |    1 +
 ...e_languages__group_todo__forward.frag.json |    1 +
 ...e_languages__group_todo__forward.vert.json |    1 +
 ...e_languages__group_todo__nodemon.json.json |    1 +
 ...e_languages__group_todo__sas_test.sas.json |    1 +
 ...nguages__group_todo__testTypings.d.ts.json |    1 +
 ...uages__group_todo__test_setup_py.test.json |    1 +
 ...e_languages__group_todo__vba_test.bas.json |    1 +
 ...languages__group_todo__wgsl_test.wgsl.json |    1 +
 ...s__path_to_test__class_method_type.py.json |    1 +
 .../counts/tests__path_to_test__empty.py.json |    1 +
 .../counts/tests__path_to_test__file.md.json  |    1 +
 .../counts/tests__path_to_test__file.py.json  |    1 +
 .../counts/tests__path_to_test__file.txt.json |    1 +
 .../tests__path_to_test__version.py.json      |    1 +
 tests/golden/legacy/trees/dot_dot.txt         |   10 +
 tests/golden/legacy/trees/more_languages.txt  | 2225 +++++++++++++++++
 .../legacy/trees/more_languages_group1.txt    |  374 +++
 .../legacy/trees/more_languages_group2.txt    |  135 +
 .../legacy/trees/more_languages_group3.txt    |  346 +++
 .../legacy/trees/more_languages_group4.txt    |  214 ++
 .../legacy/trees/more_languages_group5.txt    |  257 ++
 .../legacy/trees/more_languages_group6.txt    |  615 +++++
 .../legacy/trees/more_languages_group7.txt    |  126 +
 .../trees/more_languages_group_lisp.txt       |   22 +
 .../trees/more_languages_group_todo.txt       |  111 +
 tests/golden/legacy/trees/multi_seed.txt      |  427 ++++
 tests/golden/legacy/trees/path_to_test.txt    |   51 +
 tests/golden/legacy/trees/repo_concise.txt    |  699 ++++++
 tests/golden/legacy/trees_v1/dot_dot.txt      |   10 +
 .../golden/legacy/trees_v1/more_languages.txt | 1122 +++++++++
 .../legacy/trees_v1/more_languages_group1.txt |  146 ++
 .../legacy/trees_v1/more_languages_group2.txt |   54 +
 .../legacy/trees_v1/more_languages_group3.txt |  149 ++
 .../legacy/trees_v1/more_languages_group4.txt |  123 +
 .../legacy/trees_v1/more_languages_group5.txt |  235 ++
 .../legacy/trees_v1/more_languages_group6.txt |  300 +++
 .../legacy/trees_v1/more_languages_group7.txt |   83 +
 .../trees_v1/more_languages_group_lisp.txt    |    5 +
 .../trees_v1/more_languages_group_todo.txt    |   13 +
 tests/golden/legacy/trees_v1/multi_seed.txt   |  199 ++
 tests/golden/legacy/trees_v1/path_to_test.txt |   51 +
 247 files changed, 10379 insertions(+)
 create mode 100644 tests/golden/diff_components.py
 create mode 100644 tests/golden/generate_legacy_goldens.py
 create mode 100644 tests/golden/legacy/components/tests__dot_dot__my_test_file.py.json
 create mode 100644 tests/golden/legacy/components/tests__dot_dot__nested_dir__.env.test.json
 create mode 100644 tests/golden/legacy/components/tests__dot_dot__nested_dir__pytest.ini.json
 create mode 100644 tests/golden/legacy/components/tests__dot_dot__nested_dir__test_tp_dotdot.py.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__JavaTest.java.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__JuliaTest.jl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__KotlinTest.kt.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__LuaTest.lua.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__ObjectiveCTest.m.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__OcamlTest.ml.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__addamt.cobol.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__lesson.cbl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__test.js.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group1__test.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__PerlTest.pl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__PhpTest.php.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__PowershellTest.ps1.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__ScalaTest.scala.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__apl_test.apl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__c_test.c.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__go_test.go.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group2__test.csv.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__bash_test.sh.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__cpp_test.cpp.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__csharp_test.cs.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__hallucination.tex.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__ruby_test.rb.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__swift_test.swift.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test.capnp.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test.graphql.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test.lean.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test.proto.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test.sqlite.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test_Cargo.toml.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test_json_rpc_2_0.json.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test_openapi.yaml.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test_openrpc.json.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group3__test_pyproject.toml.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__RTest.R.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__erl_test.erl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__haskell_test.hs.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__mathematica_test.nb.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__matlab_test.m.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__rust_test.rs.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__test.zig.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__test_fsharp.fs.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__test_tcl_tk.tcl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group4__tf_test.tf.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__Makefile.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__ansible_test.yml.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__app-routing.module.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__app.component.spec.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__app.component.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__app.module.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.md.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.txt.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__environment.test.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__hello_world.pyi.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__k8s_test.yaml.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__requirements_test.txt.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__rust_todo_test.rs.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__sql_test.sql.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__standard-app-routing.module.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__test.env.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__testJsonSchema.json.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__testPackage.json.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group5__tickets.component.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__catastrophic.c.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cc.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cu.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.h.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__edge_case.hpp.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__fractal.thy.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__python_complex_class.py.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__ramda__cloneRegExp.js.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__ramda_prop.js.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__tensorflow_flags.h.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__test.f.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__torch.rst.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group6__yc.html.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group7__absurdly_huge.jsonl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group7__angular_crud.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group7__structure.py.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group7__test.metal.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group7__test.wgsl.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_lisp__LispTest.lisp.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_lisp__clojure_test.clj.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_lisp__racket_struct.rkt.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_lisp__test_scheme.scm.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__AAPLShaders.metal.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__crystal_test.cr.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__dart_test.dart.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__elixir_test.exs.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__forward.frag.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__forward.vert.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__nodemon.json.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__sas_test.sas.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__testTypings.d.ts.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__test_setup_py.test.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__vba_test.bas.json
 create mode 100644 tests/golden/legacy/components/tests__more_languages__group_todo__wgsl_test.wgsl.json
 create mode 100644 tests/golden/legacy/components/tests__path_to_test__class_method_type.py.json
 create mode 100644 tests/golden/legacy/components/tests__path_to_test__empty.py.json
 create mode 100644 tests/golden/legacy/components/tests__path_to_test__file.md.json
 create mode 100644 tests/golden/legacy/components/tests__path_to_test__file.py.json
 create mode 100644 tests/golden/legacy/components/tests__path_to_test__file.txt.json
 create mode 100644 tests/golden/legacy/components/tests__path_to_test__version.py.json
 create mode 100644 tests/golden/legacy/counts/tests__dot_dot__my_test_file.py.json
 create mode 100644 tests/golden/legacy/counts/tests__dot_dot__nested_dir__.env.test.json
 create mode 100644 tests/golden/legacy/counts/tests__dot_dot__nested_dir__pytest.ini.json
 create mode 100644 tests/golden/legacy/counts/tests__dot_dot__nested_dir__test_tp_dotdot.py.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__JavaTest.java.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__JuliaTest.jl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__KotlinTest.kt.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__LuaTest.lua.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__ObjectiveCTest.m.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__OcamlTest.ml.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__addamt.cobol.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__lesson.cbl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__test.js.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group1__test.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__PerlTest.pl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__PhpTest.php.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__PowershellTest.ps1.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__ScalaTest.scala.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__apl_test.apl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__c_test.c.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__go_test.go.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group2__test.csv.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__bash_test.sh.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__cpp_test.cpp.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__csharp_test.cs.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__hallucination.tex.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__ruby_test.rb.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__swift_test.swift.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test.capnp.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test.graphql.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test.lean.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test.proto.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test.sqlite.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test_Cargo.toml.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test_json_rpc_2_0.json.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test_openapi.yaml.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test_openrpc.json.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group3__test_pyproject.toml.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__RTest.R.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__erl_test.erl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__haskell_test.hs.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__mathematica_test.nb.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__matlab_test.m.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__rust_test.rs.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__test.zig.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__test_fsharp.fs.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__test_tcl_tk.tcl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group4__tf_test.tf.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__Makefile.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__ansible_test.yml.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__app-routing.module.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__app.component.spec.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__app.component.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__app.module.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.md.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.txt.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__environment.test.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__hello_world.pyi.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__k8s_test.yaml.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__requirements_test.txt.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__rust_todo_test.rs.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__sql_test.sql.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__standard-app-routing.module.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__test.env.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__testJsonSchema.json.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__testPackage.json.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group5__tickets.component.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__catastrophic.c.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cc.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cu.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.h.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__edge_case.hpp.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__fractal.thy.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__python_complex_class.py.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__ramda__cloneRegExp.js.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__ramda_prop.js.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__tensorflow_flags.h.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__test.f.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__torch.rst.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group6__yc.html.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group7__absurdly_huge.jsonl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group7__angular_crud.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group7__structure.py.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group7__test.metal.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group7__test.wgsl.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_lisp__LispTest.lisp.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_lisp__clojure_test.clj.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_lisp__racket_struct.rkt.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_lisp__test_scheme.scm.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__AAPLShaders.metal.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__crystal_test.cr.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__dart_test.dart.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__elixir_test.exs.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__forward.frag.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__forward.vert.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__nodemon.json.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__sas_test.sas.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__testTypings.d.ts.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__test_setup_py.test.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__vba_test.bas.json
 create mode 100644 tests/golden/legacy/counts/tests__more_languages__group_todo__wgsl_test.wgsl.json
 create mode 100644 tests/golden/legacy/counts/tests__path_to_test__class_method_type.py.json
 create mode 100644 tests/golden/legacy/counts/tests__path_to_test__empty.py.json
 create mode 100644 tests/golden/legacy/counts/tests__path_to_test__file.md.json
 create mode 100644 tests/golden/legacy/counts/tests__path_to_test__file.py.json
 create mode 100644 tests/golden/legacy/counts/tests__path_to_test__file.txt.json
 create mode 100644 tests/golden/legacy/counts/tests__path_to_test__version.py.json
 create mode 100644 tests/golden/legacy/trees/dot_dot.txt
 create mode 100644 tests/golden/legacy/trees/more_languages.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group1.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group2.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group3.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group4.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group5.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group6.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group7.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group_lisp.txt
 create mode 100644 tests/golden/legacy/trees/more_languages_group_todo.txt
 create mode 100644 tests/golden/legacy/trees/multi_seed.txt
 create mode 100644 tests/golden/legacy/trees/path_to_test.txt
 create mode 100644 tests/golden/legacy/trees/repo_concise.txt
 create mode 100644 tests/golden/legacy/trees_v1/dot_dot.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group1.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group2.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group3.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group4.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group5.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group6.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group7.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group_lisp.txt
 create mode 100644 tests/golden/legacy/trees_v1/more_languages_group_todo.txt
 create mode 100644 tests/golden/legacy/trees_v1/multi_seed.txt
 create mode 100644 tests/golden/legacy/trees_v1/path_to_test.txt

diff --git a/tests/golden/diff_components.py b/tests/golden/diff_components.py
new file mode 100644
index 0000000..ff12ac6
--- /dev/null
+++ b/tests/golden/diff_components.py
@@ -0,0 +1,59 @@
+# tests/golden/diff_components.py
+"""Diff Rust extract output against legacy goldens for given fixture paths.
+
+Usage: python tests/golden/diff_components.py <fixture-rel-path>...
+Runs the Rust example binary and prints a unified diff per fixture.
+"""
+import difflib
+import json
+import subprocess
+import sys
+from pathlib import Path
+
+REPO = Path(__file__).resolve().parents[2]
+
+
+def main() -> None:
+    paths = sys.argv[1:]
+    out = subprocess.run(
+        ["cargo", "run", "-q", "-p", "tree_plus_core", "--example", "extract", "--"]
+        + paths,
+        capture_output=True,
+        text=True,
+        cwd=REPO,
+    )
+    if out.returncode != 0:
+        print(out.stderr)
+        sys.exit(1)
+    decoder = json.JSONDecoder()
+    rest = out.stdout.strip()
+    results = []
+    while rest:
+        obj, idx = decoder.raw_decode(rest)
+        results.append(obj)
+        rest = rest[idx:].lstrip()
+    for result in results:
+        rel = result["path"]
+        golden_name = rel.replace("/", "__") + ".json"
+        golden = json.loads(
+            (REPO / "tests/golden/legacy/components" / golden_name).read_text()
+        )
+        expected = golden["components"]
+        actual = result["components"]
+        if expected == actual:
+            print(f"OK {rel}")
+            continue
+        print(f"DIFF {rel} (expected {len(expected)}, actual {len(actual)})")
+        diff = difflib.unified_diff(
+            [json.dumps(c) for c in expected],
+            [json.dumps(c) for c in actual],
+            "legacy",
+            "rust",
+            lineterm="",
+        )
+        for line in diff:
+            print(" ", line)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/tests/golden/generate_legacy_goldens.py b/tests/golden/generate_legacy_goldens.py
new file mode 100644
index 0000000..d50bd79
--- /dev/null
+++ b/tests/golden/generate_legacy_goldens.py
@@ -0,0 +1,156 @@
+# tests/golden/generate_legacy_goldens.py
+"""Generate golden outputs from the legacy Python tree_plus implementation.
+
+Run from the repository root:
+    python tests/golden/generate_legacy_goldens.py
+
+Outputs land in tests/golden/legacy/:
+- components/<sanitized_path>.json : parse_file() component lists per fixture
+- trees/<name>.txt                 : into_str() renders for directories/files
+- counts/<sanitized_path>.json     : count_tokens_lines() per fixture (wc tokenizer)
+"""
+from pathlib import Path
+import json
+import sys
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[2]))
+
+import tree_plus_src as tree_plus  # noqa: E402
+from tree_plus_src import engine  # noqa: E402
+
+REPO = Path(__file__).resolve().parents[2]
+OUT = REPO / "tests" / "golden" / "legacy"
+
+FIXTURE_DIRS = [
+    REPO / "tests" / "path_to_test",
+    REPO / "tests" / "more_languages",
+    REPO / "tests" / "dot_dot",
+]
+
+TREE_TARGETS = {
+    "path_to_test": ("tests/path_to_test",),
+    "more_languages": ("tests/more_languages",),
+    "more_languages_group1": ("tests/more_languages/group1",),
+    "more_languages_group2": ("tests/more_languages/group2",),
+    "more_languages_group3": ("tests/more_languages/group3",),
+    "more_languages_group4": ("tests/more_languages/group4",),
+    "more_languages_group5": ("tests/more_languages/group5",),
+    "more_languages_group6": ("tests/more_languages/group6",),
+    "more_languages_group7": ("tests/more_languages/group7",),
+    "more_languages_group_lisp": ("tests/more_languages/group_lisp",),
+    "more_languages_group_todo": ("tests/more_languages/group_todo",),
+    "dot_dot": ("tests/dot_dot",),
+    "multi_seed": ("tests/path_to_test", "tests/more_languages/group1"),
+}
+
+
+def sanitize(p: Path) -> str:
+    rel = p.relative_to(REPO)
+    return str(rel).replace("/", "__")
+
+
+DEFERRED_PARSERS = [
+    "parse_php", "parse_kt", "parse_swift", "parse_go", "parse_bash",
+    "parse_ps1", "parse_zig", "parse_rb", "parse_sql", "parse_graphql",
+    "parse_cs", "parse_jl", "parse_scala", "parse_java", "parse_perl",
+    "parse_hs", "parse_fsharp", "parse_lisp", "parse_erl", "parse_capnp",
+    "parse_grpc", "parse_tex", "parse_lean", "parse_fortran", "parse_tf",
+    "parse_isabelle", "parse_lua", "parse_tcl", "parse_objective_c",
+    "parse_matlab", "parse_r", "parse_mathematica", "parse_ocaml",
+    "parse_cbl", "parse_apl", "parse_metal", "parse_wgsl",
+    "parse_tensorflow_flags",
+]
+
+TS_ARTIFACT = '  return("Standalone function with parameters")'
+
+
+def _patch_for_v1_scope() -> None:
+    import importlib
+
+    # the package re-exports the function; we need the module itself
+    pf = importlib.import_module("tree_plus_src.parse_file")
+
+    for name in DEFERRED_PARSERS:
+        if hasattr(pf, name):
+            setattr(pf, name, lambda *a, **k: [])
+    original_parse_ts = pf.parse_ts
+
+    def parse_ts_no_artifact(*a, **k):
+        return [c for c in original_parse_ts(*a, **k) if c != TS_ARTIFACT]
+
+    pf.parse_ts = parse_ts_no_artifact
+    # the engine's read_file/parse caches must not leak between runs
+    pf.read_file.cache_clear()
+
+
+def main() -> None:
+    (OUT / "components").mkdir(parents=True, exist_ok=True)
+    (OUT / "trees").mkdir(parents=True, exist_ok=True)
+    (OUT / "counts").mkdir(parents=True, exist_ok=True)
+
+    n_components = 0
+    n_counts = 0
+    for fixture_dir in FIXTURE_DIRS:
+        for path in sorted(fixture_dir.rglob("*")):
+            if not path.is_file():
+                continue
+            if "__pycache__" in path.parts or path.suffix == ".pyc":
+                continue
+            # skip anything the default ignore set would hide (test artifacts
+            # like .hypothesis, caches, compiled files)
+            if tree_plus.should_ignore(path, tree_plus.DEFAULT_IGNORE):
+                continue
+            rel = str(path.relative_to(REPO))
+            try:
+                components = tree_plus.parse_file(rel)
+            except Exception as e:
+                components = [f"__EXCEPTION__: {type(e).__name__}: {e}"]
+            (OUT / "components" / f"{sanitize(path)}.json").write_text(
+                json.dumps(
+                    {"path": rel, "components": components},
+                    indent=1,
+                    ensure_ascii=False,
+                )
+            )
+            n_components += 1
+            try:
+                count = tree_plus.count_tokens_lines(rel)
+            except Exception:
+                count = None
+            payload = (
+                None
+                if count is None
+                else {"n_tokens": count.n_tokens, "n_lines": count.n_lines}
+            )
+            (OUT / "counts" / f"{sanitize(path)}.json").write_text(
+                json.dumps({"path": rel, "count": payload}, ensure_ascii=False)
+            )
+            n_counts += 1
+
+    for name, seeds in TREE_TARGETS.items():
+        root = engine.from_seeds(seeds, tokenizer_name=tree_plus.TokenizerName.WC)
+        (OUT / "trees" / f"{name}.txt").write_text(root.into_str())
+
+    # v1-scope tree goldens: legacy output with extractors outside the Rust
+    # port's version-1 scope stubbed out (deferred languages keep markers but
+    # lose components), plus documented intentional differences.
+    _patch_for_v1_scope()
+    (OUT / "trees_v1").mkdir(parents=True, exist_ok=True)
+    for name, seeds in TREE_TARGETS.items():
+        root = engine.from_seeds(seeds, tokenizer_name=tree_plus.TokenizerName.WC)
+        (OUT / "trees_v1" / f"{name}.txt").write_text(root.into_str())
+
+    # concise render of the whole repo (deterministic, no parsing)
+    root = engine.from_seeds(
+        (str(REPO),), tokenizer_name=tree_plus.TokenizerName.WC, concise=True
+    )
+    concise = root.into_str()
+    # the root label embeds the repo dir name only; keep as-is
+    (OUT / "trees" / "repo_concise.txt").write_text(concise)
+
+    print(f"wrote {n_components} component files, {n_counts} counts,"
+          f" {len(TREE_TARGETS) + 1} trees -> {OUT}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/tests/golden/legacy/components/tests__dot_dot__my_test_file.py.json b/tests/golden/legacy/components/tests__dot_dot__my_test_file.py.json
new file mode 100644
index 0000000..bd87a0a
--- /dev/null
+++ b/tests/golden/legacy/components/tests__dot_dot__my_test_file.py.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/dot_dot/my_test_file.py",
+ "components": [
+  "def dot_dot_dot()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__dot_dot__nested_dir__.env.test.json b/tests/golden/legacy/components/tests__dot_dot__nested_dir__.env.test.json
new file mode 100644
index 0000000..9020f5f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__dot_dot__nested_dir__.env.test.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/dot_dot/nested_dir/.env.test",
+ "components": [
+  "DEBUG_TREE_PLUS"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__dot_dot__nested_dir__pytest.ini.json b/tests/golden/legacy/components/tests__dot_dot__nested_dir__pytest.ini.json
new file mode 100644
index 0000000..30376bf
--- /dev/null
+++ b/tests/golden/legacy/components/tests__dot_dot__nested_dir__pytest.ini.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/dot_dot/nested_dir/pytest.ini",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__dot_dot__nested_dir__test_tp_dotdot.py.json b/tests/golden/legacy/components/tests__dot_dot__nested_dir__test_tp_dotdot.py.json
new file mode 100644
index 0000000..7013c5f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__dot_dot__nested_dir__test_tp_dotdot.py.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/dot_dot/nested_dir/test_tp_dotdot.py",
+ "components": [
+  "def ignore_tokens_lines_test(text: str) -> str",
+  "def test_tree_plus_dotdot()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json b/tests/golden/legacy/components/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
new file mode 100644
index 0000000..c3d6441
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
@@ -0,0 +1,40 @@
+{
+ "path": "tests/more_languages/group1/CUSTOMER-INVOICE.CBL",
+ "components": [
+  "IDENTIFICATION DIVISION.",
+  "PROGRAM-ID. CUSTOMER-INVOICE.",
+  "AUTHOR. JANE DOE.",
+  "DATE. 2023-12-30.",
+  "  DATE-COMPILED. 06/30/10.",
+  "    DATE-WRITTEN. 12/34/56.",
+  "ENVIRONMENT DIVISION.",
+  "INPUT-OUTPUT SECTION.",
+  "FILE-CONTROL.",
+  "    SELECT CUSTOMER-FILE.",
+  "    SELECT INVOICE-FILE.",
+  "    SELECT REPORT-FILE.",
+  "DATA DIVISION.",
+  "FILE SECTION.",
+  "FD CUSTOMER-FILE.",
+  "01 CUSTOMER-RECORD.",
+  "   05 CUSTOMER-ID.",
+  "   05 CUSTOMER-NAME.",
+  "   05 CUSTOMER-BALANCE.",
+  "FD INVOICE-FILE.",
+  "01 INVOICE-RECORD.",
+  "   05 INVOICE-ID.",
+  "   05 CUSTOMER-ID.",
+  "   05 INVOICE-AMOUNT.",
+  "FD REPORT-FILE.",
+  "01 REPORT-RECORD.",
+  "WORKING-STORAGE SECTION.",
+  "01 WS-CUSTOMER-FOUND.",
+  "01 WS-END-OF-FILE.",
+  "01 WS-TOTAL-BALANCE.",
+  "PROCEDURE DIVISION.",
+  "0000-MAIN-ROUTINE.",
+  "1000-PROCESS-RECORDS.",
+  "1100-UPDATE-CUSTOMER-BALANCE.",
+  "END PROGRAM CUSTOMER-INVOICE."
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__JavaTest.java.json b/tests/golden/legacy/components/tests__more_languages__group1__JavaTest.java.json
new file mode 100644
index 0000000..45b740d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__JavaTest.java.json
@@ -0,0 +1,29 @@
+{
+ "path": "tests/more_languages/group1/JavaTest.java",
+ "components": [
+  "abstract class LivingBeing",
+  "    abstract void breathe()",
+  "interface Communicator",
+  "    String communicate()",
+  "@Log",
+  "@Getter",
+  "@Setter",
+  "class Person extends LivingBeing implements Communicator",
+  "    Person(String name, int age)",
+  "    @Override",
+  "    void breathe()",
+  "    @Override",
+  "    public String communicate()",
+  "    void greet()",
+  "    String personalizedGreeting(String greeting, Optional<Boolean> includeAge)",
+  "@Singleton",
+  "@RestController",
+  "@SpringBootApplication",
+  "public class Example",
+  "    @Inject",
+  "    public Example(Person person)",
+  "    @RequestMapping(\"/greet\")",
+  "    String home(@RequestParam(value = \"name\", defaultValue = \"World\") String name,\n                @RequestParam(value = \"age\", defaultValue = \"30\") int age)",
+  "    public static void main(String[] args)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__JuliaTest.jl.json b/tests/golden/legacy/components/tests__more_languages__group1__JuliaTest.jl.json
new file mode 100644
index 0000000..65e67ee
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__JuliaTest.jl.json
@@ -0,0 +1,17 @@
+{
+ "path": "tests/more_languages/group1/JuliaTest.jl",
+ "components": [
+  "module JuliaTest_EdgeCase",
+  "struct Location\n    name::String \n    lat::Float32\n    lon::Float32\nend",
+  "mutable struct mPerson\n    name::String\n    age::Int\nend",
+  "Base.@kwdef mutable struct Param\n    Δt::Float64 = 0.1\n    n::Int64\n    m::Int64\nend",
+  "    sic(x,y)",
+  "welcome(l::Location)",
+  "∑(α, Ω)",
+  "function noob()\nend",
+  "function ye_olde(hello::String, world::Location)\nend",
+  "function multiline_greet(\n        p::mPerson, \n        greeting::String\n    )\nend",
+  "function julia_is_awesome(prob::DiffEqBase.AbstractDAEProblem{uType, duType, tType,\n        isinplace};\n    kwargs...) where {uType, duType, tType, isinplace}\nend",
+  "end"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__KotlinTest.kt.json b/tests/golden/legacy/components/tests__more_languages__group1__KotlinTest.kt.json
new file mode 100644
index 0000000..bedeb34
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__KotlinTest.kt.json
@@ -0,0 +1,52 @@
+{
+ "path": "tests/more_languages/group1/KotlinTest.kt",
+ "components": [
+  "data class Person(val name: String)",
+  "fun greet(person: Person)",
+  "fun <T> processItems(items: List<T>, processor: (T) -> Unit)",
+  "interface Source<out T>",
+  "    fun nextT(): T",
+  "fun MutableList<Int>.swap(index1: Int, index2: Int)",
+  "fun Any?.toString(): String",
+  "tailrec fun findFixPoint(x: Double = 1.0): Double",
+  "class GenericRepository<T>",
+  "    fun getItem(id: Int): T?",
+  "sealed interface Error",
+  "sealed class IOError(): Error",
+  "object Runner",
+  "    inline fun <reified S: SomeClass<T>, T> run() : T",
+  "infix fun Int.shl(x: Int): Int",
+  "class MyStringCollection",
+  "    infix fun add(s: String)",
+  "    fun build()",
+  "open class Base(p: Int)",
+  "class Derived(p: Int) : Base(p)",
+  "open class Shape",
+  "    open fun draw()",
+  "    fun fill()",
+  "    open fun edge(case: Int)",
+  "interface Thingy",
+  "    fun edge()",
+  "class Circle() : Shape(), Thingy",
+  "    override fun draw()",
+  "    final override fun edge(case: Int)",
+  "interface Base",
+  "    fun print()",
+  "class BaseImpl(val x: Int) : Base",
+  "    override fun print()",
+  "internal class Derived(b: Base) : Base by b",
+  "class Person constructor(firstName: String)",
+  "class People(\n    firstNames: Array<String>,\n    ages: Array<Int>(42),\n)",
+  "    fun edgeCases(): Boolean",
+  "class Alien public @Inject constructor(\n    val firstName: String,\n    val lastName: String,\n    var age: Int,\n    val pets: MutableList<Pet> = mutableListOf(),\n)",
+  "    fun objectOriented(): String",
+  " enum class IntArithmetics : BinaryOperator<Int>, IntBinaryOperator",
+  "    PLUS {\n        override fun apply(t: Int, u: Int): Int",
+  "    TIMES {\n        override fun apply(t: Int, u: Int): Int",
+  "    override fun applyAsInt(t: Int, u: Int)",
+  "fun reformat(\n    str: String,\n    normalizeCase: Boolean = true,\n    upperCaseFirstLetter: Boolean = true,\n    divideByCamelHumps: Boolean = false,\n    wordSeparator: Char = ' ',\n)",
+  "operator fun Point.unaryMinus()",
+  "abstract class Polygon",
+  "    abstract fun draw()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__LuaTest.lua.json b/tests/golden/legacy/components/tests__more_languages__group1__LuaTest.lua.json
new file mode 100644
index 0000000..c9da7e1
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__LuaTest.lua.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group1/LuaTest.lua",
+ "components": [
+  "function HelloWorld.new",
+  "function HelloWorld.greet",
+  "function say_hello"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__ObjectiveCTest.m.json b/tests/golden/legacy/components/tests__more_languages__group1__ObjectiveCTest.m.json
new file mode 100644
index 0000000..a6d7343
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__ObjectiveCTest.m.json
@@ -0,0 +1,10 @@
+{
+ "path": "tests/more_languages/group1/ObjectiveCTest.m",
+ "components": [
+  "@interface HelloWorld",
+  "@interface HelloWorld -> (void) sayHello",
+  "@implementation HelloWorld",
+  "@implementation HelloWorld -> (void) sayHello",
+  "void sayHelloWorld()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__OcamlTest.ml.json b/tests/golden/legacy/components/tests__more_languages__group1__OcamlTest.ml.json
new file mode 100644
index 0000000..e4d37de
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__OcamlTest.ml.json
@@ -0,0 +1,9 @@
+{
+ "path": "tests/more_languages/group1/OcamlTest.ml",
+ "components": [
+  "type color",
+  "class hello",
+  "class hello -> method say_hello",
+  "let main ()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__addamt.cobol.json b/tests/golden/legacy/components/tests__more_languages__group1__addamt.cobol.json
new file mode 100644
index 0000000..d906d17
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__addamt.cobol.json
@@ -0,0 +1,20 @@
+{
+ "path": "tests/more_languages/group1/addamt.cobol",
+ "components": [
+  "IDENTIFICATION DIVISION.",
+  "PROGRAM-ID.\n           ADDAMT.",
+  "DATA DIVISION.",
+  "WORKING-STORAGE SECTION.",
+  "01  KEYED-INPUT.",
+  "    05  CUST-NO-IN.",
+  "    05  AMT1-IN.",
+  "    05  AMT2-IN.",
+  "    05  AMT3-IN.",
+  "01  DISPLAYED-OUTPUT.",
+  "    05  CUST-NO-OUT.",
+  "    05  TOTAL-OUT.",
+  "01  MORE-DATA.",
+  "PROCEDURE DIVISION.",
+  "100-MAIN."
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__lesson.cbl.json b/tests/golden/legacy/components/tests__more_languages__group1__lesson.cbl.json
new file mode 100644
index 0000000..45182c6
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__lesson.cbl.json
@@ -0,0 +1,45 @@
+{
+ "path": "tests/more_languages/group1/lesson.cbl",
+ "components": [
+  "IDENTIFICATION DIVISION.",
+  "PROGRAM-ID.    CBL0002.",
+  "AUTHOR.        Otto B. Fun.",
+  "ENVIRONMENT DIVISION.",
+  "INPUT-OUTPUT SECTION.",
+  "FILE-CONTROL.",
+  "    SELECT PRINT-LINE.",
+  "    SELECT ACCT-REC.",
+  "DATA DIVISION.",
+  "FILE SECTION.",
+  "FD  PRINT-LINE.",
+  "01  PRINT-REC.",
+  "    05  ACCT-NO-O.",
+  "    05  ACCT-LIMIT-O.",
+  "    05  ACCT-BALANCE-O.",
+  "    05  LAST-NAME-O.",
+  "    05  FIRST-NAME-O.",
+  "    05  COMMENTS-O.",
+  "FD  ACCT-REC.",
+  "01  ACCT-FIELDS.",
+  "    05  ACCT-NO.",
+  "    05  ACCT-LIMIT.",
+  "    05  ACCT-BALANCE.",
+  "    05  LAST-NAME.",
+  "    05  FIRST-NAME.",
+  "    05  CLIENT-ADDR.",
+  "        10  STREET-ADDR.",
+  "        10  CITY-COUNTY.",
+  "        10  USA-STATE.",
+  "    05  RESERVED.",
+  "    05  COMMENTS.",
+  "WORKING-STORAGE SECTION.",
+  "01 FLAGS.",
+  "  05 LASTREC.",
+  "PROCEDURE DIVISION.",
+  "OPEN-FILES.",
+  "READ-NEXT-RECORD.",
+  "CLOSE-STOP.",
+  "READ-RECORD.",
+  "WRITE-RECORD."
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__test.js.json b/tests/golden/legacy/components/tests__more_languages__group1__test.js.json
new file mode 100644
index 0000000..acd47be
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__test.js.json
@@ -0,0 +1,40 @@
+{
+ "path": "tests/more_languages/group1/test.js",
+ "components": [
+  "class MyClass",
+  "  myMethod()",
+  "  async asyncMethod(a, b)",
+  "  methodWithDefaultParameters(a = 5, b = 10)",
+  "  multilineMethod(\n    c,\n    d\n  )",
+  "  multilineMethodWithDefaults(\n    t = \"tree\",\n    p = \"plus\"\n  )",
+  "function myFunction(param1, param2)",
+  "function multilineFunction(\n  param1,\n  param2\n)",
+  "const arrowFunction = () =>",
+  "const parametricArrow = (a, b) =>",
+  "function ()",
+  "function outerFunction(outerParam)",
+  "  function innerFunction(innerParam)",
+  "  innerFunction(\"inner\")",
+  "const myObject = {",
+  "  myMethod: function (stuff)",
+  "let myArrowObject = {",
+  "  myArrow: ({\n    a,\n    b,\n    c,\n  }) =>",
+  "const myAsyncArrowFunction = async () =>",
+  "function functionWithRestParameters(...args)",
+  "const namedFunctionExpression = function myNamedFunction()",
+  "const multilineArrowFunction = (\n  a,\n  b\n) =>",
+  "function functionReturningFunction()",
+  "  return function ()",
+  "function destructuringOnMultipleLines({\n  a,\n  b,\n})",
+  "const arrowFunctionWithDestructuring = ({ a, b }) =>",
+  "const multilineDestructuringArrow = ({\n  a,\n  b,\n}) =>",
+  "async function asyncFunctionWithErrorHandling()",
+  "class Car",
+  "  constructor(brand)",
+  "  present()",
+  "class Model extends Car",
+  "  constructor(brand, mod)",
+  "    super(brand)",
+  "  show()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group1__test.ts.json b/tests/golden/legacy/components/tests__more_languages__group1__test.ts.json
new file mode 100644
index 0000000..9ae0272
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group1__test.ts.json
@@ -0,0 +1,41 @@
+{
+ "path": "tests/more_languages/group1/test.ts",
+ "components": [
+  "type MyType",
+  "interface MyInterface",
+  "class TsClass",
+  "  myMethod()",
+  "  myMethodWithArgs(param1: string, param2: number): void",
+  "  static myStaticMethod<T>(param: T): T",
+  "  multilineMethod(\n    c: number,\n    d: number\n  ): number",
+  "  multilineMethodWithDefaults(\n    t: string = \"tree\",\n    p: string = \"plus\"\n  ): string",
+  "export class AdvancedComponent implements MyInterface",
+  "  async myAsyncMethod(\n    a: string,\n    b: number,\n    c: string\n  ): Promise<void>",
+  "  genericMethod<T, U>(\n    arg1: T,\n    arg2: U\n  ): [T, U]",
+  "export class TicketsComponent implements MyInterface",
+  "  async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String })",
+  "function tsFunction()",
+  "function tsFunctionSigned(\n  param1: number,\n  param2: number\n): void",
+  "export default async function tsFunctionComplicated<A, B, C>({\n  a = 1 | 2,\n  b = \"bob\",\n  c = async () => \"charlie\",\n}: {\n  a: number;\n  b: string;\n  c: () => Promise<string>;\n}): Promise<string>",
+  "  return(\"Standalone function with parameters\")",
+  "const tsArrowFunctionSigned = ({\n  a,\n  b,\n}: {\n  a: number;\n  b: string;\n}) =>",
+  "export const tsComplicatedArrow = async ({\n  a = 1 | 2,\n  b = \"bob\",\n  c = async () => \"charlie\",\n}: {\n  a: number;\n  b: string;\n  c: () => Promise<string>;\n}): Promise<string> =>",
+  "const arrowFunction = () =>",
+  "const arrow = (a: String, b: Number) =>",
+  "const asyncArrowFunction = async () =>",
+  "const asyncArrow = async (a: String, b: Number) =>",
+  "let weirdArrow = () =>",
+  "const asyncPromiseArrow = async (): Promise<void> =>",
+  "let myWeirdArrowSigned = (x: number): number =>",
+  "class Person",
+  "  constructor(private firstName: string, private lastName: string)",
+  "  getFullName(): string",
+  "  describe(): string",
+  "class Employee extends Person",
+  "  constructor(\n    firstName: string,\n    lastName: string,\n    private jobTitle: string\n  )",
+  "    super(firstName, lastName)",
+  "  describe(): string",
+  "interface Shape",
+  "interface Square extends Shape"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__PerlTest.pl.json b/tests/golden/legacy/components/tests__more_languages__group2__PerlTest.pl.json
new file mode 100644
index 0000000..052b564
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__PerlTest.pl.json
@@ -0,0 +1,9 @@
+{
+ "path": "tests/more_languages/group2/PerlTest.pl",
+ "components": [
+  "package PerlTest",
+  "package PerlTest -> sub new",
+  "package PerlTest -> sub hello",
+  "package PerlTest -> sub say_hello"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__PhpTest.php.json b/tests/golden/legacy/components/tests__more_languages__group2__PhpTest.php.json
new file mode 100644
index 0000000..42ef00f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__PhpTest.php.json
@@ -0,0 +1,10 @@
+{
+ "path": "tests/more_languages/group2/PhpTest.php",
+ "components": [
+  "class HelloWorld",
+  "class HelloWorld -> function sayHello",
+  "function greet",
+  "class Person",
+  "class Person -> function __construct"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__PowershellTest.ps1.json b/tests/golden/legacy/components/tests__more_languages__group2__PowershellTest.ps1.json
new file mode 100644
index 0000000..db3bdfc
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__PowershellTest.ps1.json
@@ -0,0 +1,27 @@
+{
+ "path": "tests/more_languages/group2/PowershellTest.ps1",
+ "components": [
+  "function Say-Nothing()",
+  "class Person",
+  "    Person([string]$name)",
+  "    [string]Greet()",
+  "    [string]GreetMany([int]$times)",
+  "    [string]GreetWithDetails([string]$greeting, [int]$times)",
+  "    [string]GreetMultiline(\n        [string]$greeting,\n        [int]$times\n    )",
+  "    NoReturn([int]$times)",
+  "    NoReturnNoArgs()",
+  "function Say-Hello([Person]$person)",
+  "function Multi-Hello([Person]$personA, [Person]$personB)",
+  "function Switch-Item",
+  "  param ([switch]$on)",
+  "function Get-SmallFiles",
+  "  param (\n      [PSDefaultValue(Help = '100')]\n      $Size = 100)",
+  "function Get-User",
+  "  [CmdletBinding(DefaultParameterSetName=\"ID\")]",
+  "  [OutputType(\"System.Int32\", ParameterSetName=\"ID\")]",
+  "  [OutputType([String], ParameterSetName=\"Name\")]",
+  "  Param (\n    [parameter(Mandatory=$true, ParameterSetName=\"ID\")]\n    [Int[]]\n    $UserID,\n    [parameter(Mandatory=$true, ParameterSetName=\"Name\")]\n    [String[]]\n    $UserName)",
+  "filter Get-ErrorLog ([switch]$Message)",
+  "function global:MultilineSignature(\n  [string]$param1,\n  [int]$param2,\n  [Parameter(Mandatory=$true)]\n  [string]$param3\n)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__ScalaTest.scala.json b/tests/golden/legacy/components/tests__more_languages__group2__ScalaTest.scala.json
new file mode 100644
index 0000000..f21daf1
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__ScalaTest.scala.json
@@ -0,0 +1,16 @@
+{
+ "path": "tests/more_languages/group2/ScalaTest.scala",
+ "components": [
+  "def sumOfSquares(x: Int, y: Int): Int",
+  "trait Bark",
+  "  def bark: String",
+  "case class Person(name: String)",
+  "class GenericClass[T](\n    val data: T,\n    val count: Int\n)",
+  "  def getData: T",
+  "object HelloWorld",
+  "  def greet(person: Person): Unit",
+  "  def main(args: Array[String]): Unit",
+  "def complexFunction(\n    a: Int,\n    b: String,\n    c: Float\n): (Int, String) Option",
+  "def sumOfSquaresShort(x: Int, y: Int): Int"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__apl_test.apl.json b/tests/golden/legacy/components/tests__more_languages__group2__apl_test.apl.json
new file mode 100644
index 0000000..b4d2954
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__apl_test.apl.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group2/apl_test.apl",
+ "components": [
+  ":Namespace HelloWorld",
+  ":Namespace HelloWorld -> hello ← 'Hello, World!'",
+  ":Namespace HelloWorld -> plus ← {⍺+⍵}"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__c_test.c.json b/tests/golden/legacy/components/tests__more_languages__group2__c_test.c.json
new file mode 100644
index 0000000..cc70367
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__c_test.c.json
@@ -0,0 +1,39 @@
+{
+ "path": "tests/more_languages/group2/c_test.c",
+ "components": [
+  "struct Point",
+  "    int x;",
+  "    int y;",
+  "struct Point getOrigin()",
+  "float mul_two_floats(float x1, float x2)",
+  "enum days",
+  "    SUN,",
+  "    MON,",
+  "    TUE,",
+  "    WED,",
+  "    THU,",
+  "    FRI,",
+  "    SAT",
+  "long add_two_longs(long x1, long x2)",
+  "double multiplyByTwo(double num)",
+  "char getFirstCharacter(char *str)",
+  "void greet(Person p)",
+  "typedef struct",
+  "    char name[50];",
+  "} Person;",
+  "int main()",
+  "int* getArrayStart(int arr[], int size)",
+  "long complexFunctionWithMultipleArguments(\n    int param1,\n    double param2,\n    char *param3,\n    struct Point point\n)",
+  "keyPattern *ACLKeyPatternCreate(sds pattern, int flags)",
+  "sds sdsCatPatternString(sds base, keyPattern *pat)",
+  "static int ACLCheckChannelAgainstList(list *reference, const char *channel, int channellen, int is_pattern)",
+  "    while((ln = listNext(&li)))",
+  "static struct config",
+  "    aeEventLoop *el;",
+  "    cliConnInfo conn_info;",
+  "    const char *hostsocket;",
+  "    int tls;",
+  "    struct cliSSLconfig sslconfig;",
+  "} config;"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__go_test.go.json b/tests/golden/legacy/components/tests__more_languages__group2__go_test.go.json
new file mode 100644
index 0000000..c2d22da
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__go_test.go.json
@@ -0,0 +1,12 @@
+{
+ "path": "tests/more_languages/group2/go_test.go",
+ "components": [
+  "type Greeting struct",
+  "func (g Greeting) sayHello()",
+  "func createGreeting(m string) Greeting",
+  "type SomethingLong struct",
+  "func (s *SomethingLong) WithAReasonableName(\n\tctx context.Context,\n\tparam1 string,\n\tparam2 int,\n\tparam3 map[string]interface{},\n\tcallback func(int) error,\n) (resultType, error)",
+  "type resultType struct",
+  "func main()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group2__test.csv.json b/tests/golden/legacy/components/tests__more_languages__group2__test.csv.json
new file mode 100644
index 0000000..00c8f4e
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group2__test.csv.json
@@ -0,0 +1,10 @@
+{
+ "path": "tests/more_languages/group2/test.csv",
+ "components": [
+  "Name",
+  "Age",
+  "Country",
+  "City",
+  "Email"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__bash_test.sh.json b/tests/golden/legacy/components/tests__more_languages__group3__bash_test.sh.json
new file mode 100644
index 0000000..345abd9
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__bash_test.sh.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group3/bash_test.sh",
+ "components": [
+  "echo_hello_world()",
+  "function fun_echo_hello_world()",
+  "export SECRET",
+  "alias md='make debug'",
+  "add_alias()",
+  "create_conda_env()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__cpp_test.cpp.json b/tests/golden/legacy/components/tests__more_languages__group3__cpp_test.cpp.json
new file mode 100644
index 0000000..d307291
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__cpp_test.cpp.json
@@ -0,0 +1,74 @@
+{
+ "path": "tests/more_languages/group3/cpp_test.cpp",
+ "components": [
+  "class Person",
+  "    std::string name;",
+  "public:",
+  "    Person(std::string n) : name(n)",
+  "    void greet()",
+  "void globalGreet()",
+  "int main()",
+  "void printMessage(const std::string &message)",
+  "template<typename T>\nvoid printVector(const std::vector<T>& vec)",
+  "struct Point",
+  "    int x, y;",
+  "    Point(int x, int y) : x(x), y(y)",
+  "class Animal",
+  "public:",
+  "    Animal(const std::string &name) : name(name)",
+  "    virtual void speak() const",
+  "    virtual ~Animal()",
+  "protected:",
+  "    std::string name;",
+  "class Dog : public Animal",
+  "public:",
+  "    Dog(const std::string &name) : Animal(name)",
+  "    void speak() const override",
+  "class Cat : public Animal",
+  "public:",
+  "    Cat(const std::string &name) : Animal(name)",
+  "    void speak() const override",
+  "nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int num_layers,\n                             int batch_size, int max_seq_length, float dropout,\n                             bool bidirectional, bool cudnn_allow_tf32,\n\t\t\t     int workspace_size, int reserve_space_size)",
+  "int main()",
+  "enum ECarTypes",
+  "  Sedan,",
+  "  Hatchback,",
+  "  SUV,",
+  "  Wagon",
+  "ECarTypes GetPreferredCarType()",
+  "enum ECarTypes : uint8_t",
+  "  Sedan,",
+  "  Hatchback,",
+  "  SUV = 254,",
+  "  Hybrid",
+  "enum class ECarTypes : uint8_t",
+  "  Sedan,",
+  "  Hatchback,",
+  "  SUV = 254,",
+  "  Hybrid",
+  "void myFunction(string fname, int age)",
+  "template <typename T> T cos(T)",
+  "template <typename T> T sin(T)",
+  "template <typename T> T sqrt(T)",
+  "template<typename T> struct VLEN",
+  "template<typename T> class arr",
+  "  private:",
+  "    static T *ralloc(size_t num)",
+  "    static void dealloc(T *ptr)",
+  "    static T *ralloc(size_t num)",
+  "    static void dealloc(T *ptr)",
+  "  public:",
+  "    arr() : p(0), sz(0)",
+  "    arr(size_t n) : p(ralloc(n)), sz(n)",
+  "    arr(arr &&other)\n      : p(other.p), sz(other.sz)",
+  "    ~arr()",
+  "    void resize(size_t n)",
+  "    T &operator[](size_t idx)",
+  "    T *data()",
+  "    size_t size() const",
+  "class Buffer",
+  " private:",
+  "  void* ptr_;",
+  "std::tuple<array, array, array> quantize(\n    const array& w,\n    int group_size,\n    int bits,\n    StreamOrDevice s)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__csharp_test.cs.json b/tests/golden/legacy/components/tests__more_languages__group3__csharp_test.cs.json
new file mode 100644
index 0000000..9dd7b8c
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__csharp_test.cs.json
@@ -0,0 +1,48 @@
+{
+ "path": "tests/more_languages/group3/csharp_test.cs",
+ "components": [
+  "public interface IExcelTemplate",
+  "    void LoadTemplate(string templateFilePath)",
+  "    void LoadData(Dictionary<string, string> data)",
+  "    void ModifyCell(string cellName, string value)",
+  "    void SaveToFile(string filePath)",
+  "public interface IGreet",
+  "    void Greet()",
+  "public enum WeekDays",
+  "public delegate void DisplayMessage(string message)",
+  "public struct Address",
+  "public static class HelperFunctions",
+  "    public static void PrintMessage(string message)",
+  "    public static int AddNumbers(int a, int b)",
+  "namespace HelloWorldApp",
+  "    class Person : IGreet",
+  "        public Person(string name, int age)",
+  "        public void Greet()",
+  "    class HelloWorld",
+  "        static void Main(string[] args)",
+  "namespace TemplateToExcelServer.Template",
+  "    public interface ITemplateObject",
+  "        string[,] GetContent()",
+  "        string[] GetContentArray()",
+  "        string[] GetFormat()",
+  "        int? GetFormatLength()",
+  "        TemplateObject SetContent(string[,] Content)",
+  "        TemplateObject SetContentArray(string[] value)",
+  "        TemplateObject SetFormat(string[] Header)",
+  "        TemplateObject SetNameOfReport(\n            ReadOnlyMemory<byte> ReportName,\n            int[] EdgeCase)",
+  "        TemplateObject SetSheetName(ReadOnlyMemory<byte> SheetName)",
+  "public class BankAccount(string accountID, string owner)",
+  "    public override string ToString() =>",
+  "var IncrementBy = (int source, int increment = 1) =>",
+  "Func<int, int, int> add = (x, y) =>",
+  "button.Click += (sender, args) =>",
+  "public Func<int, int> GetMultiplier(int factor)",
+  "public void Method(\n        int param1,\n        int param2,\n        int param3,\n        int param4,\n        int param5,\n        int param6,\n    )",
+  "System.Net.ServicePointManager.ServerCertificateValidationCallback +=\n    (se, cert, chain, sslerror) =>",
+  "class ServerCertificateValidation",
+  "    public bool OnRemoteCertificateValidation(\n        object se,\n        X509Certificate cert,\n        X509Chain chain,\n        SslPolicyErrors sslerror\n    )",
+  "s_downloadButton.Clicked += async (o, e) =>",
+  "[HttpGet, Route(\"DotNetCount\")]",
+  "static public async Task<int> GetDotNetCount(string URL)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__hallucination.tex.json b/tests/golden/legacy/components/tests__more_languages__group3__hallucination.tex.json
new file mode 100644
index 0000000..555c5f8
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__hallucination.tex.json
@@ -0,0 +1,30 @@
+{
+ "path": "tests/more_languages/group3/hallucination.tex",
+ "components": [
+  "Harnessing the Master Algorithm: Strategies for AI LLMs to Mitigate Hallucinations",
+  "Hallucinated Pedro Domingos et al.",
+  "Christmas Eve 2023",
+  "1 Introduction",
+  "2 Representation in LLMs",
+  "  2.1 Current Representational Models",
+  "  2.2 Incorporating Cognitive Structures",
+  "  2.3 Conceptual Diagrams of Advanced Representational Models",
+  "3 Evaluation Strategies",
+  "  3.1 Existing Evaluation Metrics for LLMs",
+  "  3.2 Integrating Contextual and Ethical Considerations",
+  "  3.3 Case Studies: Evaluation in Practice",
+  "4 Optimization Techniques",
+  "  4.1 Continuous Learning Models",
+  "  4.2 Adaptive Algorithms for Real-time Adjustments",
+  "  4.3 Performance Metrics Pre- and Post-Optimization",
+  "5 Interdisciplinary Insights",
+  "  5.1 Cognitive Science and AI: A Symbiotic Relationship",
+  "  5.2 Learning from Human Cognitive Processes",
+  "6 Challenges and Future Directions",
+  "  6.1 Addressing Current Limitations",
+  "  6.2 The Road Ahead: Ethical and Practical Considerations",
+  "7 Conclusion",
+  "  7.1 Summarizing Key Findings",
+  "  7.2 The Next Steps in AI Development"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__ruby_test.rb.json b/tests/golden/legacy/components/tests__more_languages__group3__ruby_test.rb.json
new file mode 100644
index 0000000..4552c04
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__ruby_test.rb.json
@@ -0,0 +1,14 @@
+{
+ "path": "tests/more_languages/group3/ruby_test.rb",
+ "components": [
+  "module Greeter",
+  "  def self.say_hello",
+  "class HelloWorld",
+  "  def say_hello",
+  "class Human",
+  "  def self.bar",
+  "  def self.bar=(value)",
+  "class Doctor < Human",
+  "  def brachial_plexus(\n      roots,\n      trunks,\n      divisions: true,\n      cords: [],\n      branches: Time.now\n    )"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__swift_test.swift.json b/tests/golden/legacy/components/tests__more_languages__group3__swift_test.swift.json
new file mode 100644
index 0000000..41f385f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__swift_test.swift.json
@@ -0,0 +1,23 @@
+{
+ "path": "tests/more_languages/group3/swift_test.swift",
+ "components": [
+  "class Person",
+  "    init(name: String)",
+  "    func greet()",
+  "    func yEdgeCase(\n        fname: String, \n        lname: String, \n        age: Int,\n        address: String, \n        phoneNumber: String\n    )",
+  "func globalGreet()",
+  "struct Point",
+  "protocol Animal",
+  "    func speak()",
+  "struct Dog: Animal",
+  "class Cat: Animal",
+  "    init(name: String)",
+  "    func speak()",
+  "enum CarType",
+  "func getPreferredCarType() -> CarType",
+  "enum CarType: UInt8",
+  "enum class CarType: UInt8",
+  "func myFunction(fname: String, age: Int)",
+  "func myFunctionWithMultipleParameters(\n    fname: String, \n    lname: String, \n    age: Int, \n    address: String, \n    phoneNumber: String\n)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test.capnp.json b/tests/golden/legacy/components/tests__more_languages__group3__test.capnp.json
new file mode 100644
index 0000000..c42af21
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test.capnp.json
@@ -0,0 +1,23 @@
+{
+ "path": "tests/more_languages/group3/test.capnp",
+ "components": [
+  "struct Employee",
+  "  id @0 :Int32",
+  "  name @1 :Text",
+  "  role @2 :Text",
+  "  skills @3 :List(Skill)",
+  "  struct Skill",
+  "    name @0 :Text",
+  "    level @1 :Level",
+  "    enum Level",
+  "      beginner @0",
+  "      intermediate @1",
+  "      expert @2",
+  "  status :union",
+  "    active @4 :Void",
+  "    onLeave @5 :Void",
+  "    retired @6 :Void",
+  "struct Company",
+  "  employees @0 :List(Employee)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test.graphql.json b/tests/golden/legacy/components/tests__more_languages__group3__test.graphql.json
new file mode 100644
index 0000000..e59dd5f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test.graphql.json
@@ -0,0 +1,19 @@
+{
+ "path": "tests/more_languages/group3/test.graphql",
+ "components": [
+  "type Query",
+  "    getBooks: [Book]",
+  "    getAuthors: [Author]",
+  "type Mutation",
+  "    addBook(title: String, author: String): Book",
+  "    removeBook(id: ID): Book",
+  "type Book",
+  "    id: ID",
+  "    title: String",
+  "    author: Author",
+  "type Author",
+  "    id: ID",
+  "    name: String",
+  "    books: [Book]"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test.lean.json b/tests/golden/legacy/components/tests__more_languages__group3__test.lean.json
new file mode 100644
index 0000000..b6734bb
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test.lean.json
@@ -0,0 +1,15 @@
+{
+ "path": "tests/more_languages/group3/test.lean",
+ "components": [
+  "# Advanced Topics in Group Theory",
+  "section GroupDynamics",
+  "lemma group_stability (G : Type*) [Group G] (H : Subgroup G)",
+  "theorem subgroup_closure {G : Type*} [Group G] (S : Set G)",
+  "axiom group_homomorphism_preservation {G H : Type*} [Group G] [Group H] (f : G → H)",
+  "end GroupDynamics",
+  "section ConstructiveApproach",
+  "lemma finite_group_order (G : Type*) [Group G] [Fintype G]",
+  "lemma complex_lemma {X Y : Type*} [SomeClass X] [AnotherClass Y]\n  (f : X → Y) (g : Y → X)",
+  "end ConstructiveApproach"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test.proto.json b/tests/golden/legacy/components/tests__more_languages__group3__test.proto.json
new file mode 100644
index 0000000..39d378a
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test.proto.json
@@ -0,0 +1,23 @@
+{
+ "path": "tests/more_languages/group3/test.proto",
+ "components": [
+  "syntax = \"proto3\"",
+  "service EmployeeService",
+  "    rpc GetEmployee(EmployeeId) returns (EmployeeInfo)",
+  "    rpc AddEmployee(EmployeeData) returns (EmployeeInfo)",
+  "    rpc UpdateEmployee(EmployeeUpdate) returns (EmployeeInfo)",
+  "message EmployeeId",
+  "    int32 id = 1",
+  "message EmployeeInfo",
+  "    int32 id = 1",
+  "    string name = 2",
+  "    string role = 3",
+  "message EmployeeData",
+  "    string name = 1",
+  "    string role = 2",
+  "message EmployeeUpdate",
+  "    int32 id = 1",
+  "    string name = 2",
+  "    string role = 3"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test.sqlite.json b/tests/golden/legacy/components/tests__more_languages__group3__test.sqlite.json
new file mode 100644
index 0000000..c604a8b
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test.sqlite.json
@@ -0,0 +1,13 @@
+{
+ "path": "tests/more_languages/group3/test.sqlite",
+ "components": [
+  "students table:",
+  "   id integer primary key",
+  "   name text not null",
+  "   age integer not null",
+  "courses table:",
+  "   id integer primary key",
+  "   title text not null",
+  "   credits integer not null"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test_Cargo.toml.json b/tests/golden/legacy/components/tests__more_languages__group3__test_Cargo.toml.json
new file mode 100644
index 0000000..a50605a
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test_Cargo.toml.json
@@ -0,0 +1,12 @@
+{
+ "path": "tests/more_languages/group3/test_Cargo.toml",
+ "components": [
+  "name: test_cargo",
+  "version: 0.1.0",
+  "description: A test Cargo.toml",
+  "license: MIT OR Apache-2.0",
+  "dependencies:",
+  "  clap 4.4",
+  "  sqlx 0.7 (features: runtime-tokio, tls-rustls)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test_json_rpc_2_0.json.json b/tests/golden/legacy/components/tests__more_languages__group3__test_json_rpc_2_0.json.json
new file mode 100644
index 0000000..764830e
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test_json_rpc_2_0.json.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group3/test_json_rpc_2_0.json",
+ "components": [
+  "jsonrpc: 2.0",
+  "method: subtract",
+  "params:",
+  "    minuend: 42",
+  "    subtrahend: 23",
+  "id: 1"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test_openapi.yaml.json b/tests/golden/legacy/components/tests__more_languages__group3__test_openapi.yaml.json
new file mode 100644
index 0000000..3a9f9d1
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test_openapi.yaml.json
@@ -0,0 +1,16 @@
+{
+ "path": "tests/more_languages/group3/test_openapi.yaml",
+ "components": [
+  "openapi: 3.0.1",
+  "    title: TODO Plugin",
+  "    description: A plugin to create and manage TODO lists using ChatGPT.",
+  "    version: v1",
+  "servers:",
+  "    - url: PLUGIN_HOSTNAME",
+  "paths:",
+  "    '/todos/{username}':",
+  "        GET (getTodos): Get the list of todos",
+  "        POST (addTodo): Add a todo to the list",
+  "        DELETE (deleteTodo): Delete a todo from the list"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test_openrpc.json.json b/tests/golden/legacy/components/tests__more_languages__group3__test_openrpc.json.json
new file mode 100644
index 0000000..0833942
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test_openrpc.json.json
@@ -0,0 +1,14 @@
+{
+ "path": "tests/more_languages/group3/test_openrpc.json",
+ "components": [
+  "openrpc: 1.2.1",
+  "info:",
+  "    title: Demo Petstore",
+  "    version: 1.0.0",
+  "methods:",
+  "    listPets: List all pets",
+  "        params:",
+  "            - limit: integer",
+  "        result: pets = An array of pets"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group3__test_pyproject.toml.json b/tests/golden/legacy/components/tests__more_languages__group3__test_pyproject.toml.json
new file mode 100644
index 0000000..dd96d4f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group3__test_pyproject.toml.json
@@ -0,0 +1,16 @@
+{
+ "path": "tests/more_languages/group3/test_pyproject.toml",
+ "components": [
+  "name: tree_plus",
+  "version: 1.0.8",
+  "description: A `tree` util enhanced with tokens, lines, and components.",
+  "License :: OSI Approved :: Apache Software License",
+  "License :: OSI Approved :: MIT License",
+  "dependencies:",
+  "    tiktoken",
+  "    PyYAML",
+  "    click",
+  "    rich",
+  "    tomli"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__RTest.R.json b/tests/golden/legacy/components/tests__more_languages__group4__RTest.R.json
new file mode 100644
index 0000000..70ff2a6
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__RTest.R.json
@@ -0,0 +1,9 @@
+{
+ "path": "tests/more_languages/group4/RTest.R",
+ "components": [
+  "class(person)",
+  "greet.Person <- function",
+  "ensure_between = function",
+  "run_intermediate_annealing_process = function"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__erl_test.erl.json b/tests/golden/legacy/components/tests__more_languages__group4__erl_test.erl.json
new file mode 100644
index 0000000..c7aeb6f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__erl_test.erl.json
@@ -0,0 +1,20 @@
+{
+ "path": "tests/more_languages/group4/erl_test.erl",
+ "components": [
+  "-module(erl_test).",
+  "-record(person).",
+  "-type ra_peer_status().",
+  "-type ra_membership().",
+  "-opaque my_opaq_type().",
+  "-type orddict(Key, Val).",
+  "-type edge(\n        Cases,\n        Pwn,\n    ).",
+  "-spec guarded(X) -> X when X :: tuple().",
+  "-spec edge_case(\n        {integer(), any()} | [any()]\n    ) -> processed, integer(), any()} | [{item, any()}].",
+  "-spec complex_function({integer(), any()} | [any()]) -> \n    {processed, integer(), any()} | [{item, any()}].",
+  "-spec list_manipulation([integer()]) -> [integer()].",
+  "-spec overload(T1, T2) -> T3\n        ; (T4, T5) -> T6.",
+  "-spec multiguard({X, integer()}) -> X when X :: atom()\n        ; ([Y]) -> Y when Y :: number().",
+  "-record(multiline).",
+  "-record(maybe_undefined)."
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__haskell_test.hs.json b/tests/golden/legacy/components/tests__more_languages__group4__haskell_test.hs.json
new file mode 100644
index 0000000..668197a
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__haskell_test.hs.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group4/haskell_test.hs",
+ "components": [
+  "data Person",
+  "greet :: Person -> String",
+  "resolveVariables ::\n  forall m fragments.\n  (MonadError QErr m, Traversable fragments) =>\n  Options.BackwardsCompatibleNullInNonNullableVariables ->\n  [G.VariableDefinition] ->\n  GH.VariableValues ->\n  [G.Directive G.Name] ->\n  G.SelectionSet fragments G.Name ->\n  m\n    ( [G.Directive Variable],\n      G.SelectionSet fragments Variable\n    )"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__mathematica_test.nb.json b/tests/golden/legacy/components/tests__more_languages__group4__mathematica_test.nb.json
new file mode 100644
index 0000000..2527c2d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__mathematica_test.nb.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group4/mathematica_test.nb",
+ "components": [
+  "person[name_]",
+  "sayHello[]",
+  "sumList[list_List]"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__matlab_test.m.json b/tests/golden/legacy/components/tests__more_languages__group4__matlab_test.m.json
new file mode 100644
index 0000000..4fae43d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__matlab_test.m.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group4/matlab_test.m",
+ "components": [
+  "classdef HelloWorld -> function greet",
+  "function loneFun"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__rust_test.rs.json b/tests/golden/legacy/components/tests__more_languages__group4__rust_test.rs.json
new file mode 100644
index 0000000..1cc0ca6
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__rust_test.rs.json
@@ -0,0 +1,50 @@
+{
+ "path": "tests/more_languages/group4/rust_test.rs",
+ "components": [
+  "fn at_beginning<'a>(&'a str)",
+  "pub enum Days<E: EdgeCase> {\n    #\\[default]\n    Sun,\n    Mon,\n    #\\[error(\"edge case {idx}, expected at least {} and at most {}\", .limits.lo, .limits.hi)]\n    Tue,\n    Wed,\n    Thu(i16, bool),\n    Fri { day: u8 },\n    Sat {\n        urday: String,\n        edge_case: E,\n    },\n}",
+  "struct Point",
+  "impl Point",
+  "    fn get_origin() -> Point",
+  "struct Person",
+  "impl Person",
+  "    fn greet(&self)",
+  "fn add_two_longs(x1: i64, x2: i64) -> i64",
+  "fn add_two_longs_longer(\n    x1: i64,\n    x2: i64,\n) -> i64",
+  "const fn multiply_by_two(num: f64) -> f64",
+  "fn get_first_character(s: &str) -> Option<char>",
+  "trait Drawable",
+  "    fn draw(&self)",
+  "impl Drawable for Point",
+  "    fn draw(&self)",
+  "fn with_generic<D: Drawable>(d: D)",
+  "fn with_generic<D>(d: D)\nwhere \n    D: Drawable",
+  "fn main()",
+  "pub struct VisibleStruct",
+  "mod my_module",
+  "    pub struct AlsoVisibleStruct<T>(T, T)",
+  "macro_rules! say_hello",
+  "#[macro_export]\nmacro_rules! hello_tree_plus",
+  "pub mod lib",
+  "    pub mod interfaces",
+  "    mod engine",
+  "pub fn flow<S1, S2, S3, S4, E, T, L>(\n    source: S1,\n    extractor: E,\n    inbox: S2,\n    transformer: T,\n    outbox: S3,\n    loader: L,\n    sink: &mut S4,\n) -> Result<(), Box<dyn Error>>\nwhere\n    S1: Extractable,\n    S2: Extractable + Loadable,\n    S3: Extractable + Loadable,\n    S4: Loadable,\n    E: Extractor<S1, S2>,\n    T: Transformer<S2, S3>,\n    L: Loader<S3, S4>",
+  "trait Container",
+  "    fn items(&self) -> impl Iterator<Item = Widget>",
+  "trait HttpService",
+  "    async fn fetch(&self, url: Url) -> HtmlBody",
+  "struct Pair<T, U>",
+  "trait Transformer<T>",
+  "    fn transform(&self, input: T) -> T",
+  "impl<T: std::ops::Add<Output = T> + Copy> Transformer<T> for Pair<T, T>",
+  "    fn transform(&self, input: T) -> T",
+  "fn main()",
+  "async fn handle_get(State(pool): State<PgPool>) -> Result<Html<String>, (StatusCode, String)> \nwhere\n    Bion: Cool",
+  "#[macro_export]\nmacro_rules! unit",
+  "            fn insert(\n                &mut self,\n                key: (),\n                value: $unit_dtype,\n            ) -> Result<Option<$unit_dtype>, ETLError>",
+  "pub async fn handle_get_axum_route(\n    Session { maybe_claims }: Session,\n    Path(RouteParams {\n        alpha,\n        bravo,\n        charlie,\n        edge_case\n    }): Path<RouteParams>,\n) -> ServerResult<Response>",
+  "fn encode_pipeline(cmds: &[Cmd], atomic: bool) -> Vec<u8>",
+  "pub async fn handle_post_yeet(\n    State(auth_backend): State<AuthBackend>,\n    Session { maybe_claims }: Session,\n    Form(yeet_form): Form<YeetForm>,\n) -> Result<Response, AuthError>",
+  "pub async fn handle_get_thingy(\n    session: Session,\n    State(ApiBackend {\n        page_cache,\n        auth_backend,\n        library_sql,\n        some_data_cache,\n        metadata_cache,\n        thingy_client,\n        ..\n    }): State<ApiBackend>,\n) -> ServerResult<Response>"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__test.zig.json b/tests/golden/legacy/components/tests__more_languages__group4__test.zig.json
new file mode 100644
index 0000000..cf9f233
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__test.zig.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group4/test.zig",
+ "components": [
+  "pub fn add(a: i32, b: i32) i32",
+  "test \"add function\"",
+  "const BunBuildOptions = struct",
+  "    pub fn updateRuntime(this: *BunBuildOptions) anyerror!void",
+  "    pub fn step(this: BunBuildOptions, b: anytype) *std.build.OptionsStep",
+  "pub fn sgemv(\n    order: Order,\n    trans: Trans,\n    m: usize,\n    n: usize,\n    alpha: f32,\n    a: []const f32,\n    lda: usize,\n    x: []const f32,\n    x_add: usize,\n    beta: f32,\n    y: []f32,\n    y_add: usize,\n) void"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__test_fsharp.fs.json b/tests/golden/legacy/components/tests__more_languages__group4__test_fsharp.fs.json
new file mode 100644
index 0000000..38162e6
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__test_fsharp.fs.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group4/test_fsharp.fs",
+ "components": [
+  "module TestFSharp",
+  "type Person = {",
+  "let add x y =",
+  "let multiply \n    (x: int) \n    (y: int): int =",
+  "let complexFunction\n    (a: int)\n    (b: string)\n    (c: float)\n    : (int * string) option =",
+  "type Result<'T> ="
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__test_tcl_tk.tcl.json b/tests/golden/legacy/components/tests__more_languages__group4__test_tcl_tk.tcl.json
new file mode 100644
index 0000000..a28d142
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__test_tcl_tk.tcl.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group4/test_tcl_tk.tcl",
+ "components": [
+  "proc sayHello {}",
+  "proc arrg { input }",
+  "proc multiLine {\n    x,\n    y\n}"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group4__tf_test.tf.json b/tests/golden/legacy/components/tests__more_languages__group4__tf_test.tf.json
new file mode 100644
index 0000000..b94f569
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group4__tf_test.tf.json
@@ -0,0 +1,12 @@
+{
+ "path": "tests/more_languages/group4/tf_test.tf",
+ "components": [
+  "provider \"aws\"",
+  "resource \"aws_instance\" \"example\"",
+  "data \"aws_ami\" \"ubuntu\"",
+  "variable \"instance_type\"",
+  "output \"instance_public_ip\"",
+  "locals",
+  "module \"vpc\""
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__Makefile.json b/tests/golden/legacy/components/tests__more_languages__group5__Makefile.json
new file mode 100644
index 0000000..37c94b5
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__Makefile.json
@@ -0,0 +1,14 @@
+{
+ "path": "tests/more_languages/group5/Makefile",
+ "components": [
+  "include dotenv/dev.env",
+  ".PHONY: dev",
+  "dev",
+  "services-down",
+  "services-stop: services-down",
+  "define CHECK_POSTGRES",
+  "damage-report",
+  "tail-logs",
+  "cloud"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__ansible_test.yml.json b/tests/golden/legacy/components/tests__more_languages__group5__ansible_test.yml.json
new file mode 100644
index 0000000..128fb5d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__ansible_test.yml.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group5/ansible_test.yml",
+ "components": [
+  "Install package",
+  "Start service",
+  "Create user"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__app-routing.module.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__app-routing.module.ts.json
new file mode 100644
index 0000000..474f900
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__app-routing.module.ts.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group5/app-routing.module.ts",
+ "components": [
+  "const routes: Routes = [\n    { path: '', redirectTo: 'login', pathMatch: 'full' },\n    { path: '*', redirectTo: 'login' },\n    { path: 'home', component: HomeComponent },\n    { path: 'login', component: LoginComponent },\n    { path: 'register', component: RegisterComponent },\n    { path: 'events', component: EventsComponent },\n    { path: 'invites', component: InvitesComponent },\n    { path: 'rewards', component: RewardsComponent },\n    { path: 'profile', component: ProfileComponent },\n];",
+  "export class AppRoutingModule"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__app.component.spec.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__app.component.spec.ts.json
new file mode 100644
index 0000000..06bade1
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__app.component.spec.ts.json
@@ -0,0 +1,10 @@
+{
+ "path": "tests/more_languages/group5/app.component.spec.ts",
+ "components": [
+  "describe 'AppComponent'",
+  "    it should create the app",
+  "    it should welcome the user",
+  "    it should welcome 'Jimbo'",
+  "    it should request login if not logged in"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__app.component.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__app.component.ts.json
new file mode 100644
index 0000000..59d0318
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__app.component.ts.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group5/app.component.ts",
+ "components": [
+  "export class AppComponent",
+  "  constructor(\n    private http: HttpClient,\n    private loginService: LoginService,\n    private stripeService: StripeService\n  )",
+  "  constructor(private loginService: LoginService)",
+  "  checkSession()",
+  "  async goToEvent(event_id: string)",
+  "  valInvitedBy(event: any, event_id: string)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__app.module.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__app.module.ts.json
new file mode 100644
index 0000000..1395c5b
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__app.module.ts.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group5/app.module.ts",
+ "components": [
+  "@NgModule({\n    declarations: [\n        AppComponent,\n        HomeComponent,\n        LoginComponent,\n        RegisterComponent,\n        EventsComponent,\n        InvitesComponent,\n        RewardsComponent,\n        ProfileComponent",
+  "export class AppModule"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.md.json b/tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.md.json
new file mode 100644
index 0000000..b8bd60f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.md.json
@@ -0,0 +1,23 @@
+{
+ "path": "tests/more_languages/group5/checkbox_test.md",
+ "components": [
+  "# My Checkbox Test",
+  "## My No Parens Test",
+  "## My Empty href Test",
+  "## My other url Test [Q&A]",
+  "## My other other url Test [Q&A]",
+  "## My 2nd other url Test [Q&A]",
+  "## My 3rd other url Test [Q&A]",
+  "- [ ] Task 1",
+  "    - [ ] No Space Task 1.1",
+  "    - [ ] Two Spaces Task 1.2",
+  "        - [ ] Subtask 1.2.1",
+  "- [ ] Task 2",
+  "- [x] Task 3",
+  "    - [ ] Subtask 3.1",
+  "- [x] Task 6",
+  "    - [x] Subtask 6.1",
+  "        - [ ] Handle edge cases",
+  "# My Codeblock Test"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.txt.json b/tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.txt.json
new file mode 100644
index 0000000..6ac8e8f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__checkbox_test.txt.json
@@ -0,0 +1,13 @@
+{
+ "path": "tests/more_languages/group5/checkbox_test.txt",
+ "components": [
+  "- [ ] fix phone number format +1",
+  "- [ ] add forgot password",
+  "- [ ] ? add email verification",
+  "- [ ] store token the right way",
+  "- [ ] test nesting of checkboxes",
+  "- [ ] user can use option to buy ticket at 2-referred price",
+  "- [ ] CTA refer 2 people to get instant lower price",
+  "- [ ] form to send referrals"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__environment.test.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__environment.test.ts.json
new file mode 100644
index 0000000..aee223d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__environment.test.ts.json
@@ -0,0 +1,10 @@
+{
+ "path": "tests/more_languages/group5/environment.test.ts",
+ "components": [
+  "environment:",
+  "   production",
+  "   cognitoUserPoolId",
+  "   cognitoAppClientId",
+  "   apiurl"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__hello_world.pyi.json b/tests/golden/legacy/components/tests__more_languages__group5__hello_world.pyi.json
new file mode 100644
index 0000000..95f66fa
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__hello_world.pyi.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group5/hello_world.pyi",
+ "components": [
+  "@final\nclass dtype(Generic[_DTypeScalar_co])",
+  "    names: None | tuple[builtins.str, ...]"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__k8s_test.yaml.json b/tests/golden/legacy/components/tests__more_languages__group5__k8s_test.yaml.json
new file mode 100644
index 0000000..e915f13
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__k8s_test.yaml.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group5/k8s_test.yaml",
+ "components": [
+  "apps/v1.Deployment -> my-app",
+  "v1.Service -> my-service",
+  "v1.ConfigMap -> my-config"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__requirements_test.txt.json b/tests/golden/legacy/components/tests__more_languages__group5__requirements_test.txt.json
new file mode 100644
index 0000000..4074c24
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__requirements_test.txt.json
@@ -0,0 +1,14 @@
+{
+ "path": "tests/more_languages/group5/requirements_test.txt",
+ "components": [
+  "psycopg2-binary",
+  "pytest",
+  "coverage",
+  "flask[async]",
+  "flask_cors",
+  "stripe",
+  "pyjwt[crypto]",
+  "cognitojwt[async]",
+  "flask-lambda"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__rust_todo_test.rs.json b/tests/golden/legacy/components/tests__more_languages__group5__rust_todo_test.rs.json
new file mode 100644
index 0000000..c31d09e
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__rust_todo_test.rs.json
@@ -0,0 +1,13 @@
+{
+ "path": "tests/more_languages/group5/rust_todo_test.rs",
+ "components": [
+  "TODO: This todo tests parse_todo",
+  "enum Color {\n    Red,\n    Blue,\n    Green,\n}",
+  "struct Point",
+  "trait Drawable",
+  "    fn draw(&self)",
+  "impl Drawable for Point",
+  "    fn draw(&self)",
+  "fn main()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__sql_test.sql.json b/tests/golden/legacy/components/tests__more_languages__group5__sql_test.sql.json
new file mode 100644
index 0000000..20c37cd
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__sql_test.sql.json
@@ -0,0 +1,27 @@
+{
+ "path": "tests/more_languages/group5/sql_test.sql",
+ "components": [
+  "CREATE TABLE promoters",
+  "   user_id serial PRIMARY KEY,",
+  "   type varchar(20) NOT NULL,",
+  "   username varchar(20) NOT NULL,",
+  "   password varchar(20) NOT NULL,",
+  "   email varchar(30) NOT NULL,",
+  "   phone varchar(20) NOT NULL,",
+  "   promocode varchar(20),",
+  "   info json,",
+  "   going text[],",
+  "   invites text[],",
+  "   balance integer NOT NULL,",
+  "   rewards text[],",
+  "   created timestamp",
+  "CREATE TABLE events",
+  "   event_id serial PRIMARY KEY,",
+  "   name varchar(64) NOT NULL,",
+  "   date varchar(64) NOT NULL,",
+  "   location varchar(64) NOT NULL,",
+  "   performer varchar(64) NOT NULL,",
+  "   rewards json,",
+  "   created timestamp"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__standard-app-routing.module.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__standard-app-routing.module.ts.json
new file mode 100644
index 0000000..9787404
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__standard-app-routing.module.ts.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/more_languages/group5/standard-app-routing.module.ts",
+ "components": [
+  "const routes: Routes = [\n  { path: '', component: HomeComponent },\n  {\n    path: 'heroes',\n    component: HeroesListComponent,\n    children: [\n      { path: ':id', component: HeroDetailComponent },\n      { path: 'new', component: HeroFormComponent },\n    ],\n  },\n  { path: '**', component: PageNotFoundComponent },\n];"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__test.env.json b/tests/golden/legacy/components/tests__more_languages__group5__test.env.json
new file mode 100644
index 0000000..76e2d01
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__test.env.json
@@ -0,0 +1,25 @@
+{
+ "path": "tests/more_languages/group5/test.env",
+ "components": [
+  "PROMO_PATH",
+  "PRODUCTION",
+  "SQL_SCHEMA_PATH",
+  "DB_LOGS",
+  "DB_LOG",
+  "PGPASSWORD",
+  "PGDATABASE",
+  "PGHOST",
+  "PGPORT",
+  "PGUSER",
+  "SERVER_LOG",
+  "SERVER_LOGS",
+  "API_URL",
+  "APP_LOGS",
+  "APP_LOG",
+  "APP_URL",
+  "COGNITO_USER_POOL_ID",
+  "COGNITO_APP_CLIENT_ID",
+  "AWS_REGION",
+  "STRIPE_SECRET_KEY"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__testJsonSchema.json.json b/tests/golden/legacy/components/tests__more_languages__group5__testJsonSchema.json.json
new file mode 100644
index 0000000..caa85e4
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__testJsonSchema.json.json
@@ -0,0 +1,9 @@
+{
+ "path": "tests/more_languages/group5/testJsonSchema.json",
+ "components": [
+  "$schema: http://json-schema.org/draft-07/schema#",
+  "type: object",
+  "title: random_test",
+  "description: A promoter's activites related to events"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__testPackage.json.json b/tests/golden/legacy/components/tests__more_languages__group5__testPackage.json.json
new file mode 100644
index 0000000..7e17045
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__testPackage.json.json
@@ -0,0 +1,13 @@
+{
+ "path": "tests/more_languages/group5/testPackage.json",
+ "components": [
+  "name: 'promo-app'",
+  "version: 0.0.0",
+  "scripts:",
+  "    ng: 'ng'",
+  "    start: 'ng serve'",
+  "    build: 'ng build'",
+  "    watch: 'ng build --watch --configuration development'",
+  "    test: 'ng test'"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group5__tickets.component.ts.json b/tests/golden/legacy/components/tests__more_languages__group5__tickets.component.ts.json
new file mode 100644
index 0000000..cda6526
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group5__tickets.component.ts.json
@@ -0,0 +1,54 @@
+{
+ "path": "tests/more_languages/group5/tickets.component.ts",
+ "components": [
+  "interface EnrichedTicket extends Ticket",
+  "interface SpinConfig",
+  "interface RotationState",
+  "interface SpeakInput",
+  "const formatSpeakInput = (input: SpeakInput): string =>",
+  "function hourToSpeech(hour: number, minute: number, period: string): string",
+  "export class TicketsComponent implements AfterViewInit",
+  "  speak(input: SpeakInput)",
+  "  speakEvent(ticket: EnrichedTicket): void",
+  "  formatEvent(ticket: EnrichedTicket): string",
+  "  speakVenue(ticket: EnrichedTicket): void",
+  "  formatDate(date: Date, oneLiner: boolean = false): string",
+  "  formatDateForSpeech(date: Date): string",
+  "  async spinQRCode(\n    event: PointerEvent,\n    config: SpinConfig = DEFAULT_SPIN_CONFIG\n  )",
+  "  private animateRotation(\n    imgElement: HTMLElement,\n    targetRotation: number,\n    config: SpinConfig,\n    cleanup: () => void\n  )",
+  "    const animate = (currentTime: number) =>",
+  "        requestAnimationFrame(animate)",
+  "        cleanup()",
+  "    requestAnimationFrame(animate)",
+  "  private getNext90Degree(currentRotation: number): number",
+  "  private getCurrentRotation(matrix: string): number",
+  "  ngAfterViewInit()",
+  "      const mouseEnterListener = () =>",
+  "      const mouseLeaveListener = () =>",
+  "  ngOnDestroy()",
+  "  toggleColumn(event: MatOptionSelectionChange, column: string)",
+  "  adjustColumns(event?: Event)",
+  "  onResize(event: Event)",
+  "  async ngOnInit()",
+  "  async loadTickets(): Promise<void>",
+  "  onDateRangeChange(\n    type: \"start\" | \"end\",\n    event: MatDatepickerInputEvent<Date>\n  )",
+  "  applyFilter(column: string): void",
+  "  formatDateForComparison(date: Date): string",
+  "  constructor(private renderer: Renderer2)",
+  "  onFilterChange(event: Event, column: string)",
+  "  onLatitudeChange(event: Event)",
+  "  onLongitudeChange(event: Event)",
+  "  onRadiusChange(event: Event)",
+  "  sortData(sort: Sort): void",
+  "  onRowClick(event: Event, row: any)",
+  "function isDate(value: Date | undefined | null): value is Date",
+  "function isNonNullNumber(value: number | null): value is number",
+  "function hasLocation(\n  ticket: any\n): ticket is",
+  "const create_faker_ticket = async () =>",
+  "function compare(a: number | string, b: number | string, isAsc: boolean)",
+  "function compare_dates(a: Date, b: Date, isAsc: boolean)",
+  "async function mockMoreTickets(): Promise<Ticket[]>",
+  "const mockTickets = async () =>",
+  "const renderQRCode = async (text: String): Promise<string> =>"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json b/tests/golden/legacy/components/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json
new file mode 100644
index 0000000..995b401
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json
@@ -0,0 +1,35 @@
+{
+ "path": "tests/more_languages/group6/Microsoft.PowerShell_profile.ps1",
+ "components": [
+  "function Log($message)",
+  "function Remove-ChocolateyFromPath",
+  "function Show-Profiles",
+  "function Show-Path",
+  "function Show-Error($err)",
+  "function Get-ScoopPackagePath",
+  "\tparam(\n\t\t[Parameter(Mandatory = $true)]\n\t\t[string]$PackageName)",
+  "function Check-Command",
+  "\tparam(\n\t\t[Parameter(Mandatory = $true)]\n\t\t[string]$Name)",
+  "function Add-ToPath",
+  "\tparam(\n\t\t[Parameter(Mandatory = $true)]\n\t\t[string]$PathToAdd)",
+  "function Install-Scoop",
+  "function Scoop-Install",
+  "\tparam(\n\t\t[Parameter(Mandatory = $true)]\n\t\t[string]$Name,\n\t\t[string]$PathToAdd)",
+  "function Start-CondaEnv",
+  "function Install-PipPackage",
+  "\tparam(\n        [Parameter(Mandatory = $true)]\n\t\t[string]$PackageName)",
+  "function Install-VSBuildTools",
+  "function Install-Crate",
+  "\tparam(\n        [Parameter(Mandatory = $true)]\n\t\t[string]$CrateName)",
+  "function Get-ScoopVersion",
+  "function Get-Version",
+  "    param(\n        [Parameter(Mandatory = $true)]\n        [string]$ExecutablePath,\n        [string]$ExecutableName)",
+  "function Show-Requirements",
+  "\tfunction Measure-Status",
+  "\t\tparam(\n\t\t\t[Parameter(Mandatory = $true)]\n\t\t\t[string]$Name)",
+  "function Find-Profile",
+  "function Edit-Profile",
+  "function Set-Profile",
+  "function Show-Profile"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__catastrophic.c.json b/tests/golden/legacy/components/tests__more_languages__group6__catastrophic.c.json
new file mode 100644
index 0000000..a9305a5
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__catastrophic.c.json
@@ -0,0 +1,196 @@
+{
+ "path": "tests/more_languages/group6/catastrophic.c",
+ "components": [
+  "TODO: technically we should use a proper parser",
+  "struct Point",
+  "    int x;",
+  "    int y;",
+  "struct Point getOrigin()",
+  "float mul_two_floats(float x1, float x2)",
+  "enum days",
+  "    SUN,",
+  "    MON,",
+  "    TUE,",
+  "    WED,",
+  "    THU,",
+  "    FRI,",
+  "    SAT",
+  "enum worker_pool_flags",
+  "        POOL_BH                 = 1 << 0,",
+  "        POOL_MANAGER_ACTIVE     = 1 << 1,",
+  "        POOL_DISASSOCIATED      = 1 << 2,",
+  "        POOL_BH_DRAINING        = 1 << 3,",
+  "enum worker_flags",
+  "        WORKER_DIE              = 1 << 1,",
+  "        WORKER_IDLE             = 1 << 2,",
+  "        WORKER_PREP             = 1 << 3,",
+  "        WORKER_CPU_INTENSIVE    = 1 << 6,",
+  "        WORKER_UNBOUND          = 1 << 7,",
+  "        WORKER_REBOUND          = 1 << 8,",
+  "        WORKER_NOT_RUNNING      = WORKER_PREP | WORKER_CPU_INTENSIVE |\n                                  WORKER_UNBOUND | WORKER_REBOUND,",
+  "struct worker_pool",
+  "  raw_spinlock_t    lock;",
+  "  int      cpu;",
+  "  int      node;",
+  "  int      id;",
+  "  unsigned int    flags;",
+  "  unsigned long    watchdog_ts;",
+  "  bool      cpu_stall;",
+  "  int      nr_running;",
+  "  struct list_head  worklist;",
+  "  int      nr_workers;",
+  "  int      nr_idle;",
+  "  struct list_head  idle_list;",
+  "  struct timer_list  idle_timer;",
+  "  struct work_struct      idle_cull_work;",
+  "  struct timer_list  mayday_timer;",
+  "  struct worker    *manager;",
+  "  struct list_head  workers;",
+  "  struct ida    worker_ida;",
+  "  struct workqueue_attrs  *attrs;",
+  "  struct hlist_node  hash_node;",
+  "  int      refcnt;",
+  "  struct rcu_head    rcu;",
+  "long add_two_longs(long x1, long x2)",
+  "double multiplyByTwo(double num)",
+  "char getFirstCharacter(char *str)",
+  "void greet(Person p)",
+  "typedef struct",
+  "    char name[50];",
+  "} Person;",
+  "typedef struct PersonA",
+  "    char name[50];",
+  "} PersonB;",
+  "int main()",
+  "int* getArrayStart(int arr[], int size)",
+  "long complexFunctionWithMultipleArguments(\n    int param1,\n    double param2,\n    char *param3,\n    struct Point point\n)",
+  "keyPattern *ACLKeyPatternCreate(sds pattern, int flags)",
+  "sds sdsCatPatternString(sds base, keyPattern *pat)",
+  "static int ACLCheckChannelAgainstList(list *reference, const char *channel, int channellen, int is_pattern)",
+  "    while((ln = listNext(&li)))",
+  "static struct config",
+  "    aeEventLoop *el;",
+  "    cliConnInfo conn_info;",
+  "    const char *hostsocket;",
+  "    int tls;",
+  "    struct cliSSLconfig sslconfig;",
+  "} config;",
+  "class Person",
+  "    std::string name;",
+  "public:",
+  "    Person(std::string n) : name(n)",
+  "    void greet()",
+  "void globalGreet()",
+  "int main()",
+  "void printMessage(const std::string &message)",
+  "template<typename T>\nvoid printVector(const std::vector<T>& vec)",
+  "struct foo",
+  "    char x;",
+  "    struct foo_in",
+  "          char* y;",
+  "          short z;",
+  "    } inner;",
+  "struct Point",
+  "    int x, y;",
+  "    Point(int x, int y) : x(x), y(y)",
+  "class Animal",
+  "  public:",
+  "    Animal(const std::string &name) : name(name)",
+  "    virtual void speak() const",
+  "    virtual ~Animal()",
+  "protected:",
+  "    std::string name;",
+  "class Dog : public Animal",
+  "  public:",
+  "    Dog(const std::string &name) : Animal(name)",
+  "    void speak() const override",
+  "class Cat : public Animal",
+  "  public:",
+  "    Cat(const std::string &name) : Animal(name)",
+  "    void speak() const override",
+  "class CatDog: public Animal, public Cat, public Dog",
+  "  public:",
+  "      CatDog(const std::string &name) : Animal(name)",
+  "      int meow_bark()",
+  "nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int num_layers,\n                             int batch_size, int max_seq_length, float dropout,\n                             bool bidirectional, bool cudnn_allow_tf32,\n\t\t\t     int workspace_size, int reserve_space_size)",
+  "int main()",
+  "enum ECarTypes",
+  "  Sedan,",
+  "  Hatchback,",
+  "  SUV,",
+  "  Wagon",
+  "ECarTypes GetPreferredCarType()",
+  "enum ECarTypes : uint8_t",
+  "  Sedan,",
+  "  Hatchback,",
+  "  SUV = 254,",
+  "  Hybrid",
+  "enum class ECarTypes : uint8_t",
+  "  Sedan,",
+  "  Hatchback,",
+  "  SUV = 254,",
+  "  Hybrid",
+  "void myFunction(string fname, int age)",
+  "template <typename T> T cos(T)",
+  "template <typename T> T sin(T)",
+  "template <typename T> T sqrt(T)",
+  "template<typename T> struct VLEN",
+  "template<typename T> class arr",
+  "  private:",
+  "    static T *ralloc(size_t num)",
+  "    static void dealloc(T *ptr)",
+  "    static T *ralloc(size_t num)",
+  "    static void dealloc(T *ptr)",
+  "  public:",
+  "    arr() : p(0), sz(0)",
+  "    arr(size_t n) : p(ralloc(n)), sz(n)",
+  "    arr(arr &&other)\n      : p(other.p), sz(other.sz)",
+  "    ~arr()",
+  "    void resize(size_t n)",
+  "    T &operator[](size_t idx)",
+  "    T *data()",
+  "    size_t size() const",
+  "class Buffer",
+  " private:",
+  "  void* ptr_;",
+  "std::tuple<array, array, array> quantize(\n    const array& w,\n    int group_size,\n    int bits,\n    StreamOrDevice s)",
+  "#define PY_SSIZE_T_CLEAN",
+  "#define PLATFORM_IS_X86",
+  "#define PLATFORM_WINDOWS",
+  "#define GETCPUID(a, b, c, d, a_inp, c_inp)",
+  "static int GetXCR0EAX()",
+  "#define GETCPUID(a, b, c, d, a_inp, c_inp)",
+  "static int GetXCR0EAX()",
+  "  asm(\"XGETBV\" : \"=a\"(eax), \"=d\"(edx) : \"c\"(0))",
+  "static void ReportMissingCpuFeature(const char* name)",
+  "static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)",
+  "static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)",
+  "static PyMethodDef cpu_feature_guard_methods[]",
+  "static struct PyModuleDef cpu_feature_guard_module",
+  "#define EXPORT_SYMBOL __declspec(dllexport)",
+  "#define EXPORT_SYMBOL __attribute__ ((visibility(\"default\")))",
+  "EXPORT_SYMBOL PyMODINIT_FUNC PyInit_cpu_feature_guard(void)",
+  "typedef struct",
+  "    GPT2Config config;",
+  "    ParameterTensors params;",
+  "    size_t param_sizes[NUM_PARAMETER_TENSORS];",
+  "    float* params_memory;",
+  "    size_t num_parameters;",
+  "    ParameterTensors grads;",
+  "    float* grads_memory;",
+  "    float* m_memory;",
+  "    float* v_memory;",
+  "    ActivationTensors acts;",
+  "    size_t act_sizes[NUM_ACTIVATION_TENSORS];",
+  "    float* acts_memory;",
+  "    size_t num_activations;",
+  "    ActivationTensors grads_acts;",
+  "    float* grads_acts_memory;",
+  "    int batch_size;",
+  "    int seq_len;",
+  "    int* inputs;",
+  "    int* targets;",
+  "    float mean_loss;",
+  "} GPT2;"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cc.json b/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cc.json
new file mode 100644
index 0000000..e1c7a7f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cc.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group6/cpp_examples_impl.cc",
+ "components": [
+  "PYBIND11_MODULE(cpp_examples, m)",
+  "    m.def(\"add\", &add<int>, \"An example function to add two numbers.\")"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cu.json b/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cu.json
new file mode 100644
index 0000000..9c58fdb
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.cu.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group6/cpp_examples_impl.cu",
+ "components": [
+  "template <typename T>\nT add(T a, T b)",
+  "template <>\nint add<int>(int a, int b)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.h.json b/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.h.json
new file mode 100644
index 0000000..49c3962
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__cpp_examples_impl.h.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group6/cpp_examples_impl.h",
+ "components": [
+  "template <typename T>\nT add(T a, T b)",
+  "template <>\nint add<int>(int, int)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__edge_case.hpp.json b/tests/golden/legacy/components/tests__more_languages__group6__edge_case.hpp.json
new file mode 100644
index 0000000..077ea00
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__edge_case.hpp.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group6/edge_case.hpp",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__fractal.thy.json b/tests/golden/legacy/components/tests__more_languages__group6__fractal.thy.json
new file mode 100644
index 0000000..9b9a4e9
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__fractal.thy.json
@@ -0,0 +1,30 @@
+{
+ "path": "tests/more_languages/group6/fractal.thy",
+ "components": [
+  "Title:      fractal.thy",
+  "Author:     Isabelle/HOL Contributors!",
+  "Author:     edge cases r us",
+  "theory Simplified_Ring",
+  "section ‹Basic Algebraic Structures›",
+  "class everything = nothing + itself",
+  "subsection ‹Monoids›",
+  "definition ring_hom :: \"[('a, 'm) ring_scheme, ('b, 'n) ring_scheme] => ('a => 'b) set\"",
+  "fun example_fun :: \"nat ⇒ nat\"",
+  "locale monoid =\n  fixes G (structure)\n  assumes m_closed: \"⟦x ∈ carrier G; y ∈ carrier G⟧ ⟹  x ⊗ y ∈ carrier G\"\n    and m_assoc: \"⟦x ∈ carrier G; y ∈ carrier G; z ∈ carrier G⟧ ⟹  (x ⊗ y) ⊗ z = x ⊗ (y ⊗ z)\"\n    and one_closed: \"𝟭 ∈ carrier G\"\n    and l_one: \"x ∈ carrier G ⟹  𝟭 ⊗ x = x\"\n    and r_one: \"x ∈ carrier G ⟹  x ⊗ 𝟭 = x\"",
+  "subsection ‹Groups›",
+  "locale group = monoid +\n  assumes Units_closed: \"x ∈ Units G ⟹  x ∈ carrier G\"\n    and l_inv_ex: \"x ∈ carrier G ⟹  ∃ y ∈ carrier G. y ⊗ x = 𝟭\"\n    and r_inv_ex: \"x ∈ carrier G ⟹  ∃ y ∈ carrier G. x ⊗ y = 𝟭\"",
+  "subsection ‹Rings›",
+  "locale ring = abelian_group R + monoid R +\n  assumes l_distr: \"⟦x ∈ carrier R; y ∈ carrier R; z ∈ carrier R⟧ ⟹  (x ⊕ y) ⊗ z = x ⊗ z ⊕ y ⊗ z\"\n    and r_distr: \"⟦x ∈ carrier R; y ∈ carrier R; z ∈ carrier R⟧ ⟹  z ⊗ (x ⊕ y) = z ⊗ x ⊕ z ⊗ y\"",
+  "locale commutative_ring = ring +\n  assumes m_commutative: \"⟦x ∈ carrier R; y ∈ carrier R⟧ ⟹  x ⊗ y = y ⊗ x\"",
+  "locale domain = commutative_ring +\n  assumes no_zero_divisors: \"⟦a ⊗ b = 𝟬; a ∈ carrier R; b ∈ carrier R⟧ ⟹  a = 𝟬 ∨ b = 𝟬\"",
+  "locale field = domain +\n  assumes inv_ex: \"x ∈ carrier R - {𝟬} ⟹  inv x ∈ carrier R\"",
+  "subsection ‹Morphisms›",
+  "lemma example_lemma: \"example_fun n = n\"",
+  "qualified lemma gcd_0:\n  \"gcd a 0 = normalize a\"",
+  "lemma abelian_monoidI:\n  fixes R (structure)\n      and f :: \"'edge::{} ⇒ 'case::{}\"\n  assumes \"⋀x y. ⟦ x ∈ carrier R; y ∈ carrier R ⟧ ⟹  x ⊕ y ∈ carrier R\"\n      and \"𝟬 ∈ carrier R\"\n      and \"⋀x y z. ⟦ x ∈ carrier R; y ∈ carrier R; z ∈ carrier R ⟧ ⟹  (x ⊕ y) ⊕ z = x ⊕ (y ⊕ z)\"\n  shows \"abelian_monoid R\"",
+  "lemma euclidean_size_gcd_le1 [simp]:\n  assumes \"a ≠ 0\"\n  shows \"euclidean_size (gcd a b) ≤ euclidean_size a\"",
+  "theorem Residue_theorem:\n  fixes S pts::\"complex set\" and f::\"complex ⇒ complex\"\n    and g::\"real ⇒ complex\"\n  assumes \"open S\" \"connected S\" \"finite pts\" and\n          holo:\"f holomorphic_on S-pts\" and\n          \"valid_path g\" and\n          loop:\"pathfinish g = pathstart g\" and\n          \"path_image g ⊆ S-pts\" and\n          homo:\"∀z. (z ∉ S) ⟶  winding_number g z  = 0\"\n  shows \"contour_integral g f = 2 * pi * 𝗂 *(∑p ∈ pts. winding_number g p * residue f p)\"",
+  "corollary fps_coeff_residues_bigo':\n  fixes f :: \"complex ⇒ complex\" and r :: real\n  assumes exp: \"f has_fps_expansion F\"\n  assumes \"open A\" \"connected A\" \"cball 0 r ⊆ A\" \"r > 0\" \n  assumes \"f holomorphic_on A - S\" \"S ⊆ ball 0 r\" \"finite S\" \"0 ∉ S\"\n  assumes \"eventually (λn. g n = -(∑z ∈ S. residue (λz. f z / z ^ Suc n) z)) sequentially\"\n             (is \"eventually (λn. _ = -?g' n) _\")\n  shows   \"(λn. fps_nth F n - g n) ∈ O(λn. 1 / r ^ n)\" (is \"(λn. ?c n - _) ∈ O(_)\")",
+  "end"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__python_complex_class.py.json b/tests/golden/legacy/components/tests__more_languages__group6__python_complex_class.py.json
new file mode 100644
index 0000000..50da2b1
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__python_complex_class.py.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/more_languages/group6/python_complex_class.py",
+ "components": [
+  "class Box(Space[NDArray[Any]])"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__ramda__cloneRegExp.js.json b/tests/golden/legacy/components/tests__more_languages__group6__ramda__cloneRegExp.js.json
new file mode 100644
index 0000000..b5ece94
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__ramda__cloneRegExp.js.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/more_languages/group6/ramda__cloneRegExp.js",
+ "components": [
+  "export default function _cloneRegExp(pattern)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__ramda_prop.js.json b/tests/golden/legacy/components/tests__more_languages__group6__ramda_prop.js.json
new file mode 100644
index 0000000..3dffe56
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__ramda_prop.js.json
@@ -0,0 +1,9 @@
+{
+ "path": "tests/more_languages/group6/ramda_prop.js",
+ "components": [
+  "/**\n * Returns a function that when supplied an object returns the indicated\n * property of that object, if it exists.\n * @category Object\n * @typedefn Idx = String | Int | Symbol\n * @sig Idx -> {s: a} -> a | Undefined\n * @param {String|Number} p The property name or array index\n * @param {Object} obj The object to query\n * @return {*} The value at `obj.p`.\n */\nvar prop = _curry2(function prop(p, obj)",
+  "/**\n * Solves equations of the form a * x = b\n * @param {{\n *  z: number\n * }} x\n */\nfunction foo(x)",
+  "/**\n * Deconstructs an array field from the input documents to output a document for each element.\n * Each output document is the input document with the value of the array field replaced by the element.\n * @category Object\n * @sig String -> {k: [v]} -> [{k: v}]\n * @param {String} key The key to determine which property of the object should be unwound.\n * @param {Object} object The object containing the list to unwind at the property named by the key.\n * @return {List} A list of new objects, each having the given key associated to an item from the unwound list.\n */\nvar unwind = _curry2(function(key, object)",
+  "  return _map(function(item)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__tensorflow_flags.h.json b/tests/golden/legacy/components/tests__more_languages__group6__tensorflow_flags.h.json
new file mode 100644
index 0000000..3e56f70
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__tensorflow_flags.h.json
@@ -0,0 +1,102 @@
+{
+ "path": "tests/more_languages/group6/tensorflow_flags.h",
+ "components": [
+  "TF_DECLARE_FLAG('test_only_experiment_1')",
+  "TF_DECLARE_FLAG('test_only_experiment_2')",
+  "TF_DECLARE_FLAG('enable_nested_function_shape_inference'):\n\tAllow ops such as tf.cond to invoke the ShapeRefiner on their nested functions.",
+  "TF_DECLARE_FLAG('enable_quantized_dtypes_training'):\n\tSet quantized dtypes, like tf.qint8, to be trainable.",
+  "TF_DECLARE_FLAG('graph_building_optimization'):\n\tOptimize graph building for faster tf.function tracing.",
+  "TF_DECLARE_FLAG('saved_model_fingerprinting'):\n\tAdd fingerprint to SavedModels.",
+  "TF_DECLARE_FLAG('more_stack_traces'):\n\tEnable experimental code that preserves and propagates graph node stack traces in C++.",
+  "TF_DECLARE_FLAG('publish_function_graphs'):\n\tEnables the publication of partitioned function graphs via StatsPublisherInterface. Disabling this flag can reduce memory consumption.",
+  "TF_DECLARE_FLAG('enable_aggressive_constant_replication'):\n\tReplicate constants across CPU devices and even for local CPUs within the same task if available.",
+  "TF_DECLARE_FLAG('enable_colocation_key_propagation_in_while_op_lowering'):\n\tIf true, colocation key attributes for the ops will be propagated during while op lowering to switch/merge ops.",
+  "Flag('tf_xla_auto_jit'):\n\tControl compilation of operators into XLA computations on CPU and GPU devices.  0 = use ConfigProto setting; -1 = off; 1 = on for things very likely to be improved; 2 = on for everything; (experimental) fusible = only for Tensorflow operations that XLA knows how to fuse. If set to single-gpu(<N>) then this resolves to <N> for single-GPU graphs (graphs that have at least one node placed on a GPU and no more than one GPU is in use through the entire graph) and 0 otherwise.  Experimental.",
+  "Flag('tf_xla_min_cluster_size'):\n\tMinimum number of operators in an XLA compilation. Ignored for operators placed on an XLA device or operators explicitly marked for compilation.",
+  "Flag('tf_xla_max_cluster_size'):\n\tMaximum number of operators in an XLA compilation.",
+  "Flag('tf_xla_cluster_exclude_ops'):\n\t(experimental) Exclude the operations from auto-clustering. If multiple, separate them with commas. Where, Some_other_ops.",
+  "Flag('tf_xla_clustering_debug'):\n\tDump graphs during XLA compilation.",
+  "Flag('tf_xla_cpu_global_jit'):\n\tEnables global JIT compilation for CPU via SessionOptions.",
+  "Flag('tf_xla_clustering_fuel'):\n\tPlaces an artificial limit on the number of ops marked as eligible for clustering.",
+  "Flag('tf_xla_disable_deadness_safety_checks_for_debugging'):\n\tDisable deadness related safety checks when clustering (this is unsound).",
+  "Flag('tf_xla_disable_resource_variable_safety_checks_for_debugging'):\n\tDisable resource variables related safety checks when clustering (this is unsound).",
+  "Flag('tf_xla_deterministic_cluster_names'):\n\tCauses the function names assigned by auto clustering to be deterministic from run to run.",
+  "Flag('tf_xla_persistent_cache_directory'):\n\tIf non-empty, JIT-compiled executables are saved to and loaded from the specified file system directory path. Empty by default.",
+  "Flag('tf_xla_persistent_cache_device_types'):\n\tIf non-empty, the persistent cache will only be used for the specified devices (comma separated). Each device type should be able to be converted to.",
+  "Flag('tf_xla_persistent_cache_read_only'):\n\tIf true, the persistent cache will be read-only.",
+  "Flag('tf_xla_disable_strict_signature_checks'):\n\tIf true, entires loaded into the XLA compile cache will not have their signatures checked strictly. Defaults to false.",
+  "Flag('tf_xla_persistent_cache_prefix'):\n\tSpecifies the persistance cache prefix. Default is.",
+  "Flag('tf_xla_sparse_core_disable_table_stacking'):\n\tDisable table stacking for all the tables passed to the SparseCore mid level API.",
+  "Flag('tf_xla_sparse_core_minibatch_max_division_level'):\n\tMax level of division to split input data into minibatches.",
+  "Flag('tf_xla_sparse_core_stacking_mem_limit_bytes'):\n\tIf non-zero, limits the size of the activations for a given table to be below these many bytes.",
+  "Flag('tf_xla_sparse_core_stacking_table_shard_limit_bytes'):\n\tIf non-zero, limits the size of any table shard to be below these many bytes.",
+  "Flag('always_specialize')",
+  "Flag('cost_driven_async_parallel_for')",
+  "Flag('enable_crash_reproducer')",
+  "Flag('log_query_of_death')",
+  "Flag('vectorize')",
+  "Flag('tf_xla_enable_lazy_compilation')",
+  "Flag('tf_xla_print_cluster_outputs'):\n\tIf true then insert Print nodes to print out values produced by XLA clusters.",
+  "Flag('tf_xla_check_cluster_input_numerics'):\n\tIf true then insert CheckNumerics nodes to check all cluster inputs.",
+  "Flag('tf_xla_check_cluster_output_numerics'):\n\tIf true then insert CheckNumerics nodes to check all cluster outputs.",
+  "Flag('tf_xla_disable_constant_folding'):\n\tIf true then disables constant folding on TF graph before XLA compilation.",
+  "Flag('tf_xla_disable_full_embedding_pipelining'):\n\tIf true then disables full embedding pipelining and instead use strict SparseCore / TensorCore sequencing.",
+  "Flag('tf_xla_embedding_parallel_iterations'):\n\tIf >0 then use this many parallel iterations in embedding_pipelining and embedding_sequency. By default, use the parallel_iterations on the original model WhileOp.",
+  "Flag('tf_xla_compile_on_demand'):\n\tSwitch a device into 'on-demand' mode, where instead of autoclustering ops are compiled one by one just-in-time.",
+  "Flag('tf_xla_enable_xla_devices'):\n\tGenerate XLA_* devices, where placing a computation on such a device forces compilation by XLA. Deprecated.",
+  "Flag('tf_xla_always_defer_compilation')",
+  "Flag('tf_xla_async_compilation'):\n\tWhen lazy compilation is enabled, asynchronous compilation starts the cluster compilation in the background, and the fallback path is executed until the compilation has finished.",
+  "Flag('tf_xla_use_device_api_for_xla_launch'):\n\tIf true, uses Device API (PjRt) for single device compilation and execution of functions marked for JIT compilation i.e. jit_compile=True. Defaults to false.",
+  "Flag('tf_xla_use_device_api_for_compile_on_demand'):\n\tIf true, uses Device API (PjRt) for compiling and executing ops one by one in 'on-demand' mode. Defaults to false.",
+  "Flag('tf_xla_use_device_api_for_auto_jit'):\n\tIf true, uses Device API (PjRt) for compilation and execution when auto-clustering is enabled. Defaults to false.",
+  "Flag('tf_xla_use_device_api'):\n\tIf true, uses Device API (PjRt) for compilation and execution of ops one-by-one in 'on-demand' mode, for functions marked for JIT compilation, or when auto-clustering is enabled. Defaults to false.",
+  "Flag('tf_xla_enable_device_api_for_gpu'):\n\tIf true, uses Device API (PjRt) for TF GPU device. This is a helper flag so that individual tests can turn on PjRt for GPU specifically.",
+  "Flag('tf_xla_call_module_disabled_checks'):\n\tA comma-sepated list of directives specifying the safety checks to be skipped when compiling XlaCallModuleOp. See the op documentation for the recognized values.",
+  "Flag('tf_mlir_enable_mlir_bridge'):\n\tEnables experimental MLIR-Based TensorFlow Compiler Bridge.",
+  "Flag('tf_mlir_enable_merge_control_flow_pass'):\n\tEnables MergeControlFlow pass for MLIR-Based TensorFlow Compiler Bridge.",
+  "Flag('tf_mlir_enable_convert_control_to_data_outputs_pass'):\n\tEnables MLIR-Based TensorFlow Compiler Bridge.",
+  "Flag('tf_mlir_enable_strict_clusters'):\n\tDo not allow clusters that have cyclic control dependencies.",
+  "Flag('tf_mlir_enable_multiple_local_cpu_devices'):\n\tEnable multiple local CPU devices. CPU ops which are outside compiled inside the tpu cluster will also be replicated across multiple cpu devices.",
+  "Flag('tf_dump_graphs_in_tfg'):\n\tWhen tf_dump_graphs_in_tfg is true, graphs after transformations are dumped in MLIR TFG dialect and not in GraphDef.",
+  "Flag('tf_mlir_enable_generic_outside_compilation'):\n\tEnables OutsideCompilation passes for MLIR-Based TensorFlow Generic Compiler Bridge.",
+  "Flag('tf_mlir_enable_tpu_variable_runtime_reformatting_pass'):\n\tEnables TPUVariableRuntimeReformatting pass for MLIR-Based TensorFlow Compiler Bridge. This enables weight update sharding and creates TPUReshardVariables ops.",
+  "TF_PY_DECLARE_FLAG('test_only_experiment_1')",
+  "TF_PY_DECLARE_FLAG('test_only_experiment_2')",
+  "TF_PY_DECLARE_FLAG('enable_nested_function_shape_inference')",
+  "TF_PY_DECLARE_FLAG('enable_quantized_dtypes_training')",
+  "TF_PY_DECLARE_FLAG('graph_building_optimization')",
+  "TF_PY_DECLARE_FLAG('op_building_optimization')",
+  "TF_PY_DECLARE_FLAG('saved_model_fingerprinting')",
+  "TF_PY_DECLARE_FLAG('tf_shape_default_int64')",
+  "TF_PY_DECLARE_FLAG('more_stack_traces')",
+  "TF_PY_DECLARE_FLAG('publish_function_graphs')",
+  "TF_PY_DECLARE_FLAG('enable_aggressive_constant_replication')",
+  "TF_PY_DECLARE_FLAG('enable_colocation_key_propagation_in_while_op_lowering')",
+  "#define TENSORFLOW_CORE_CONFIG_FLAG_DEFS_H_",
+  "class Flags",
+  " public:",
+  "bool SetterForXlaAutoJitFlag(const string& value)",
+  "bool SetterForXlaCallModuleDisabledChecks(const string& value)",
+  "void AppendMarkForCompilationPassFlagsInternal(std::vector<Flag>* flag_list)",
+  "void AllocateAndParseJitRtFlags()",
+  "void AllocateAndParseFlags()",
+  "void ResetFlags()",
+  "bool SetXlaAutoJitFlagFromFlagString(const string& value)",
+  "BuildXlaOpsPassFlags* GetBuildXlaOpsPassFlags()",
+  "MarkForCompilationPassFlags* GetMarkForCompilationPassFlags()",
+  "XlaSparseCoreFlags* GetXlaSparseCoreFlags()",
+  "XlaDeviceFlags* GetXlaDeviceFlags()",
+  "XlaOpsCommonFlags* GetXlaOpsCommonFlags()",
+  "XlaCallModuleFlags* GetXlaCallModuleFlags()",
+  "MlirCommonFlags* GetMlirCommonFlags()",
+  "void ResetJitCompilerFlags()",
+  "const JitRtFlags& GetJitRtFlags()",
+  "ConfigProto::Experimental::MlirBridgeRollout GetMlirBridgeRolloutState(\n    std::optional<const ConfigProto> config_proto)",
+  "void AppendMarkForCompilationPassFlags(std::vector<Flag>* flag_list)",
+  "void DisableXlaCompilation()",
+  "void EnableXlaCompilation()",
+  "bool FailOnXlaCompilation()",
+  "#define TF_PY_DECLARE_FLAG(flag_name)",
+  "PYBIND11_MODULE(flags_pybind, m)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__test.f.json b/tests/golden/legacy/components/tests__more_languages__group6__test.f.json
new file mode 100644
index 0000000..620d4d7
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__test.f.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group6/test.f",
+ "components": [
+  "MODULE basic_mod",
+  "    TYPE :: person\n        CHARACTER(LEN=50) :: name\n        INTEGER :: age\n    END TYPE person",
+  "    SUBROUTINE short_hello(happy, path)\n    END SUBROUTINE short_hello",
+  "    SUBROUTINE long_hello(\n        p,\n        message\n    )\n    END SUBROUTINE long_hello",
+  "END MODULE basic_mod",
+  "PROGRAM HelloFortran\nEND PROGRAM HelloFortran"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__torch.rst.json b/tests/golden/legacy/components/tests__more_languages__group6__torch.rst.json
new file mode 100644
index 0000000..e1ddc0d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__torch.rst.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group6/torch.rst",
+ "components": [
+  "# libtorch (C++-only)",
+  "- Building libtorch using Python"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group6__yc.html.json b/tests/golden/legacy/components/tests__more_languages__group6__yc.html.json
new file mode 100644
index 0000000..db98f04
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group6__yc.html.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group6/yc.html",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group7__absurdly_huge.jsonl.json b/tests/golden/legacy/components/tests__more_languages__group7__absurdly_huge.jsonl.json
new file mode 100644
index 0000000..72267e5
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group7__absurdly_huge.jsonl.json
@@ -0,0 +1,14 @@
+{
+ "path": "tests/more_languages/group7/absurdly_huge.jsonl",
+ "components": [
+  "SMILES: str",
+  "Yield: float",
+  "Temperature: int",
+  "Pressure: float",
+  "Solvent: str",
+  "Success: bool",
+  "Reaction_Conditions: dict",
+  "Products: list",
+  "EdgeCasesMissed: None"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group7__angular_crud.ts.json b/tests/golden/legacy/components/tests__more_languages__group7__angular_crud.ts.json
new file mode 100644
index 0000000..f0009a9
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group7__angular_crud.ts.json
@@ -0,0 +1,18 @@
+{
+ "path": "tests/more_languages/group7/angular_crud.ts",
+ "components": [
+  "interface DBCommand<T = any>",
+  "export class IndexedDbService",
+  "    constructor()",
+  "    async create_connection({ db_name = 'client_db', table_name }: DBCommand)",
+  "                upgrade(db)",
+  "    async create_model({ db_name, table_name, model }: DBCommand)",
+  "        verify_matching({ table_name, model })",
+  "    async read_key({ db_name, table_name, key }: DBCommand)",
+  "    async update_model({ db_name, table_name, model }: DBCommand)",
+  "        verify_matching({ table_name, model })",
+  "    async delete_key({ db_name, table_name, key }: DBCommand)",
+  "    async list_table({\n        db_name,\n        table_name,\n        where,\n    }: DBCommand & { where?: { [key: string]: string | number } })",
+  "    async search_table(criteria: SearchCriteria)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group7__structure.py.json b/tests/golden/legacy/components/tests__more_languages__group7__structure.py.json
new file mode 100644
index 0000000..b5acd6c
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group7__structure.py.json
@@ -0,0 +1,35 @@
+{
+ "path": "tests/more_languages/group7/structure.py",
+ "components": [
+  "@runtime_checkable\nclass DataClass(Protocol)",
+  "    __dataclass_fields__: dict",
+  "class MyInteger(Enum)",
+  "    ONE = 1",
+  "    TWO = 2",
+  "    THREE = 42",
+  "class MyString(Enum)",
+  "    AAA1 = \"aaa\"",
+  "    BB_B = \"\"\"edge\ncase\"\"\"",
+  "@dataclass(frozen=True, slots=True, kw_only=True)\nclass Tool",
+  "    name: str",
+  "    description: str",
+  "    input_model: DataClass",
+  "    output_model: DataClass",
+  "    def execute(self, *args, **kwargs)",
+  "    @property\n    def edge_case(self) -> str",
+  "    def should_still_see_me(self, x: bool = True) -> \"Tool\"",
+  "@dataclass\nclass MyInput[T]",
+  "    name: str",
+  "    rank: MyInteger",
+  "    serial_n: int",
+  "@dataclass\nclass Thingy",
+  "    is_edge_case: bool",
+  "@dataclass\nclass MyOutput",
+  "    orders: str",
+  "class MyTools(Enum)",
+  "    TOOL_A = Tool(\n        name=\"complicated\",\n        description=\"edge case!\",\n        input_model=MyInput[Thingy],\n        output_model=MyOutput,\n    )",
+  "    TOOL_B = Tool(\n        name=\"\"\"super\ncomplicated\n\"\"\",\n        description=\"edge case!\",\n        input_model=MyInput,\n        output_model=MyOutput,\n    )",
+  "@final\nclass dtype(Generic[_DTypeScalar_co])",
+  "    names: None | tuple[builtins.str, ...]"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group7__test.metal.json b/tests/golden/legacy/components/tests__more_languages__group7__test.metal.json
new file mode 100644
index 0000000..a4b724d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group7__test.metal.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group7/test.metal",
+ "components": [
+  "struct MyData",
+  "kernel void myKernel(device MyData* data [[buffer(0)]],\n                     uint id [[thread_position_in_grid]])",
+  "float myHelperFunction(float x, float y)",
+  "vertex float4 vertexShader(const device packed_float3* vertex_array [[buffer(0)]],\n                           unsigned int vid [[vertex_id]])",
+  "fragment half4 fragmentShader(float4 P [[position]])",
+  "float3 computeNormalMap(ColorInOut in, texture2d<float> normalMapTexture)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group7__test.wgsl.json b/tests/golden/legacy/components/tests__more_languages__group7__test.wgsl.json
new file mode 100644
index 0000000..3e5ed6e
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group7__test.wgsl.json
@@ -0,0 +1,20 @@
+{
+ "path": "tests/more_languages/group7/test.wgsl",
+ "components": [
+  "alias MyVec = vec4<f32>",
+  "alias AnotherVec = vec2<f32>",
+  "struct VertexInput",
+  "struct VertexOutput",
+  "struct MyUniforms",
+  "@group(0) @binding(0) var<uniform> u_mvp: mat4x4<f32>",
+  "@group(0) @binding(1) var<uniform> u_color: MyVec",
+  "@group(1) @binding(0) var my_texture: texture_2d<f32>",
+  "@group(1) @binding(1) var my_sampler: sampler",
+  "@vertex\nfn vs_main(in: VertexInput) -> VertexOutput",
+  "@fragment\nfn fs_main(in: VertexOutput) -> @location(0) vec4<f32>",
+  "@compute @workgroup_size(8, 8, 1)\nfn cs_main(@builtin(global_invocation_id) global_id: vec3<u32>)",
+  "fn helper_function(val: f32) -> f32",
+  "struct AnotherStruct",
+  "@compute\n@workgroup_size(8, 8, 1)\nfn multi_line_edge_case(\n    @builtin(global_invocation_id)\n    globalId       : vec3<u32>,\n    @group(1)\n    @binding(0)\n    srcTexture     : texture_2d<f32>,\n    @group(1)\n    @binding(1)\n    srcSampler     : sampler,\n    @group(0)\n    @binding(0)\n    uniformsPtr    : ptr<uniform, MyUniforms>,\n    storageBuffer  : ptr<storage, array<vec4<f32>, 64>, read_write>,\n)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_lisp__LispTest.lisp.json b/tests/golden/legacy/components/tests__more_languages__group_lisp__LispTest.lisp.json
new file mode 100644
index 0000000..d0644e9
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_lisp__LispTest.lisp.json
@@ -0,0 +1,7 @@
+{
+ "path": "tests/more_languages/group_lisp/LispTest.lisp",
+ "components": [
+  "defstruct person",
+  "defun greet"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_lisp__clojure_test.clj.json b/tests/golden/legacy/components/tests__more_languages__group_lisp__clojure_test.clj.json
new file mode 100644
index 0000000..f804808
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_lisp__clojure_test.clj.json
@@ -0,0 +1,13 @@
+{
+ "path": "tests/more_languages/group_lisp/clojure_test.clj",
+ "components": [
+  "defprotocol P",
+  "defrecord Person",
+  "defn -main",
+  "ns bion.likes_trees",
+  "def repo-url",
+  "defn config",
+  "defmacro with-os",
+  "defrecord SetFullElement"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_lisp__racket_struct.rkt.json b/tests/golden/legacy/components/tests__more_languages__group_lisp__racket_struct.rkt.json
new file mode 100644
index 0000000..3b3ef8d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_lisp__racket_struct.rkt.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/more_languages/group_lisp/racket_struct.rkt",
+ "components": [
+  "struct point"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_lisp__test_scheme.scm.json b/tests/golden/legacy/components/tests__more_languages__group_lisp__test_scheme.scm.json
new file mode 100644
index 0000000..444dfd3
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_lisp__test_scheme.scm.json
@@ -0,0 +1,11 @@
+{
+ "path": "tests/more_languages/group_lisp/test_scheme.scm",
+ "components": [
+  "define topological-sort",
+  "  define table",
+  "  define queue",
+  "  define result",
+  "  define set-up",
+  "  define traverse"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__AAPLShaders.metal.json b/tests/golden/legacy/components/tests__more_languages__group_todo__AAPLShaders.metal.json
new file mode 100644
index 0000000..bfe72ca
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__AAPLShaders.metal.json
@@ -0,0 +1,30 @@
+{
+ "path": "tests/more_languages/group_todo/AAPLShaders.metal",
+ "components": [
+  "struct LightingParameters",
+  "float Geometry(float Ndotv, float alphaG)",
+  "float3 computeNormalMap(ColorInOut in, texture2d<float> normalMapTexture)",
+  "float3 computeDiffuse(LightingParameters parameters)",
+  "float Distribution(float NdotH, float roughness)",
+  "float3 computeSpecular(LightingParameters parameters)",
+  "float4 equirectangularSample(float3 direction, sampler s, texture2d<float> image)",
+  "LightingParameters calculateParameters(ColorInOut in,\n                                       AAPLCameraData cameraData,\n                                       constant AAPLLightData& lightData,\n                                       texture2d<float>   baseColorMap,\n                                       texture2d<float>   normalMap,\n                                       texture2d<float>   metallicMap,\n                                       texture2d<float>   roughnessMap,\n                                       texture2d<float>   ambientOcclusionMap,\n                                       texture2d<float>   skydomeMap)",
+  "struct SkyboxVertex",
+  "struct SkyboxV2F",
+  "vertex SkyboxV2F skyboxVertex(SkyboxVertex in [[stage_in]],\n                                 constant AAPLCameraData& cameraData [[buffer(BufferIndexCameraData)]])",
+  "fragment float4 skyboxFragment(SkyboxV2F v [[stage_in]], texture2d<float> skytexture [[texture(0)]])",
+  "vertex ColorInOut vertexShader(Vertex in [[stage_in]],\n                               constant AAPLInstanceTransform& instanceTransform [[ buffer(BufferIndexInstanceTransforms) ]],\n                               constant AAPLCameraData& cameraData [[ buffer(BufferIndexCameraData) ]])",
+  "float2 calculateScreenCoord( float3 ndcpos )",
+  "fragment float4 fragmentShader(\n                    ColorInOut                  in                    [[stage_in]],\n                    constant AAPLCameraData&    cameraData            [[ buffer(BufferIndexCameraData) ]],\n                    constant AAPLLightData&     lightData             [[ buffer(BufferIndexLightData) ]],\n                    constant AAPLSubmeshKeypath&submeshKeypath        [[ buffer(BufferIndexSubmeshKeypath)]],\n                    constant Scene*             pScene                [[ buffer(SceneIndex)]],\n                    texture2d<float>            skydomeMap            [[ texture(AAPLSkyDomeTexture) ]],\n                    texture2d<float>            rtReflections         [[ texture(AAPLTextureIndexReflections), function_constant(is_raytracing_enabled)]])",
+  "fragment float4 reflectionShader(ColorInOut in [[stage_in]],\n                                 texture2d<float> rtReflections [[texture(AAPLTextureIndexReflections)]])",
+  "struct ThinGBufferOut",
+  "fragment ThinGBufferOut gBufferFragmentShader(ColorInOut in [[stage_in]])",
+  "kernel void rtReflection(\n             texture2d< float, access::write >      outImage                [[texture(OutImageIndex)]],\n             texture2d< float >                     positions               [[texture(ThinGBufferPositionIndex)]],\n             texture2d< float >                     directions              [[texture(ThinGBufferDirectionIndex)]],\n             texture2d< float >                     skydomeMap              [[texture(AAPLSkyDomeTexture)]],\n             constant AAPLInstanceTransform*        instanceTransforms      [[buffer(BufferIndexInstanceTransforms)]],\n             constant AAPLCameraData&               cameraData              [[buffer(BufferIndexCameraData)]],\n             constant AAPLLightData&                lightData               [[buffer(BufferIndexLightData)]],\n             constant Scene*                        pScene                  [[buffer(SceneIndex)]],\n             instance_acceleration_structure        accelerationStructure   [[buffer(AccelerationStructureIndex)]],\n             uint2 tid [[thread_position_in_grid]])",
+  "else if ( intersection.type == raytracing::intersection_type::none )",
+  "struct VertexInOut",
+  "vertex VertexInOut vertexPassthrough( uint vid [[vertex_id]] )",
+  "fragment float4 fragmentPassthrough( VertexInOut in [[stage_in]], texture2d< float > tin )",
+  "fragment float4 fragmentBloomThreshold( VertexInOut in [[stage_in]],\n                                       texture2d< float > tin [[texture(0)]],\n                                       constant float* threshold [[buffer(0)]] )",
+  "fragment float4 fragmentPostprocessMerge( VertexInOut in [[stage_in]],\n                                         constant float& exposure [[buffer(0)]],\n                                         texture2d< float > texture0 [[texture(0)]],\n                                         texture2d< float > texture1 [[texture(1)]])"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__crystal_test.cr.json b/tests/golden/legacy/components/tests__more_languages__group_todo__crystal_test.cr.json
new file mode 100644
index 0000000..8a5ab21
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__crystal_test.cr.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/crystal_test.cr",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__dart_test.dart.json b/tests/golden/legacy/components/tests__more_languages__group_todo__dart_test.dart.json
new file mode 100644
index 0000000..6fdb845
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__dart_test.dart.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/dart_test.dart",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__elixir_test.exs.json b/tests/golden/legacy/components/tests__more_languages__group_todo__elixir_test.exs.json
new file mode 100644
index 0000000..6de993c
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__elixir_test.exs.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/elixir_test.exs",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__forward.frag.json b/tests/golden/legacy/components/tests__more_languages__group_todo__forward.frag.json
new file mode 100644
index 0000000..8494545
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__forward.frag.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/forward.frag",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__forward.vert.json b/tests/golden/legacy/components/tests__more_languages__group_todo__forward.vert.json
new file mode 100644
index 0000000..7d485fa
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__forward.vert.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/forward.vert",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__nodemon.json.json b/tests/golden/legacy/components/tests__more_languages__group_todo__nodemon.json.json
new file mode 100644
index 0000000..a566978
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__nodemon.json.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/nodemon.json",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__sas_test.sas.json b/tests/golden/legacy/components/tests__more_languages__group_todo__sas_test.sas.json
new file mode 100644
index 0000000..68f236a
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__sas_test.sas.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/sas_test.sas",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__testTypings.d.ts.json b/tests/golden/legacy/components/tests__more_languages__group_todo__testTypings.d.ts.json
new file mode 100644
index 0000000..bf27acc
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__testTypings.d.ts.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/testTypings.d.ts",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__test_setup_py.test.json b/tests/golden/legacy/components/tests__more_languages__group_todo__test_setup_py.test.json
new file mode 100644
index 0000000..2ece290
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__test_setup_py.test.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/test_setup_py.test",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__vba_test.bas.json b/tests/golden/legacy/components/tests__more_languages__group_todo__vba_test.bas.json
new file mode 100644
index 0000000..758031d
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__vba_test.bas.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/more_languages/group_todo/vba_test.bas",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__more_languages__group_todo__wgsl_test.wgsl.json b/tests/golden/legacy/components/tests__more_languages__group_todo__wgsl_test.wgsl.json
new file mode 100644
index 0000000..91e72f5
--- /dev/null
+++ b/tests/golden/legacy/components/tests__more_languages__group_todo__wgsl_test.wgsl.json
@@ -0,0 +1,8 @@
+{
+ "path": "tests/more_languages/group_todo/wgsl_test.wgsl",
+ "components": [
+  "@binding(0) @group(0) var<uniform> frame : u32",
+  "@vertex\nfn vtx_main(@builtin(vertex_index) vertex_index : u32) -> @builtin(position) vec4f",
+  "@fragment\nfn frag_main() -> @location(0) vec4f"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__path_to_test__class_method_type.py.json b/tests/golden/legacy/components/tests__path_to_test__class_method_type.py.json
new file mode 100644
index 0000000..6faaf76
--- /dev/null
+++ b/tests/golden/legacy/components/tests__path_to_test__class_method_type.py.json
@@ -0,0 +1,33 @@
+{
+ "path": "tests/path_to_test/class_method_type.py",
+ "components": [
+  "T = TypeVar(\"T\")",
+  "def parse_py(contents: str) -> List[str]",
+  "class MyClass",
+  "    @staticmethod\n    def physical_element_aval(dtype) -> core.ShapedArray",
+  "    def my_method(self)",
+  "    @staticmethod\n    def my_typed_method(obj: dict) -> int",
+  "    def my_multiline_signature_method(\n        self,\n        alice: str = None,\n        bob: int = None,\n    ) -> tuple",
+  "@lru_cache(maxsize=None)\ndef my_multiline_signature_function(\n    tree: tuple = (),\n    plus: str = \"+\",\n) -> tuple",
+  "class LogLevelEnum(str, Enum)",
+  "    CRITICAL = \"CRITICAL\"",
+  "    GREETING = \"GREETING\"",
+  "    WARNING = \"WARNING\"",
+  "    ERROR = \"ERROR\"",
+  "    DEBUG = \"DEBUG\"",
+  "    INFO = \"INFO\"",
+  "    OFF = \"OFF\"",
+  "class Thingy(BaseModel)",
+  "    metric: float",
+  "@dataclass\nclass TestDataclass",
+  "    tree: str",
+  "A = TypeVar(\"A\", str, bytes)",
+  "def omega_yikes(file: str, expected: List[str]) -> bool",
+  "def ice[T](args: Iterable[T] = ())",
+  "class list[T]",
+  "    def __getitem__(self, index: int, /) -> T",
+  "    @classmethod\n    def from_code(cls, toolbox, code: bytes, score=None) -> \"Thingy\"",
+  "    @classmethod\n    def from_str(cls, toolbox, string: str, score=None) -> \"Thingy\"",
+  "class Router(hk.Module)"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__path_to_test__empty.py.json b/tests/golden/legacy/components/tests__path_to_test__empty.py.json
new file mode 100644
index 0000000..c297b5f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__path_to_test__empty.py.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/path_to_test/empty.py",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__path_to_test__file.md.json b/tests/golden/legacy/components/tests__path_to_test__file.md.json
new file mode 100644
index 0000000..8da3d5f
--- /dev/null
+++ b/tests/golden/legacy/components/tests__path_to_test__file.md.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/path_to_test/file.md",
+ "components": [
+  "# Hello, world!"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__path_to_test__file.py.json b/tests/golden/legacy/components/tests__path_to_test__file.py.json
new file mode 100644
index 0000000..479f3b5
--- /dev/null
+++ b/tests/golden/legacy/components/tests__path_to_test__file.py.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/path_to_test/file.py",
+ "components": [
+  "def hello_world()"
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__path_to_test__file.txt.json b/tests/golden/legacy/components/tests__path_to_test__file.txt.json
new file mode 100644
index 0000000..40c85e2
--- /dev/null
+++ b/tests/golden/legacy/components/tests__path_to_test__file.txt.json
@@ -0,0 +1,4 @@
+{
+ "path": "tests/path_to_test/file.txt",
+ "components": []
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/components/tests__path_to_test__version.py.json b/tests/golden/legacy/components/tests__path_to_test__version.py.json
new file mode 100644
index 0000000..0383bdb
--- /dev/null
+++ b/tests/golden/legacy/components/tests__path_to_test__version.py.json
@@ -0,0 +1,6 @@
+{
+ "path": "tests/path_to_test/version.py",
+ "components": [
+  "__version__ = \"1.2.3\""
+ ]
+}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__dot_dot__my_test_file.py.json b/tests/golden/legacy/counts/tests__dot_dot__my_test_file.py.json
new file mode 100644
index 0000000..420bcb8
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__dot_dot__my_test_file.py.json
@@ -0,0 +1 @@
+{"path": "tests/dot_dot/my_test_file.py", "count": {"n_tokens": 7, "n_lines": 2}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__dot_dot__nested_dir__.env.test.json b/tests/golden/legacy/counts/tests__dot_dot__nested_dir__.env.test.json
new file mode 100644
index 0000000..e978f74
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__dot_dot__nested_dir__.env.test.json
@@ -0,0 +1 @@
+{"path": "tests/dot_dot/nested_dir/.env.test", "count": {"n_tokens": 4, "n_lines": 0}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__dot_dot__nested_dir__pytest.ini.json b/tests/golden/legacy/counts/tests__dot_dot__nested_dir__pytest.ini.json
new file mode 100644
index 0000000..9014079
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__dot_dot__nested_dir__pytest.ini.json
@@ -0,0 +1 @@
+{"path": "tests/dot_dot/nested_dir/pytest.ini", "count": {"n_tokens": 20, "n_lines": 4}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__dot_dot__nested_dir__test_tp_dotdot.py.json b/tests/golden/legacy/counts/tests__dot_dot__nested_dir__test_tp_dotdot.py.json
new file mode 100644
index 0000000..a7a8ceb
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__dot_dot__nested_dir__test_tp_dotdot.py.json
@@ -0,0 +1 @@
+{"path": "tests/dot_dot/nested_dir/test_tp_dotdot.py", "count": {"n_tokens": 362, "n_lines": 52}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json b/tests/golden/legacy/counts/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
new file mode 100644
index 0000000..94c34f9
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/CUSTOMER-INVOICE.CBL", "count": {"n_tokens": 412, "n_lines": 60}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__JavaTest.java.json b/tests/golden/legacy/counts/tests__more_languages__group1__JavaTest.java.json
new file mode 100644
index 0000000..d609ed8
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__JavaTest.java.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/JavaTest.java", "count": {"n_tokens": 578, "n_lines": 86}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__JuliaTest.jl.json b/tests/golden/legacy/counts/tests__more_languages__group1__JuliaTest.jl.json
new file mode 100644
index 0000000..d80c234
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__JuliaTest.jl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/JuliaTest.jl", "count": {"n_tokens": 381, "n_lines": 63}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__KotlinTest.kt.json b/tests/golden/legacy/counts/tests__more_languages__group1__KotlinTest.kt.json
new file mode 100644
index 0000000..a3ecb30
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__KotlinTest.kt.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/KotlinTest.kt", "count": {"n_tokens": 974, "n_lines": 171}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__LuaTest.lua.json b/tests/golden/legacy/counts/tests__more_languages__group1__LuaTest.lua.json
new file mode 100644
index 0000000..3cfe8ec
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__LuaTest.lua.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/LuaTest.lua", "count": {"n_tokens": 83, "n_lines": 16}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__ObjectiveCTest.m.json b/tests/golden/legacy/counts/tests__more_languages__group1__ObjectiveCTest.m.json
new file mode 100644
index 0000000..aeeab27
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__ObjectiveCTest.m.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/ObjectiveCTest.m", "count": {"n_tokens": 62, "n_lines": 16}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__OcamlTest.ml.json b/tests/golden/legacy/counts/tests__more_languages__group1__OcamlTest.ml.json
new file mode 100644
index 0000000..9746ea6
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__OcamlTest.ml.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/OcamlTest.ml", "count": {"n_tokens": 49, "n_lines": 12}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__addamt.cobol.json b/tests/golden/legacy/counts/tests__more_languages__group1__addamt.cobol.json
new file mode 100644
index 0000000..1771956
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__addamt.cobol.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/addamt.cobol", "count": {"n_tokens": 441, "n_lines": 40}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__lesson.cbl.json b/tests/golden/legacy/counts/tests__more_languages__group1__lesson.cbl.json
new file mode 100644
index 0000000..6d5f22c
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__lesson.cbl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/lesson.cbl", "count": {"n_tokens": 635, "n_lines": 78}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__test.js.json b/tests/golden/legacy/counts/tests__more_languages__group1__test.js.json
new file mode 100644
index 0000000..f36977c
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__test.js.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/test.js", "count": {"n_tokens": 757, "n_lines": 154}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group1__test.ts.json b/tests/golden/legacy/counts/tests__more_languages__group1__test.ts.json
new file mode 100644
index 0000000..1e61cdc
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group1__test.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group1/test.ts", "count": {"n_tokens": 832, "n_lines": 165}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__PerlTest.pl.json b/tests/golden/legacy/counts/tests__more_languages__group2__PerlTest.pl.json
new file mode 100644
index 0000000..ffbf1c6
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__PerlTest.pl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/PerlTest.pl", "count": {"n_tokens": 63, "n_lines": 20}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__PhpTest.php.json b/tests/golden/legacy/counts/tests__more_languages__group2__PhpTest.php.json
new file mode 100644
index 0000000..04772fc
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__PhpTest.php.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/PhpTest.php", "count": {"n_tokens": 70, "n_lines": 19}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__PowershellTest.ps1.json b/tests/golden/legacy/counts/tests__more_languages__group2__PowershellTest.ps1.json
new file mode 100644
index 0000000..2b51259
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__PowershellTest.ps1.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/PowershellTest.ps1", "count": {"n_tokens": 459, "n_lines": 89}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__ScalaTest.scala.json b/tests/golden/legacy/counts/tests__more_languages__group2__ScalaTest.scala.json
new file mode 100644
index 0000000..e56d238
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__ScalaTest.scala.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/ScalaTest.scala", "count": {"n_tokens": 171, "n_lines": 40}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__apl_test.apl.json b/tests/golden/legacy/counts/tests__more_languages__group2__apl_test.apl.json
new file mode 100644
index 0000000..bffd8b0
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__apl_test.apl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/apl_test.apl", "count": {"n_tokens": 28, "n_lines": 5}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__c_test.c.json b/tests/golden/legacy/counts/tests__more_languages__group2__c_test.c.json
new file mode 100644
index 0000000..4a7a6b3
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__c_test.c.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/c_test.c", "count": {"n_tokens": 837, "n_lines": 142}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__go_test.go.json b/tests/golden/legacy/counts/tests__more_languages__group2__go_test.go.json
new file mode 100644
index 0000000..5dbdc0c
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__go_test.go.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/go_test.go", "count": {"n_tokens": 179, "n_lines": 46}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group2__test.csv.json b/tests/golden/legacy/counts/tests__more_languages__group2__test.csv.json
new file mode 100644
index 0000000..cde6a1f
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group2__test.csv.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group2/test.csv", "count": null}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__bash_test.sh.json b/tests/golden/legacy/counts/tests__more_languages__group3__bash_test.sh.json
new file mode 100644
index 0000000..dde9049
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__bash_test.sh.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/bash_test.sh", "count": {"n_tokens": 127, "n_lines": 22}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__cpp_test.cpp.json b/tests/golden/legacy/counts/tests__more_languages__group3__cpp_test.cpp.json
new file mode 100644
index 0000000..aa377d1
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__cpp_test.cpp.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/cpp_test.cpp", "count": {"n_tokens": 1670, "n_lines": 259}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__csharp_test.cs.json b/tests/golden/legacy/counts/tests__more_languages__group3__csharp_test.cs.json
new file mode 100644
index 0000000..8b73841
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__csharp_test.cs.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/csharp_test.cs", "count": {"n_tokens": 957, "n_lines": 146}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__hallucination.tex.json b/tests/golden/legacy/counts/tests__more_languages__group3__hallucination.tex.json
new file mode 100644
index 0000000..d3986c7
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__hallucination.tex.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/hallucination.tex", "count": {"n_tokens": 1633, "n_lines": 126}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__ruby_test.rb.json b/tests/golden/legacy/counts/tests__more_languages__group3__ruby_test.rb.json
new file mode 100644
index 0000000..fccfc0c
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__ruby_test.rb.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/ruby_test.rb", "count": {"n_tokens": 138, "n_lines": 37}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__swift_test.swift.json b/tests/golden/legacy/counts/tests__more_languages__group3__swift_test.swift.json
new file mode 100644
index 0000000..09aabd8
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__swift_test.swift.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/swift_test.swift", "count": {"n_tokens": 469, "n_lines": 110}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test.capnp.json b/tests/golden/legacy/counts/tests__more_languages__group3__test.capnp.json
new file mode 100644
index 0000000..ca14f50
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test.capnp.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test.capnp", "count": {"n_tokens": 117, "n_lines": 30}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test.graphql.json b/tests/golden/legacy/counts/tests__more_languages__group3__test.graphql.json
new file mode 100644
index 0000000..f4b1aa9
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test.graphql.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test.graphql", "count": {"n_tokens": 66, "n_lines": 21}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test.lean.json b/tests/golden/legacy/counts/tests__more_languages__group3__test.lean.json
new file mode 100644
index 0000000..0a5ade6
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test.lean.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test.lean", "count": {"n_tokens": 289, "n_lines": 42}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test.proto.json b/tests/golden/legacy/counts/tests__more_languages__group3__test.proto.json
new file mode 100644
index 0000000..3443075
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test.proto.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test.proto", "count": {"n_tokens": 142, "n_lines": 34}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test.sqlite.json b/tests/golden/legacy/counts/tests__more_languages__group3__test.sqlite.json
new file mode 100644
index 0000000..799c090
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test.sqlite.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test.sqlite", "count": null}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test_Cargo.toml.json b/tests/golden/legacy/counts/tests__more_languages__group3__test_Cargo.toml.json
new file mode 100644
index 0000000..0a5a448
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test_Cargo.toml.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test_Cargo.toml", "count": {"n_tokens": 119, "n_lines": 18}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test_json_rpc_2_0.json.json b/tests/golden/legacy/counts/tests__more_languages__group3__test_json_rpc_2_0.json.json
new file mode 100644
index 0000000..91d3368
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test_json_rpc_2_0.json.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test_json_rpc_2_0.json", "count": {"n_tokens": 26, "n_lines": 6}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test_openapi.yaml.json b/tests/golden/legacy/counts/tests__more_languages__group3__test_openapi.yaml.json
new file mode 100644
index 0000000..d9a8384
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test_openapi.yaml.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test_openapi.yaml", "count": {"n_tokens": 753, "n_lines": 92}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test_openrpc.json.json b/tests/golden/legacy/counts/tests__more_languages__group3__test_openrpc.json.json
new file mode 100644
index 0000000..e8242e1
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test_openrpc.json.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test_openrpc.json", "count": {"n_tokens": 225, "n_lines": 44}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group3__test_pyproject.toml.json b/tests/golden/legacy/counts/tests__more_languages__group3__test_pyproject.toml.json
new file mode 100644
index 0000000..3bba2c7
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group3__test_pyproject.toml.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group3/test_pyproject.toml", "count": {"n_tokens": 304, "n_lines": 39}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__RTest.R.json b/tests/golden/legacy/counts/tests__more_languages__group4__RTest.R.json
new file mode 100644
index 0000000..5ed60a1
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__RTest.R.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/RTest.R", "count": {"n_tokens": 367, "n_lines": 46}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__erl_test.erl.json b/tests/golden/legacy/counts/tests__more_languages__group4__erl_test.erl.json
new file mode 100644
index 0000000..24cbc9b
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__erl_test.erl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/erl_test.erl", "count": {"n_tokens": 480, "n_lines": 68}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__haskell_test.hs.json b/tests/golden/legacy/counts/tests__more_languages__group4__haskell_test.hs.json
new file mode 100644
index 0000000..834fc76
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__haskell_test.hs.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/haskell_test.hs", "count": {"n_tokens": 414, "n_lines": 41}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__mathematica_test.nb.json b/tests/golden/legacy/counts/tests__more_languages__group4__mathematica_test.nb.json
new file mode 100644
index 0000000..78d9c43
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__mathematica_test.nb.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/mathematica_test.nb", "count": {"n_tokens": 133, "n_lines": 21}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__matlab_test.m.json b/tests/golden/legacy/counts/tests__more_languages__group4__matlab_test.m.json
new file mode 100644
index 0000000..eac0086
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__matlab_test.m.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/matlab_test.m", "count": {"n_tokens": 48, "n_lines": 12}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__rust_test.rs.json b/tests/golden/legacy/counts/tests__more_languages__group4__rust_test.rs.json
new file mode 100644
index 0000000..1de30f6
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__rust_test.rs.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/rust_test.rs", "count": {"n_tokens": 1368, "n_lines": 259}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__test.zig.json b/tests/golden/legacy/counts/tests__more_languages__group4__test.zig.json
new file mode 100644
index 0000000..ecfb7f3
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__test.zig.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/test.zig", "count": {"n_tokens": 397, "n_lines": 60}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__test_fsharp.fs.json b/tests/golden/legacy/counts/tests__more_languages__group4__test_fsharp.fs.json
new file mode 100644
index 0000000..f02c099
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__test_fsharp.fs.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/test_fsharp.fs", "count": {"n_tokens": 92, "n_lines": 27}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__test_tcl_tk.tcl.json b/tests/golden/legacy/counts/tests__more_languages__group4__test_tcl_tk.tcl.json
new file mode 100644
index 0000000..7dc2a06
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__test_tcl_tk.tcl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/test_tcl_tk.tcl", "count": {"n_tokens": 54, "n_lines": 16}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group4__tf_test.tf.json b/tests/golden/legacy/counts/tests__more_languages__group4__tf_test.tf.json
new file mode 100644
index 0000000..79a7efe
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group4__tf_test.tf.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group4/tf_test.tf", "count": {"n_tokens": 202, "n_lines": 38}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__Makefile.json b/tests/golden/legacy/counts/tests__more_languages__group5__Makefile.json
new file mode 100644
index 0000000..6d78ef9
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__Makefile.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/Makefile", "count": {"n_tokens": 714, "n_lines": 84}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__ansible_test.yml.json b/tests/golden/legacy/counts/tests__more_languages__group5__ansible_test.yml.json
new file mode 100644
index 0000000..849c20e
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__ansible_test.yml.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/ansible_test.yml", "count": {"n_tokens": 55, "n_lines": 14}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__app-routing.module.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__app-routing.module.ts.json
new file mode 100644
index 0000000..b87fca8
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__app-routing.module.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/app-routing.module.ts", "count": {"n_tokens": 287, "n_lines": 28}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__app.component.spec.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__app.component.spec.ts.json
new file mode 100644
index 0000000..ee298bf
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__app.component.spec.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/app.component.spec.ts", "count": {"n_tokens": 410, "n_lines": 47}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__app.component.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__app.component.ts.json
new file mode 100644
index 0000000..13f3ef0
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__app.component.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/app.component.ts", "count": {"n_tokens": 271, "n_lines": 45}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__app.module.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__app.module.ts.json
new file mode 100644
index 0000000..e07b2b5
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__app.module.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/app.module.ts", "count": {"n_tokens": 374, "n_lines": 43}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.md.json b/tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.md.json
new file mode 100644
index 0000000..ff3ddd4
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.md.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/checkbox_test.md", "count": {"n_tokens": 191, "n_lines": 29}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.txt.json b/tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.txt.json
new file mode 100644
index 0000000..d284fb3
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__checkbox_test.txt.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/checkbox_test.txt", "count": {"n_tokens": 257, "n_lines": 33}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__environment.test.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__environment.test.ts.json
new file mode 100644
index 0000000..cc69a6f
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__environment.test.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/environment.test.ts", "count": {"n_tokens": 197, "n_lines": 19}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__hello_world.pyi.json b/tests/golden/legacy/counts/tests__more_languages__group5__hello_world.pyi.json
new file mode 100644
index 0000000..437c324
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__hello_world.pyi.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/hello_world.pyi", "count": {"n_tokens": 22, "n_lines": 3}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__k8s_test.yaml.json b/tests/golden/legacy/counts/tests__more_languages__group5__k8s_test.yaml.json
new file mode 100644
index 0000000..4292243
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__k8s_test.yaml.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/k8s_test.yaml", "count": {"n_tokens": 140, "n_lines": 37}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__requirements_test.txt.json b/tests/golden/legacy/counts/tests__more_languages__group5__requirements_test.txt.json
new file mode 100644
index 0000000..181e2dc
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__requirements_test.txt.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/requirements_test.txt", "count": {"n_tokens": 29, "n_lines": 10}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__rust_todo_test.rs.json b/tests/golden/legacy/counts/tests__more_languages__group5__rust_todo_test.rs.json
new file mode 100644
index 0000000..1abd88d
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__rust_todo_test.rs.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/rust_todo_test.rs", "count": {"n_tokens": 92, "n_lines": 26}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__sql_test.sql.json b/tests/golden/legacy/counts/tests__more_languages__group5__sql_test.sql.json
new file mode 100644
index 0000000..24de39a
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__sql_test.sql.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/sql_test.sql", "count": {"n_tokens": 270, "n_lines": 51}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__standard-app-routing.module.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__standard-app-routing.module.ts.json
new file mode 100644
index 0000000..09a2581
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__standard-app-routing.module.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/standard-app-routing.module.ts", "count": {"n_tokens": 100, "n_lines": 16}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__test.env.json b/tests/golden/legacy/counts/tests__more_languages__group5__test.env.json
new file mode 100644
index 0000000..f730c12
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__test.env.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/test.env", "count": {"n_tokens": 190, "n_lines": 25}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__testJsonSchema.json.json b/tests/golden/legacy/counts/tests__more_languages__group5__testJsonSchema.json.json
new file mode 100644
index 0000000..9b32c46
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__testJsonSchema.json.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/testJsonSchema.json", "count": {"n_tokens": 421, "n_lines": 48}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__testPackage.json.json b/tests/golden/legacy/counts/tests__more_languages__group5__testPackage.json.json
new file mode 100644
index 0000000..506d040
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__testPackage.json.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/testPackage.json", "count": {"n_tokens": 349, "n_lines": 43}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group5__tickets.component.ts.json b/tests/golden/legacy/counts/tests__more_languages__group5__tickets.component.ts.json
new file mode 100644
index 0000000..6a1a7e4
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group5__tickets.component.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group5/tickets.component.ts", "count": {"n_tokens": 7160, "n_lines": 903}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json b/tests/golden/legacy/counts/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json
new file mode 100644
index 0000000..a44b201
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__Microsoft.PowerShell_profile.ps1.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/Microsoft.PowerShell_profile.ps1", "count": {"n_tokens": 3346, "n_lines": 497}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__catastrophic.c.json b/tests/golden/legacy/counts/tests__more_languages__group6__catastrophic.c.json
new file mode 100644
index 0000000..b06ea26
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__catastrophic.c.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/catastrophic.c", "count": {"n_tokens": 5339, "n_lines": 754}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cc.json b/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cc.json
new file mode 100644
index 0000000..74a36c0
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cc.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/cpp_examples_impl.cc", "count": {"n_tokens": 60, "n_lines": 10}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cu.json b/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cu.json
new file mode 100644
index 0000000..d3d35d0
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.cu.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/cpp_examples_impl.cu", "count": {"n_tokens": 37, "n_lines": 10}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.h.json b/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.h.json
new file mode 100644
index 0000000..87ff03e
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__cpp_examples_impl.h.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/cpp_examples_impl.h", "count": {"n_tokens": 22, "n_lines": 6}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__edge_case.hpp.json b/tests/golden/legacy/counts/tests__more_languages__group6__edge_case.hpp.json
new file mode 100644
index 0000000..fd7eec6
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__edge_case.hpp.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/edge_case.hpp", "count": {"n_tokens": 426, "n_lines": 28}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__fractal.thy.json b/tests/golden/legacy/counts/tests__more_languages__group6__fractal.thy.json
new file mode 100644
index 0000000..0c10d28
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__fractal.thy.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/fractal.thy", "count": {"n_tokens": 1712, "n_lines": 147}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__python_complex_class.py.json b/tests/golden/legacy/counts/tests__more_languages__group6__python_complex_class.py.json
new file mode 100644
index 0000000..552722b
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__python_complex_class.py.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/python_complex_class.py", "count": {"n_tokens": 10, "n_lines": 2}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__ramda__cloneRegExp.js.json b/tests/golden/legacy/counts/tests__more_languages__group6__ramda__cloneRegExp.js.json
new file mode 100644
index 0000000..69e6c0e
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__ramda__cloneRegExp.js.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/ramda__cloneRegExp.js", "count": {"n_tokens": 173, "n_lines": 9}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__ramda_prop.js.json b/tests/golden/legacy/counts/tests__more_languages__group6__ramda_prop.js.json
new file mode 100644
index 0000000..617c804
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__ramda_prop.js.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/ramda_prop.js", "count": {"n_tokens": 646, "n_lines": 85}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__tensorflow_flags.h.json b/tests/golden/legacy/counts/tests__more_languages__group6__tensorflow_flags.h.json
new file mode 100644
index 0000000..8ec45ac
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__tensorflow_flags.h.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/tensorflow_flags.h", "count": {"n_tokens": 7628, "n_lines": 668}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__test.f.json b/tests/golden/legacy/counts/tests__more_languages__group6__test.f.json
new file mode 100644
index 0000000..3342682
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__test.f.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/test.f", "count": {"n_tokens": 181, "n_lines": 30}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__torch.rst.json b/tests/golden/legacy/counts/tests__more_languages__group6__torch.rst.json
new file mode 100644
index 0000000..e45381d
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__torch.rst.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/torch.rst", "count": {"n_tokens": 60, "n_lines": 8}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group6__yc.html.json b/tests/golden/legacy/counts/tests__more_languages__group6__yc.html.json
new file mode 100644
index 0000000..6cf052f
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group6__yc.html.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group6/yc.html", "count": {"n_tokens": 9063, "n_lines": 169}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group7__absurdly_huge.jsonl.json b/tests/golden/legacy/counts/tests__more_languages__group7__absurdly_huge.jsonl.json
new file mode 100644
index 0000000..14e1246
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group7__absurdly_huge.jsonl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group7/absurdly_huge.jsonl", "count": {"n_tokens": 8347, "n_lines": 126}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group7__angular_crud.ts.json b/tests/golden/legacy/counts/tests__more_languages__group7__angular_crud.ts.json
new file mode 100644
index 0000000..7b415c4
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group7__angular_crud.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group7/angular_crud.ts", "count": {"n_tokens": 1192, "n_lines": 148}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group7__structure.py.json b/tests/golden/legacy/counts/tests__more_languages__group7__structure.py.json
new file mode 100644
index 0000000..a72b9aa
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group7__structure.py.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group7/structure.py", "count": {"n_tokens": 400, "n_lines": 92}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group7__test.metal.json b/tests/golden/legacy/counts/tests__more_languages__group7__test.metal.json
new file mode 100644
index 0000000..3ccf32a
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group7__test.metal.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group7/test.metal", "count": {"n_tokens": 272, "n_lines": 34}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group7__test.wgsl.json b/tests/golden/legacy/counts/tests__more_languages__group7__test.wgsl.json
new file mode 100644
index 0000000..6abca09
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group7__test.wgsl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group7/test.wgsl", "count": {"n_tokens": 528, "n_lines": 87}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_lisp__LispTest.lisp.json b/tests/golden/legacy/counts/tests__more_languages__group_lisp__LispTest.lisp.json
new file mode 100644
index 0000000..d20eb39
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_lisp__LispTest.lisp.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_lisp/LispTest.lisp", "count": {"n_tokens": 25, "n_lines": 6}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_lisp__clojure_test.clj.json b/tests/golden/legacy/counts/tests__more_languages__group_lisp__clojure_test.clj.json
new file mode 100644
index 0000000..f967933
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_lisp__clojure_test.clj.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_lisp/clojure_test.clj", "count": {"n_tokens": 682, "n_lines": 85}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_lisp__racket_struct.rkt.json b/tests/golden/legacy/counts/tests__more_languages__group_lisp__racket_struct.rkt.json
new file mode 100644
index 0000000..caef743
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_lisp__racket_struct.rkt.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_lisp/racket_struct.rkt", "count": {"n_tokens": 14, "n_lines": 1}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_lisp__test_scheme.scm.json b/tests/golden/legacy/counts/tests__more_languages__group_lisp__test_scheme.scm.json
new file mode 100644
index 0000000..79f8691
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_lisp__test_scheme.scm.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_lisp/test_scheme.scm", "count": {"n_tokens": 360, "n_lines": 44}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__AAPLShaders.metal.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__AAPLShaders.metal.json
new file mode 100644
index 0000000..b2ed24a
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__AAPLShaders.metal.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/AAPLShaders.metal", "count": {"n_tokens": 5780, "n_lines": 566}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__crystal_test.cr.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__crystal_test.cr.json
new file mode 100644
index 0000000..994d649
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__crystal_test.cr.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/crystal_test.cr", "count": {"n_tokens": 48, "n_lines": 15}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__dart_test.dart.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__dart_test.dart.json
new file mode 100644
index 0000000..12befcf
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__dart_test.dart.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/dart_test.dart", "count": {"n_tokens": 108, "n_lines": 24}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__elixir_test.exs.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__elixir_test.exs.json
new file mode 100644
index 0000000..e733a3d
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__elixir_test.exs.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/elixir_test.exs", "count": {"n_tokens": 39, "n_lines": 10}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__forward.frag.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__forward.frag.json
new file mode 100644
index 0000000..7cb73d6
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__forward.frag.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/forward.frag", "count": {"n_tokens": 739, "n_lines": 87}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__forward.vert.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__forward.vert.json
new file mode 100644
index 0000000..61ce20b
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__forward.vert.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/forward.vert", "count": {"n_tokens": 359, "n_lines": 48}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__nodemon.json.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__nodemon.json.json
new file mode 100644
index 0000000..f1517fa
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__nodemon.json.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/nodemon.json", "count": {"n_tokens": 118, "n_lines": 20}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__sas_test.sas.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__sas_test.sas.json
new file mode 100644
index 0000000..409afe8
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__sas_test.sas.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/sas_test.sas", "count": {"n_tokens": 97, "n_lines": 22}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__testTypings.d.ts.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__testTypings.d.ts.json
new file mode 100644
index 0000000..b9fceca
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__testTypings.d.ts.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/testTypings.d.ts", "count": {"n_tokens": 158, "n_lines": 23}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__test_setup_py.test.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__test_setup_py.test.json
new file mode 100644
index 0000000..6d094ea
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__test_setup_py.test.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/test_setup_py.test", "count": {"n_tokens": 133, "n_lines": 24}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__vba_test.bas.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__vba_test.bas.json
new file mode 100644
index 0000000..eb12356
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__vba_test.bas.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/vba_test.bas", "count": {"n_tokens": 67, "n_lines": 16}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__more_languages__group_todo__wgsl_test.wgsl.json b/tests/golden/legacy/counts/tests__more_languages__group_todo__wgsl_test.wgsl.json
new file mode 100644
index 0000000..5cff913
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__more_languages__group_todo__wgsl_test.wgsl.json
@@ -0,0 +1 @@
+{"path": "tests/more_languages/group_todo/wgsl_test.wgsl", "count": {"n_tokens": 94, "n_lines": 17}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__path_to_test__class_method_type.py.json b/tests/golden/legacy/counts/tests__path_to_test__class_method_type.py.json
new file mode 100644
index 0000000..3f64fea
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__path_to_test__class_method_type.py.json
@@ -0,0 +1 @@
+{"path": "tests/path_to_test/class_method_type.py", "count": {"n_tokens": 525, "n_lines": 101}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__path_to_test__empty.py.json b/tests/golden/legacy/counts/tests__path_to_test__empty.py.json
new file mode 100644
index 0000000..473ba6d
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__path_to_test__empty.py.json
@@ -0,0 +1 @@
+{"path": "tests/path_to_test/empty.py", "count": {"n_tokens": 0, "n_lines": 0}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__path_to_test__file.md.json b/tests/golden/legacy/counts/tests__path_to_test__file.md.json
new file mode 100644
index 0000000..a7af78b
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__path_to_test__file.md.json
@@ -0,0 +1 @@
+{"path": "tests/path_to_test/file.md", "count": {"n_tokens": 11, "n_lines": 2}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__path_to_test__file.py.json b/tests/golden/legacy/counts/tests__path_to_test__file.py.json
new file mode 100644
index 0000000..241f305
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__path_to_test__file.py.json
@@ -0,0 +1 @@
+{"path": "tests/path_to_test/file.py", "count": {"n_tokens": 18, "n_lines": 3}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__path_to_test__file.txt.json b/tests/golden/legacy/counts/tests__path_to_test__file.txt.json
new file mode 100644
index 0000000..d14312a
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__path_to_test__file.txt.json
@@ -0,0 +1 @@
+{"path": "tests/path_to_test/file.txt", "count": {"n_tokens": 10, "n_lines": 2}}
\ No newline at end of file
diff --git a/tests/golden/legacy/counts/tests__path_to_test__version.py.json b/tests/golden/legacy/counts/tests__path_to_test__version.py.json
new file mode 100644
index 0000000..6270cd3
--- /dev/null
+++ b/tests/golden/legacy/counts/tests__path_to_test__version.py.json
@@ -0,0 +1 @@
+{"path": "tests/path_to_test/version.py", "count": {"n_tokens": 13, "n_lines": 2}}
\ No newline at end of file
diff --git a/tests/golden/legacy/trees/dot_dot.txt b/tests/golden/legacy/trees/dot_dot.txt
new file mode 100644
index 0000000..d6db06d
--- /dev/null
+++ b/tests/golden/legacy/trees/dot_dot.txt
@@ -0,0 +1,10 @@
+📁 dot_dot (2 folders, 4 files)
+├── 📄 my_test_file.py (7 tokens, 2 lines)
+│   └── def dot_dot_dot()
+└── 📁 nested_dir (1 folder, 3 files)
+    ├── 📄 .env.test (4 tokens, 0 lines)
+    │   └── DEBUG_TREE_PLUS
+    ├── 📄 pytest.ini (20 tokens, 4 lines)
+    └── 📄 test_tp_dotdot.py (362 tokens, 52 lines)
+        ├── def ignore_tokens_lines_test(text: str) -> str
+        └── def test_tree_plus_dotdot()
diff --git a/tests/golden/legacy/trees/more_languages.txt b/tests/golden/legacy/trees/more_languages.txt
new file mode 100644
index 0000000..f199132
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages.txt
@@ -0,0 +1,2225 @@
+📁 more_languages (10 folders, 99 files)
+├── 📁 group1 (1 folder, 11 files)
+│   ├── 📄 addamt.cobol (441 tokens, 40 lines)
+│   │   ├── IDENTIFICATION DIVISION.
+│   │   ├── PROGRAM-ID.
+│   │   │              ADDAMT.
+│   │   ├── DATA DIVISION.
+│   │   ├── WORKING-STORAGE SECTION.
+│   │   ├── 01  KEYED-INPUT.
+│   │   ├──     05  CUST-NO-IN.
+│   │   ├──     05  AMT1-IN.
+│   │   ├──     05  AMT2-IN.
+│   │   ├──     05  AMT3-IN.
+│   │   ├── 01  DISPLAYED-OUTPUT.
+│   │   ├──     05  CUST-NO-OUT.
+│   │   ├──     05  TOTAL-OUT.
+│   │   ├── 01  MORE-DATA.
+│   │   ├── PROCEDURE DIVISION.
+│   │   └── 100-MAIN.
+│   ├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+│   │   ├── IDENTIFICATION DIVISION.
+│   │   ├── PROGRAM-ID. CUSTOMER-INVOICE.
+│   │   ├── AUTHOR. JANE DOE.
+│   │   ├── DATE. 2023-12-30.
+│   │   ├──   DATE-COMPILED. 06/30/10.
+│   │   ├──     DATE-WRITTEN. 12/34/56.
+│   │   ├── ENVIRONMENT DIVISION.
+│   │   ├── INPUT-OUTPUT SECTION.
+│   │   ├── FILE-CONTROL.
+│   │   ├──     SELECT CUSTOMER-FILE.
+│   │   ├──     SELECT INVOICE-FILE.
+│   │   ├──     SELECT REPORT-FILE.
+│   │   ├── DATA DIVISION.
+│   │   ├── FILE SECTION.
+│   │   ├── FD CUSTOMER-FILE.
+│   │   ├── 01 CUSTOMER-RECORD.
+│   │   ├──    05 CUSTOMER-ID.
+│   │   ├──    05 CUSTOMER-NAME.
+│   │   ├──    05 CUSTOMER-BALANCE.
+│   │   ├── FD INVOICE-FILE.
+│   │   ├── 01 INVOICE-RECORD.
+│   │   ├──    05 INVOICE-ID.
+│   │   ├──    05 CUSTOMER-ID.
+│   │   ├──    05 INVOICE-AMOUNT.
+│   │   ├── FD REPORT-FILE.
+│   │   ├── 01 REPORT-RECORD.
+│   │   ├── WORKING-STORAGE SECTION.
+│   │   ├── 01 WS-CUSTOMER-FOUND.
+│   │   ├── 01 WS-END-OF-FILE.
+│   │   ├── 01 WS-TOTAL-BALANCE.
+│   │   ├── PROCEDURE DIVISION.
+│   │   ├── 0000-MAIN-ROUTINE.
+│   │   ├── 1000-PROCESS-RECORDS.
+│   │   ├── 1100-UPDATE-CUSTOMER-BALANCE.
+│   │   └── END PROGRAM CUSTOMER-INVOICE.
+│   ├── 📄 JavaTest.java (578 tokens, 86 lines)
+│   │   ├── abstract class LivingBeing
+│   │   ├──     abstract void breathe()
+│   │   ├── interface Communicator
+│   │   ├──     String communicate()
+│   │   ├── @Log
+│   │   ├── @Getter
+│   │   ├── @Setter
+│   │   ├── class Person extends LivingBeing implements Communicator
+│   │   ├──     Person(String name, int age)
+│   │   ├──     @Override
+│   │   ├──     void breathe()
+│   │   ├──     @Override
+│   │   ├──     public String communicate()
+│   │   ├──     void greet()
+│   │   ├──     String personalizedGreeting(String greeting, Optional<Boolean>
+│   │   │   includeAge)
+│   │   ├── @Singleton
+│   │   ├── @RestController
+│   │   ├── @SpringBootApplication
+│   │   ├── public class Example
+│   │   ├──     @Inject
+│   │   ├──     public Example(Person person)
+│   │   ├──     @RequestMapping("/greet")
+│   │   ├──     String home(@RequestParam(value = "name", defaultValue =
+│   │   │   "World") String name,
+│   │   │                   @RequestParam(value = "age", defaultValue = "30")
+│   │   │   int age)
+│   │   └──     public static void main(String[] args)
+│   ├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+│   │   ├── module JuliaTest_EdgeCase
+│   │   ├── struct Location
+│   │   │       name::String
+│   │   │       lat::Float32
+│   │   │       lon::Float32
+│   │   │   end
+│   │   ├── mutable struct mPerson
+│   │   │       name::String
+│   │   │       age::Int
+│   │   │   end
+│   │   ├── Base.@kwdef mutable struct Param
+│   │   │       Δt::Float64 = 0.1
+│   │   │       n::Int64
+│   │   │       m::Int64
+│   │   │   end
+│   │   ├──     sic(x,y)
+│   │   ├── welcome(l::Location)
+│   │   ├── ∑(α, Ω)
+│   │   ├── function noob()
+│   │   │   end
+│   │   ├── function ye_olde(hello::String, world::Location)
+│   │   │   end
+│   │   ├── function multiline_greet(
+│   │   │           p::mPerson,
+│   │   │           greeting::String
+│   │   │       )
+│   │   │   end
+│   │   ├── function julia_is_awesome(prob::DiffEqBase.AbstractDAEProblem{uType,
+│   │   │   duType, tType,
+│   │   │           isinplace};
+│   │   │       kwargs...) where {uType, duType, tType, isinplace}
+│   │   │   end
+│   │   └── end
+│   ├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+│   │   ├── data class Person(val name: String)
+│   │   ├── fun greet(person: Person)
+│   │   ├── fun <T> processItems(items: List<T>, processor: (T) -> Unit)
+│   │   ├── interface Source<out T>
+│   │   ├──     fun nextT(): T
+│   │   ├── fun MutableList<Int>.swap(index1: Int, index2: Int)
+│   │   ├── fun Any?.toString(): String
+│   │   ├── tailrec fun findFixPoint(x: Double = 1.0): Double
+│   │   ├── class GenericRepository<T>
+│   │   ├──     fun getItem(id: Int): T?
+│   │   ├── sealed interface Error
+│   │   ├── sealed class IOError(): Error
+│   │   ├── object Runner
+│   │   ├──     inline fun <reified S: SomeClass<T>, T> run() : T
+│   │   ├── infix fun Int.shl(x: Int): Int
+│   │   ├── class MyStringCollection
+│   │   ├──     infix fun add(s: String)
+│   │   ├──     fun build()
+│   │   ├── open class Base(p: Int)
+│   │   ├── class Derived(p: Int) : Base(p)
+│   │   ├── open class Shape
+│   │   ├──     open fun draw()
+│   │   ├──     fun fill()
+│   │   ├──     open fun edge(case: Int)
+│   │   ├── interface Thingy
+│   │   ├──     fun edge()
+│   │   ├── class Circle() : Shape(), Thingy
+│   │   ├──     override fun draw()
+│   │   ├──     final override fun edge(case: Int)
+│   │   ├── interface Base
+│   │   ├──     fun print()
+│   │   ├── class BaseImpl(val x: Int) : Base
+│   │   ├──     override fun print()
+│   │   ├── internal class Derived(b: Base) : Base by b
+│   │   ├── class Person constructor(firstName: String)
+│   │   ├── class People(
+│   │   │       firstNames: Array<String>,
+│   │   │       ages: Array<Int>(42),
+│   │   │   )
+│   │   ├──     fun edgeCases(): Boolean
+│   │   ├── class Alien public @Inject constructor(
+│   │   │       val firstName: String,
+│   │   │       val lastName: String,
+│   │   │       var age: Int,
+│   │   │       val pets: MutableList<Pet> = mutableListOf(),
+│   │   │   )
+│   │   ├──     fun objectOriented(): String
+│   │   ├──  enum class IntArithmetics : BinaryOperator<Int>, IntBinaryOperator
+│   │   ├──     PLUS {
+│   │   │           override fun apply(t: Int, u: Int): Int
+│   │   ├──     TIMES {
+│   │   │           override fun apply(t: Int, u: Int): Int
+│   │   ├──     override fun applyAsInt(t: Int, u: Int)
+│   │   ├── fun reformat(
+│   │   │       str: String,
+│   │   │       normalizeCase: Boolean = true,
+│   │   │       upperCaseFirstLetter: Boolean = true,
+│   │   │       divideByCamelHumps: Boolean = false,
+│   │   │       wordSeparator: Char = ' ',
+│   │   │   )
+│   │   ├── operator fun Point.unaryMinus()
+│   │   ├── abstract class Polygon
+│   │   └──     abstract fun draw()
+│   ├── 📄 lesson.cbl (635 tokens, 78 lines)
+│   │   ├── IDENTIFICATION DIVISION.
+│   │   ├── PROGRAM-ID.    CBL0002.
+│   │   ├── AUTHOR.        Otto B. Fun.
+│   │   ├── ENVIRONMENT DIVISION.
+│   │   ├── INPUT-OUTPUT SECTION.
+│   │   ├── FILE-CONTROL.
+│   │   ├──     SELECT PRINT-LINE.
+│   │   ├──     SELECT ACCT-REC.
+│   │   ├── DATA DIVISION.
+│   │   ├── FILE SECTION.
+│   │   ├── FD  PRINT-LINE.
+│   │   ├── 01  PRINT-REC.
+│   │   ├──     05  ACCT-NO-O.
+│   │   ├──     05  ACCT-LIMIT-O.
+│   │   ├──     05  ACCT-BALANCE-O.
+│   │   ├──     05  LAST-NAME-O.
+│   │   ├──     05  FIRST-NAME-O.
+│   │   ├──     05  COMMENTS-O.
+│   │   ├── FD  ACCT-REC.
+│   │   ├── 01  ACCT-FIELDS.
+│   │   ├──     05  ACCT-NO.
+│   │   ├──     05  ACCT-LIMIT.
+│   │   ├──     05  ACCT-BALANCE.
+│   │   ├──     05  LAST-NAME.
+│   │   ├──     05  FIRST-NAME.
+│   │   ├──     05  CLIENT-ADDR.
+│   │   ├──         10  STREET-ADDR.
+│   │   ├──         10  CITY-COUNTY.
+│   │   ├──         10  USA-STATE.
+│   │   ├──     05  RESERVED.
+│   │   ├──     05  COMMENTS.
+│   │   ├── WORKING-STORAGE SECTION.
+│   │   ├── 01 FLAGS.
+│   │   ├──   05 LASTREC.
+│   │   ├── PROCEDURE DIVISION.
+│   │   ├── OPEN-FILES.
+│   │   ├── READ-NEXT-RECORD.
+│   │   ├── CLOSE-STOP.
+│   │   ├── READ-RECORD.
+│   │   └── WRITE-RECORD.
+│   ├── 📄 LuaTest.lua (83 tokens, 16 lines)
+│   │   ├── function HelloWorld.new
+│   │   ├── function HelloWorld.greet
+│   │   └── function say_hello
+│   ├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+│   │   ├── @interface HelloWorld
+│   │   ├── @interface HelloWorld -> (void) sayHello
+│   │   ├── @implementation HelloWorld
+│   │   ├── @implementation HelloWorld -> (void) sayHello
+│   │   └── void sayHelloWorld()
+│   ├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+│   │   ├── type color
+│   │   ├── class hello
+│   │   ├── class hello -> method say_hello
+│   │   └── let main ()
+│   ├── 📄 test.js (757 tokens, 154 lines)
+│   │   ├── class MyClass
+│   │   ├──   myMethod()
+│   │   ├──   async asyncMethod(a, b)
+│   │   ├──   methodWithDefaultParameters(a = 5, b = 10)
+│   │   ├──   multilineMethod(
+│   │   │       c,
+│   │   │       d
+│   │   │     )
+│   │   ├──   multilineMethodWithDefaults(
+│   │   │       t = "tree",
+│   │   │       p = "plus"
+│   │   │     )
+│   │   ├── function myFunction(param1, param2)
+│   │   ├── function multilineFunction(
+│   │   │     param1,
+│   │   │     param2
+│   │   │   )
+│   │   ├── const arrowFunction = () =>
+│   │   ├── const parametricArrow = (a, b) =>
+│   │   ├── function ()
+│   │   ├── function outerFunction(outerParam)
+│   │   ├──   function innerFunction(innerParam)
+│   │   ├──   innerFunction("inner")
+│   │   ├── const myObject = {
+│   │   ├──   myMethod: function (stuff)
+│   │   ├── let myArrowObject = {
+│   │   ├──   myArrow: ({
+│   │   │       a,
+│   │   │       b,
+│   │   │       c,
+│   │   │     }) =>
+│   │   ├── const myAsyncArrowFunction = async () =>
+│   │   ├── function functionWithRestParameters(...args)
+│   │   ├── const namedFunctionExpression = function myNamedFunction()
+│   │   ├── const multilineArrowFunction = (
+│   │   │     a,
+│   │   │     b
+│   │   │   ) =>
+│   │   ├── function functionReturningFunction()
+│   │   ├──   return function ()
+│   │   ├── function destructuringOnMultipleLines({
+│   │   │     a,
+│   │   │     b,
+│   │   │   })
+│   │   ├── const arrowFunctionWithDestructuring = ({ a, b }) =>
+│   │   ├── const multilineDestructuringArrow = ({
+│   │   │     a,
+│   │   │     b,
+│   │   │   }) =>
+│   │   ├── async function asyncFunctionWithErrorHandling()
+│   │   ├── class Car
+│   │   ├──   constructor(brand)
+│   │   ├──   present()
+│   │   ├── class Model extends Car
+│   │   ├──   constructor(brand, mod)
+│   │   ├──     super(brand)
+│   │   └──   show()
+│   └── 📄 test.ts (832 tokens, 165 lines)
+│       ├── type MyType
+│       ├── interface MyInterface
+│       ├── class TsClass
+│       ├──   myMethod()
+│       ├──   myMethodWithArgs(param1: string, param2: number): void
+│       ├──   static myStaticMethod<T>(param: T): T
+│       ├──   multilineMethod(
+│       │       c: number,
+│       │       d: number
+│       │     ): number
+│       ├──   multilineMethodWithDefaults(
+│       │       t: string = "tree",
+│       │       p: string = "plus"
+│       │     ): string
+│       ├── export class AdvancedComponent implements MyInterface
+│       ├──   async myAsyncMethod(
+│       │       a: string,
+│       │       b: number,
+│       │       c: string
+│       │     ): Promise<void>
+│       ├──   genericMethod<T, U>(
+│       │       arg1: T,
+│       │       arg2: U
+│       │     ): [T, U]
+│       ├── export class TicketsComponent implements MyInterface
+│       ├──   async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String
+│       │   })
+│       ├── function tsFunction()
+│       ├── function tsFunctionSigned(
+│       │     param1: number,
+│       │     param2: number
+│       │   ): void
+│       ├── export default async function tsFunctionComplicated<A, B, C>({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string>
+│       ├──   return("Standalone function with parameters")
+│       ├── const tsArrowFunctionSigned = ({
+│       │     a,
+│       │     b,
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │   }) =>
+│       ├── export const tsComplicatedArrow = async ({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string> =>
+│       ├── const arrowFunction = () =>
+│       ├── const arrow = (a: String, b: Number) =>
+│       ├── const asyncArrowFunction = async () =>
+│       ├── const asyncArrow = async (a: String, b: Number) =>
+│       ├── let weirdArrow = () =>
+│       ├── const asyncPromiseArrow = async (): Promise<void> =>
+│       ├── let myWeirdArrowSigned = (x: number): number =>
+│       ├── class Person
+│       ├──   constructor(private firstName: string, private lastName: string)
+│       ├──   getFullName(): string
+│       ├──   describe(): string
+│       ├── class Employee extends Person
+│       ├──   constructor(
+│       │       firstName: string,
+│       │       lastName: string,
+│       │       private jobTitle: string
+│       │     )
+│       ├──     super(firstName, lastName)
+│       ├──   describe(): string
+│       ├── interface Shape
+│       └── interface Square extends Shape
+├── 📁 group2 (1 folder, 8 files)
+│   ├── 📄 apl_test.apl (28 tokens, 5 lines)
+│   │   ├── :Namespace HelloWorld
+│   │   ├── :Namespace HelloWorld -> hello ← 'Hello, World!'
+│   │   └── :Namespace HelloWorld -> plus ← {⍺+⍵}
+│   ├── 📄 c_test.c (837 tokens, 142 lines)
+│   │   ├── struct Point
+│   │   ├──     int x;
+│   │   ├──     int y;
+│   │   ├── struct Point getOrigin()
+│   │   ├── float mul_two_floats(float x1, float x2)
+│   │   ├── enum days
+│   │   ├──     SUN,
+│   │   ├──     MON,
+│   │   ├──     TUE,
+│   │   ├──     WED,
+│   │   ├──     THU,
+│   │   ├──     FRI,
+│   │   ├──     SAT
+│   │   ├── long add_two_longs(long x1, long x2)
+│   │   ├── double multiplyByTwo(double num)
+│   │   ├── char getFirstCharacter(char *str)
+│   │   ├── void greet(Person p)
+│   │   ├── typedef struct
+│   │   ├──     char name[50];
+│   │   ├── } Person;
+│   │   ├── int main()
+│   │   ├── int* getArrayStart(int arr[], int size)
+│   │   ├── long complexFunctionWithMultipleArguments(
+│   │   │       int param1,
+│   │   │       double param2,
+│   │   │       char *param3,
+│   │   │       struct Point point
+│   │   │   )
+│   │   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   │   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   │   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   │   *channel, int channellen, int is_pattern)
+│   │   ├──     while((ln = listNext(&li)))
+│   │   ├── static struct config
+│   │   ├──     aeEventLoop *el;
+│   │   ├──     cliConnInfo conn_info;
+│   │   ├──     const char *hostsocket;
+│   │   ├──     int tls;
+│   │   ├──     struct cliSSLconfig sslconfig;
+│   │   └── } config;
+│   ├── 📄 go_test.go (179 tokens, 46 lines)
+│   │   ├── type Greeting struct
+│   │   ├── func (g Greeting) sayHello()
+│   │   ├── func createGreeting(m string) Greeting
+│   │   ├── type SomethingLong struct
+│   │   ├── func (s *SomethingLong) WithAReasonableName(
+│   │   │           ctx context.Context,
+│   │   │           param1 string,
+│   │   │           param2 int,
+│   │   │           param3 map[string]interface{},
+│   │   │           callback func(int) error,
+│   │   │   ) (resultType, error)
+│   │   ├── type resultType struct
+│   │   └── func main()
+│   ├── 📄 PerlTest.pl (63 tokens, 20 lines)
+│   │   ├── package PerlTest
+│   │   ├── package PerlTest -> sub new
+│   │   ├── package PerlTest -> sub hello
+│   │   └── package PerlTest -> sub say_hello
+│   ├── 📄 PhpTest.php (70 tokens, 19 lines)
+│   │   ├── class HelloWorld
+│   │   ├── class HelloWorld -> function sayHello
+│   │   ├── function greet
+│   │   ├── class Person
+│   │   └── class Person -> function __construct
+│   ├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)
+│   │   ├── function Say-Nothing()
+│   │   ├── class Person
+│   │   ├──     Person([string]$name)
+│   │   ├──     [string]Greet()
+│   │   ├──     [string]GreetMany([int]$times)
+│   │   ├──     [string]GreetWithDetails([string]$greeting, [int]$times)
+│   │   ├──     [string]GreetMultiline(
+│   │   │           [string]$greeting,
+│   │   │           [int]$times
+│   │   │       )
+│   │   ├──     NoReturn([int]$times)
+│   │   ├──     NoReturnNoArgs()
+│   │   ├── function Say-Hello([Person]$person)
+│   │   ├── function Multi-Hello([Person]$personA, [Person]$personB)
+│   │   ├── function Switch-Item
+│   │   ├──   param ([switch]$on)
+│   │   ├── function Get-SmallFiles
+│   │   ├──   param (
+│   │   │         [PSDefaultValue(Help = '100')]
+│   │   │         $Size = 100)
+│   │   ├── function Get-User
+│   │   ├──   [CmdletBinding(DefaultParameterSetName="ID")]
+│   │   ├──   [OutputType("System.Int32", ParameterSetName="ID")]
+│   │   ├──   [OutputType([String], ParameterSetName="Name")]
+│   │   ├──   Param (
+│   │   │       [parameter(Mandatory=$true, ParameterSetName="ID")]
+│   │   │       [Int[]]
+│   │   │       $UserID,
+│   │   │       [parameter(Mandatory=$true, ParameterSetName="Name")]
+│   │   │       [String[]]
+│   │   │       $UserName)
+│   │   ├── filter Get-ErrorLog ([switch]$Message)
+│   │   └── function global:MultilineSignature(
+│   │         [string]$param1,
+│   │         [int]$param2,
+│   │         [Parameter(Mandatory=$true)]
+│   │         [string]$param3
+│   │       )
+│   ├── 📄 ScalaTest.scala (171 tokens, 40 lines)
+│   │   ├── def sumOfSquares(x: Int, y: Int): Int
+│   │   ├── trait Bark
+│   │   ├──   def bark: String
+│   │   ├── case class Person(name: String)
+│   │   ├── class GenericClass[T](
+│   │   │       val data: T,
+│   │   │       val count: Int
+│   │   │   )
+│   │   ├──   def getData: T
+│   │   ├── object HelloWorld
+│   │   ├──   def greet(person: Person): Unit
+│   │   ├──   def main(args: Array[String]): Unit
+│   │   ├── def complexFunction(
+│   │   │       a: Int,
+│   │   │       b: String,
+│   │   │       c: Float
+│   │   │   ): (Int, String) Option
+│   │   └── def sumOfSquaresShort(x: Int, y: Int): Int
+│   └── 📄 test.csv (0 tokens, 0 lines)
+│       ├── Name
+│       ├── Age
+│       ├── Country
+│       ├── City
+│       └── Email
+├── 📁 group3 (1 folder, 16 files)
+│   ├── 📄 bash_test.sh (127 tokens, 22 lines)
+│   │   ├── echo_hello_world()
+│   │   ├── function fun_echo_hello_world()
+│   │   ├── export SECRET
+│   │   ├── alias md='make debug'
+│   │   ├── add_alias()
+│   │   └── create_conda_env()
+│   ├── 📄 cpp_test.cpp (1,670 tokens, 259 lines)
+│   │   ├── class Person
+│   │   ├──     std::string name;
+│   │   ├── public:
+│   │   ├──     Person(std::string n) : name(n)
+│   │   ├──     void greet()
+│   │   ├── void globalGreet()
+│   │   ├── int main()
+│   │   ├── void printMessage(const std::string &message)
+│   │   ├── template<typename T>
+│   │   │   void printVector(const std::vector<T>& vec)
+│   │   ├── struct Point
+│   │   ├──     int x, y;
+│   │   ├──     Point(int x, int y) : x(x), y(y)
+│   │   ├── class Animal
+│   │   ├── public:
+│   │   ├──     Animal(const std::string &name) : name(name)
+│   │   ├──     virtual void speak() const
+│   │   ├──     virtual ~Animal()
+│   │   ├── protected:
+│   │   ├──     std::string name;
+│   │   ├── class Dog : public Animal
+│   │   ├── public:
+│   │   ├──     Dog(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── class Cat : public Animal
+│   │   ├── public:
+│   │   ├──     Cat(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   │   num_layers,
+│   │   │                                int batch_size, int max_seq_length,
+│   │   │   float dropout,
+│   │   │                                bool bidirectional, bool
+│   │   │   cudnn_allow_tf32,
+│   │   │                                int workspace_size, int
+│   │   │   reserve_space_size)
+│   │   ├── int main()
+│   │   ├── enum ECarTypes
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV,
+│   │   ├──   Wagon
+│   │   ├── ECarTypes GetPreferredCarType()
+│   │   ├── enum ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── enum class ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── void myFunction(string fname, int age)
+│   │   ├── template <typename T> T cos(T)
+│   │   ├── template <typename T> T sin(T)
+│   │   ├── template <typename T> T sqrt(T)
+│   │   ├── template<typename T> struct VLEN
+│   │   ├── template<typename T> class arr
+│   │   ├──   private:
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──   public:
+│   │   ├──     arr() : p(0), sz(0)
+│   │   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   │   ├──     arr(arr &&other)
+│   │   │         : p(other.p), sz(other.sz)
+│   │   ├──     ~arr()
+│   │   ├──     void resize(size_t n)
+│   │   ├──     T &operator[](size_t idx)
+│   │   ├──     T *data()
+│   │   ├──     size_t size() const
+│   │   ├── class Buffer
+│   │   ├──  private:
+│   │   ├──   void* ptr_;
+│   │   └── std::tuple<array, array, array> quantize(
+│   │           const array& w,
+│   │           int group_size,
+│   │           int bits,
+│   │           StreamOrDevice s)
+│   ├── 📄 csharp_test.cs (957 tokens, 146 lines)
+│   │   ├── public interface IExcelTemplate
+│   │   ├──     void LoadTemplate(string templateFilePath)
+│   │   ├──     void LoadData(Dictionary<string, string> data)
+│   │   ├──     void ModifyCell(string cellName, string value)
+│   │   ├──     void SaveToFile(string filePath)
+│   │   ├── public interface IGreet
+│   │   ├──     void Greet()
+│   │   ├── public enum WeekDays
+│   │   ├── public delegate void DisplayMessage(string message)
+│   │   ├── public struct Address
+│   │   ├── public static class HelperFunctions
+│   │   ├──     public static void PrintMessage(string message)
+│   │   ├──     public static int AddNumbers(int a, int b)
+│   │   ├── namespace HelloWorldApp
+│   │   ├──     class Person : IGreet
+│   │   ├──         public Person(string name, int age)
+│   │   ├──         public void Greet()
+│   │   ├──     class HelloWorld
+│   │   ├──         static void Main(string[] args)
+│   │   ├── namespace TemplateToExcelServer.Template
+│   │   ├──     public interface ITemplateObject
+│   │   ├──         string[,] GetContent()
+│   │   ├──         string[] GetContentArray()
+│   │   ├──         string[] GetFormat()
+│   │   ├──         int? GetFormatLength()
+│   │   ├──         TemplateObject SetContent(string[,] Content)
+│   │   ├──         TemplateObject SetContentArray(string[] value)
+│   │   ├──         TemplateObject SetFormat(string[] Header)
+│   │   ├──         TemplateObject SetNameOfReport(
+│   │   │               ReadOnlyMemory<byte> ReportName,
+│   │   │               int[] EdgeCase)
+│   │   ├──         TemplateObject SetSheetName(ReadOnlyMemory<byte> SheetName)
+│   │   ├── public class BankAccount(string accountID, string owner)
+│   │   ├──     public override string ToString() =>
+│   │   ├── var IncrementBy = (int source, int increment = 1) =>
+│   │   ├── Func<int, int, int> add = (x, y) =>
+│   │   ├── button.Click += (sender, args) =>
+│   │   ├── public Func<int, int> GetMultiplier(int factor)
+│   │   ├── public void Method(
+│   │   │           int param1,
+│   │   │           int param2,
+│   │   │           int param3,
+│   │   │           int param4,
+│   │   │           int param5,
+│   │   │           int param6,
+│   │   │       )
+│   │   ├── System.Net.ServicePointManager.ServerCertificateValidationCallback
+│   │   │   +=
+│   │   │       (se, cert, chain, sslerror) =>
+│   │   ├── class ServerCertificateValidation
+│   │   ├──     public bool OnRemoteCertificateValidation(
+│   │   │           object se,
+│   │   │           X509Certificate cert,
+│   │   │           X509Chain chain,
+│   │   │           SslPolicyErrors sslerror
+│   │   │       )
+│   │   ├── s_downloadButton.Clicked += async (o, e) =>
+│   │   ├── [HttpGet, Route("DotNetCount")]
+│   │   └── static public async Task<int> GetDotNetCount(string URL)
+│   ├── 📄 hallucination.tex (1,633 tokens, 126 lines)
+│   │   ├── Harnessing the Master Algorithm: Strategies for AI LLMs to Mitigate
+│   │   │   Hallucinations
+│   │   ├── Hallucinated Pedro Domingos et al.
+│   │   ├── Christmas Eve 2023
+│   │   ├── 1 Introduction
+│   │   ├── 2 Representation in LLMs
+│   │   ├──   2.1 Current Representational Models
+│   │   ├──   2.2 Incorporating Cognitive Structures
+│   │   ├──   2.3 Conceptual Diagrams of Advanced Representational Models
+│   │   ├── 3 Evaluation Strategies
+│   │   ├──   3.1 Existing Evaluation Metrics for LLMs
+│   │   ├──   3.2 Integrating Contextual and Ethical Considerations
+│   │   ├──   3.3 Case Studies: Evaluation in Practice
+│   │   ├── 4 Optimization Techniques
+│   │   ├──   4.1 Continuous Learning Models
+│   │   ├──   4.2 Adaptive Algorithms for Real-time Adjustments
+│   │   ├──   4.3 Performance Metrics Pre- and Post-Optimization
+│   │   ├── 5 Interdisciplinary Insights
+│   │   ├──   5.1 Cognitive Science and AI: A Symbiotic Relationship
+│   │   ├──   5.2 Learning from Human Cognitive Processes
+│   │   ├── 6 Challenges and Future Directions
+│   │   ├──   6.1 Addressing Current Limitations
+│   │   ├──   6.2 The Road Ahead: Ethical and Practical Considerations
+│   │   ├── 7 Conclusion
+│   │   ├──   7.1 Summarizing Key Findings
+│   │   └──   7.2 The Next Steps in AI Development
+│   ├── 📄 ruby_test.rb (138 tokens, 37 lines)
+│   │   ├── module Greeter
+│   │   ├──   def self.say_hello
+│   │   ├── class HelloWorld
+│   │   ├──   def say_hello
+│   │   ├── class Human
+│   │   ├──   def self.bar
+│   │   ├──   def self.bar=(value)
+│   │   ├── class Doctor < Human
+│   │   └──   def brachial_plexus(
+│   │             roots,
+│   │             trunks,
+│   │             divisions: true,
+│   │             cords: [],
+│   │             branches: Time.now
+│   │           )
+│   ├── 📄 swift_test.swift (469 tokens, 110 lines)
+│   │   ├── class Person
+│   │   ├──     init(name: String)
+│   │   ├──     func greet()
+│   │   ├──     func yEdgeCase(
+│   │   │           fname: String,
+│   │   │           lname: String,
+│   │   │           age: Int,
+│   │   │           address: String,
+│   │   │           phoneNumber: String
+│   │   │       )
+│   │   ├── func globalGreet()
+│   │   ├── struct Point
+│   │   ├── protocol Animal
+│   │   ├──     func speak()
+│   │   ├── struct Dog: Animal
+│   │   ├── class Cat: Animal
+│   │   ├──     init(name: String)
+│   │   ├──     func speak()
+│   │   ├── enum CarType
+│   │   ├── func getPreferredCarType() -> CarType
+│   │   ├── enum CarType: UInt8
+│   │   ├── enum class CarType: UInt8
+│   │   ├── func myFunction(fname: String, age: Int)
+│   │   └── func myFunctionWithMultipleParameters(
+│   │           fname: String,
+│   │           lname: String,
+│   │           age: Int,
+│   │           address: String,
+│   │           phoneNumber: String
+│   │       )
+│   ├── 📄 test.lean (289 tokens, 42 lines)
+│   │   ├── # Advanced Topics in Group Theory
+│   │   ├── section GroupDynamics
+│   │   ├── lemma group_stability (G : Type*) [Group G] (H : Subgroup G)
+│   │   ├── theorem subgroup_closure {G : Type*} [Group G] (S : Set G)
+│   │   ├── axiom group_homomorphism_preservation {G H : Type*} [Group G] [Group
+│   │   │   H] (f : G → H)
+│   │   ├── end GroupDynamics
+│   │   ├── section ConstructiveApproach
+│   │   ├── lemma finite_group_order (G : Type*) [Group G] [Fintype G]
+│   │   ├── lemma complex_lemma {X Y : Type*} [SomeClass X] [AnotherClass Y]
+│   │   │     (f : X → Y) (g : Y → X)
+│   │   └── end ConstructiveApproach
+│   ├── 📄 test.capnp (117 tokens, 30 lines)
+│   │   ├── struct Employee
+│   │   ├──   id @0 :Int32
+│   │   ├──   name @1 :Text
+│   │   ├──   role @2 :Text
+│   │   ├──   skills @3 :List(Skill)
+│   │   ├──   struct Skill
+│   │   ├──     name @0 :Text
+│   │   ├──     level @1 :Level
+│   │   ├──     enum Level
+│   │   ├──       beginner @0
+│   │   ├──       intermediate @1
+│   │   ├──       expert @2
+│   │   ├──   status :union
+│   │   ├──     active @4 :Void
+│   │   ├──     onLeave @5 :Void
+│   │   ├──     retired @6 :Void
+│   │   ├── struct Company
+│   │   └──   employees @0 :List(Employee)
+│   ├── 📄 test.graphql (66 tokens, 21 lines)
+│   │   ├── type Query
+│   │   ├──     getBooks: [Book]
+│   │   ├──     getAuthors: [Author]
+│   │   ├── type Mutation
+│   │   ├──     addBook(title: String, author: String): Book
+│   │   ├──     removeBook(id: ID): Book
+│   │   ├── type Book
+│   │   ├──     id: ID
+│   │   ├──     title: String
+│   │   ├──     author: Author
+│   │   ├── type Author
+│   │   ├──     id: ID
+│   │   ├──     name: String
+│   │   └──     books: [Book]
+│   ├── 📄 test.proto (142 tokens, 34 lines)
+│   │   ├── syntax = "proto3"
+│   │   ├── service EmployeeService
+│   │   ├──     rpc GetEmployee(EmployeeId) returns (EmployeeInfo)
+│   │   ├──     rpc AddEmployee(EmployeeData) returns (EmployeeInfo)
+│   │   ├──     rpc UpdateEmployee(EmployeeUpdate) returns (EmployeeInfo)
+│   │   ├── message EmployeeId
+│   │   ├──     int32 id = 1
+│   │   ├── message EmployeeInfo
+│   │   ├──     int32 id = 1
+│   │   ├──     string name = 2
+│   │   ├──     string role = 3
+│   │   ├── message EmployeeData
+│   │   ├──     string name = 1
+│   │   ├──     string role = 2
+│   │   ├── message EmployeeUpdate
+│   │   ├──     int32 id = 1
+│   │   ├──     string name = 2
+│   │   └──     string role = 3
+│   ├── 📄 test.sqlite (0 tokens, 0 lines)
+│   │   ├── students table:
+│   │   ├──    id integer primary key
+│   │   ├──    name text not null
+│   │   ├──    age integer not null
+│   │   ├── courses table:
+│   │   ├──    id integer primary key
+│   │   ├──    title text not null
+│   │   └──    credits integer not null
+│   ├── 📄 test_Cargo.toml (119 tokens, 18 lines)
+│   │   ├── name: test_cargo
+│   │   ├── version: 0.1.0
+│   │   ├── description: A test Cargo.toml
+│   │   ├── license: MIT OR Apache-2.0
+│   │   ├── dependencies:
+│   │   ├──   clap 4.4
+│   │   └──   sqlx 0.7 (features: runtime-tokio, tls-rustls)
+│   ├── 📄 test_json_rpc_2_0.json (26 tokens, 6 lines)
+│   │   ├── jsonrpc: 2.0
+│   │   ├── method: subtract
+│   │   ├── params:
+│   │   ├──     minuend: 42
+│   │   ├──     subtrahend: 23
+│   │   └── id: 1
+│   ├── 📄 test_openapi.yaml (753 tokens, 92 lines)
+│   │   ├── openapi: 3.0.1
+│   │   ├──     title: TODO Plugin
+│   │   ├──     description: A plugin to create and manage TODO lists using
+│   │   │   ChatGPT.
+│   │   ├──     version: v1
+│   │   ├── servers:
+│   │   ├──     - url: PLUGIN_HOSTNAME
+│   │   ├── paths:
+│   │   ├──     '/todos/{username}':
+│   │   ├──         GET (getTodos): Get the list of todos
+│   │   ├──         POST (addTodo): Add a todo to the list
+│   │   └──         DELETE (deleteTodo): Delete a todo from the list
+│   ├── 📄 test_openrpc.json (225 tokens, 44 lines)
+│   │   ├── openrpc: 1.2.1
+│   │   ├── info:
+│   │   ├──     title: Demo Petstore
+│   │   ├──     version: 1.0.0
+│   │   ├── methods:
+│   │   ├──     listPets: List all pets
+│   │   ├──         params:
+│   │   ├──             - limit: integer
+│   │   └──         result: pets = An array of pets
+│   └── 📄 test_pyproject.toml (304 tokens, 39 lines)
+│       ├── name: tree_plus
+│       ├── version: 1.0.8
+│       ├── description: A `tree` util enhanced with tokens, lines, and
+│       │   components.
+│       ├── License :: OSI Approved :: Apache Software License
+│       ├── License :: OSI Approved :: MIT License
+│       ├── dependencies:
+│       ├──     tiktoken
+│       ├──     PyYAML
+│       ├──     click
+│       ├──     rich
+│       └──     tomli
+├── 📁 group4 (1 folder, 10 files)
+│   ├── 📄 erl_test.erl (480 tokens, 68 lines)
+│   │   ├── -module(erl_test).
+│   │   ├── -record(person).
+│   │   ├── -type ra_peer_status().
+│   │   ├── -type ra_membership().
+│   │   ├── -opaque my_opaq_type().
+│   │   ├── -type orddict(Key, Val).
+│   │   ├── -type edge(
+│   │   │           Cases,
+│   │   │           Pwn,
+│   │   │       ).
+│   │   ├── -spec guarded(X) -> X when X :: tuple().
+│   │   ├── -spec edge_case(
+│   │   │           {integer(), any()} | [any()]
+│   │   │       ) -> processed, integer(), any()} | [{item, any()}].
+│   │   ├── -spec complex_function({integer(), any()} | [any()]) ->
+│   │   │       {processed, integer(), any()} | [{item, any()}].
+│   │   ├── -spec list_manipulation([integer()]) -> [integer()].
+│   │   ├── -spec overload(T1, T2) -> T3
+│   │   │           ; (T4, T5) -> T6.
+│   │   ├── -spec multiguard({X, integer()}) -> X when X :: atom()
+│   │   │           ; ([Y]) -> Y when Y :: number().
+│   │   ├── -record(multiline).
+│   │   └── -record(maybe_undefined).
+│   ├── 📄 haskell_test.hs (414 tokens, 41 lines)
+│   │   ├── data Person
+│   │   ├── greet :: Person -> String
+│   │   └── resolveVariables ::
+│   │         forall m fragments.
+│   │         (MonadError QErr m, Traversable fragments) =>
+│   │         Options.BackwardsCompatibleNullInNonNullableVariables ->
+│   │         [G.VariableDefinition] ->
+│   │         GH.VariableValues ->
+│   │         [G.Directive G.Name] ->
+│   │         G.SelectionSet fragments G.Name ->
+│   │         m
+│   │           ( [G.Directive Variable],
+│   │             G.SelectionSet fragments Variable
+│   │           )
+│   ├── 📄 mathematica_test.nb (133 tokens, 21 lines)
+│   │   ├── person[name_]
+│   │   ├── sayHello[]
+│   │   └── sumList[list_List]
+│   ├── 📄 matlab_test.m (48 tokens, 12 lines)
+│   │   ├── classdef HelloWorld -> function greet
+│   │   └── function loneFun
+│   ├── 📄 RTest.R (367 tokens, 46 lines)
+│   │   ├── class(person)
+│   │   ├── greet.Person <- function
+│   │   ├── ensure_between = function
+│   │   └── run_intermediate_annealing_process = function
+│   ├── 📄 rust_test.rs (1,368 tokens, 259 lines)
+│   │   ├── fn at_beginning<'a>(&'a str)
+│   │   ├── pub enum Days<E: EdgeCase> {
+│   │   │       #\[default]
+│   │   │       Sun,
+│   │   │       Mon,
+│   │   │       #\[error("edge case {idx}, expected at least {} and at most {}",
+│   │   │   .limits.lo, .limits.hi)]
+│   │   │       Tue,
+│   │   │       Wed,
+│   │   │       Thu(i16, bool),
+│   │   │       Fri { day: u8 },
+│   │   │       Sat {
+│   │   │           urday: String,
+│   │   │           edge_case: E,
+│   │   │       },
+│   │   │   }
+│   │   ├── struct Point
+│   │   ├── impl Point
+│   │   ├──     fn get_origin() -> Point
+│   │   ├── struct Person
+│   │   ├── impl Person
+│   │   ├──     fn greet(&self)
+│   │   ├── fn add_two_longs(x1: i64, x2: i64) -> i64
+│   │   ├── fn add_two_longs_longer(
+│   │   │       x1: i64,
+│   │   │       x2: i64,
+│   │   │   ) -> i64
+│   │   ├── const fn multiply_by_two(num: f64) -> f64
+│   │   ├── fn get_first_character(s: &str) -> Option<char>
+│   │   ├── trait Drawable
+│   │   ├──     fn draw(&self)
+│   │   ├── impl Drawable for Point
+│   │   ├──     fn draw(&self)
+│   │   ├── fn with_generic<D: Drawable>(d: D)
+│   │   ├── fn with_generic<D>(d: D)
+│   │   │   where
+│   │   │       D: Drawable
+│   │   ├── fn main()
+│   │   ├── pub struct VisibleStruct
+│   │   ├── mod my_module
+│   │   ├──     pub struct AlsoVisibleStruct<T>(T, T)
+│   │   ├── macro_rules! say_hello
+│   │   ├── #[macro_export]
+│   │   │   macro_rules! hello_tree_plus
+│   │   ├── pub mod lib
+│   │   ├──     pub mod interfaces
+│   │   ├──     mod engine
+│   │   ├── pub fn flow<S1, S2, S3, S4, E, T, L>(
+│   │   │       source: S1,
+│   │   │       extractor: E,
+│   │   │       inbox: S2,
+│   │   │       transformer: T,
+│   │   │       outbox: S3,
+│   │   │       loader: L,
+│   │   │       sink: &mut S4,
+│   │   │   ) -> Result<(), Box<dyn Error>>
+│   │   │   where
+│   │   │       S1: Extractable,
+│   │   │       S2: Extractable + Loadable,
+│   │   │       S3: Extractable + Loadable,
+│   │   │       S4: Loadable,
+│   │   │       E: Extractor<S1, S2>,
+│   │   │       T: Transformer<S2, S3>,
+│   │   │       L: Loader<S3, S4>
+│   │   ├── trait Container
+│   │   ├──     fn items(&self) -> impl Iterator<Item = Widget>
+│   │   ├── trait HttpService
+│   │   ├──     async fn fetch(&self, url: Url) -> HtmlBody
+│   │   ├── struct Pair<T, U>
+│   │   ├── trait Transformer<T>
+│   │   ├──     fn transform(&self, input: T) -> T
+│   │   ├── impl<T: std::ops::Add<Output = T> + Copy> Transformer<T> for Pair<T,
+│   │   │   T>
+│   │   ├──     fn transform(&self, input: T) -> T
+│   │   ├── fn main()
+│   │   ├── async fn handle_get(State(pool): State<PgPool>) ->
+│   │   │   Result<Html<String>, (StatusCode, String)>
+│   │   │   where
+│   │   │       Bion: Cool
+│   │   ├── #[macro_export]
+│   │   │   macro_rules! unit
+│   │   ├──             fn insert(
+│   │   │                   &mut self,
+│   │   │                   key: (),
+│   │   │                   value: $unit_dtype,
+│   │   │               ) -> Result<Option<$unit_dtype>, ETLError>
+│   │   ├── pub async fn handle_get_axum_route(
+│   │   │       Session { maybe_claims }: Session,
+│   │   │       Path(RouteParams {
+│   │   │           alpha,
+│   │   │           bravo,
+│   │   │           charlie,
+│   │   │           edge_case
+│   │   │       }): Path<RouteParams>,
+│   │   │   ) -> ServerResult<Response>
+│   │   ├── fn encode_pipeline(cmds: &[Cmd], atomic: bool) -> Vec<u8>
+│   │   ├── pub async fn handle_post_yeet(
+│   │   │       State(auth_backend): State<AuthBackend>,
+│   │   │       Session { maybe_claims }: Session,
+│   │   │       Form(yeet_form): Form<YeetForm>,
+│   │   │   ) -> Result<Response, AuthError>
+│   │   └── pub async fn handle_get_thingy(
+│   │           session: Session,
+│   │           State(ApiBackend {
+│   │               page_cache,
+│   │               auth_backend,
+│   │               library_sql,
+│   │               some_data_cache,
+│   │               metadata_cache,
+│   │               thingy_client,
+│   │               ..
+│   │           }): State<ApiBackend>,
+│   │       ) -> ServerResult<Response>
+│   ├── 📄 test.zig (397 tokens, 60 lines)
+│   │   ├── pub fn add(a: i32, b: i32) i32
+│   │   ├── test "add function"
+│   │   ├── const BunBuildOptions = struct
+│   │   ├──     pub fn updateRuntime(this: *BunBuildOptions) anyerror!void
+│   │   ├──     pub fn step(this: BunBuildOptions, b: anytype)
+│   │   │   *std.build.OptionsStep
+│   │   └── pub fn sgemv(
+│   │           order: Order,
+│   │           trans: Trans,
+│   │           m: usize,
+│   │           n: usize,
+│   │           alpha: f32,
+│   │           a: []const f32,
+│   │           lda: usize,
+│   │           x: []const f32,
+│   │           x_add: usize,
+│   │           beta: f32,
+│   │           y: []f32,
+│   │           y_add: usize,
+│   │       ) void
+│   ├── 📄 test_fsharp.fs (92 tokens, 27 lines)
+│   │   ├── module TestFSharp
+│   │   ├── type Person = {
+│   │   ├── let add x y =
+│   │   ├── let multiply
+│   │   │       (x: int)
+│   │   │       (y: int): int =
+│   │   ├── let complexFunction
+│   │   │       (a: int)
+│   │   │       (b: string)
+│   │   │       (c: float)
+│   │   │       : (int * string) option =
+│   │   └── type Result<'T> =
+│   ├── 📄 test_tcl_tk.tcl (54 tokens, 16 lines)
+│   │   ├── proc sayHello {}
+│   │   ├── proc arrg { input }
+│   │   └── proc multiLine {
+│   │           x,
+│   │           y
+│   │       }
+│   └── 📄 tf_test.tf (202 tokens, 38 lines)
+│       ├── provider "aws"
+│       ├── resource "aws_instance" "example"
+│       ├── data "aws_ami" "ubuntu"
+│       ├── variable "instance_type"
+│       ├── output "instance_public_ip"
+│       ├── locals
+│       └── module "vpc"
+├── 📁 group5 (1 folder, 19 files)
+│   ├── 📄 ansible_test.yml (55 tokens, 14 lines)
+│   │   ├── Install package
+│   │   ├── Start service
+│   │   └── Create user
+│   ├── 📄 app-routing.module.ts (287 tokens, 28 lines)
+│   │   ├── const routes: Routes = [
+│   │   │       { path: '', redirectTo: 'login', pathMatch: 'full' },
+│   │   │       { path: '*', redirectTo: 'login' },
+│   │   │       { path: 'home', component: HomeComponent },
+│   │   │       { path: 'login', component: LoginComponent },
+│   │   │       { path: 'register', component: RegisterComponent },
+│   │   │       { path: 'events', component: EventsComponent },
+│   │   │       { path: 'invites', component: InvitesComponent },
+│   │   │       { path: 'rewards', component: RewardsComponent },
+│   │   │       { path: 'profile', component: ProfileComponent },
+│   │   │   ];
+│   │   └── export class AppRoutingModule
+│   ├── 📄 app.component.spec.ts (410 tokens, 47 lines)
+│   │   ├── describe 'AppComponent'
+│   │   ├──     it should create the app
+│   │   ├──     it should welcome the user
+│   │   ├──     it should welcome 'Jimbo'
+│   │   └──     it should request login if not logged in
+│   ├── 📄 app.component.ts (271 tokens, 45 lines)
+│   │   ├── export class AppComponent
+│   │   ├──   constructor(
+│   │   │       private http: HttpClient,
+│   │   │       private loginService: LoginService,
+│   │   │       private stripeService: StripeService
+│   │   │     )
+│   │   ├──   constructor(private loginService: LoginService)
+│   │   ├──   checkSession()
+│   │   ├──   async goToEvent(event_id: string)
+│   │   └──   valInvitedBy(event: any, event_id: string)
+│   ├── 📄 app.module.ts (374 tokens, 43 lines)
+│   │   ├── @NgModule({
+│   │   │       declarations: [
+│   │   │           AppComponent,
+│   │   │           HomeComponent,
+│   │   │           LoginComponent,
+│   │   │           RegisterComponent,
+│   │   │           EventsComponent,
+│   │   │           InvitesComponent,
+│   │   │           RewardsComponent,
+│   │   │           ProfileComponent
+│   │   └── export class AppModule
+│   ├── 📄 checkbox_test.md (191 tokens, 29 lines)
+│   │   ├── # My Checkbox Test
+│   │   ├── ## My No Parens Test
+│   │   ├── ## My Empty href Test
+│   │   ├── ## My other url Test [Q&A]
+│   │   ├── ## My other other url Test [Q&A]
+│   │   ├── ## My 2nd other url Test [Q&A]
+│   │   ├── ## My 3rd other url Test [Q&A]
+│   │   ├── - [ ] Task 1
+│   │   ├──     - [ ] No Space Task 1.1
+│   │   ├──     - [ ] Two Spaces Task 1.2
+│   │   ├──         - [ ] Subtask 1.2.1
+│   │   ├── - [ ] Task 2
+│   │   ├── - [x] Task 3
+│   │   ├──     - [ ] Subtask 3.1
+│   │   ├── - [x] Task 6
+│   │   ├──     - [x] Subtask 6.1
+│   │   ├──         - [ ] Handle edge cases
+│   │   └── # My Codeblock Test
+│   ├── 📄 checkbox_test.txt (257 tokens, 33 lines)
+│   │   ├── - [ ] fix phone number format +1
+│   │   ├── - [ ] add forgot password
+│   │   ├── - [ ] ? add email verification
+│   │   ├── - [ ] store token the right way
+│   │   ├── - [ ] test nesting of checkboxes
+│   │   ├── - [ ] user can use option to buy ticket at 2-referred price
+│   │   ├── - [ ] CTA refer 2 people to get instant lower price
+│   │   └── - [ ] form to send referrals
+│   ├── 📄 environment.test.ts (197 tokens, 19 lines)
+│   │   ├── environment:
+│   │   ├──    production
+│   │   ├──    cognitoUserPoolId
+│   │   ├──    cognitoAppClientId
+│   │   └──    apiurl
+│   ├── 📄 hello_world.pyi (22 tokens, 3 lines)
+│   │   ├── @final
+│   │   │   class dtype(Generic[_DTypeScalar_co])
+│   │   └──     names: None | tuple[builtins.str, ...]
+│   ├── 📄 k8s_test.yaml (140 tokens, 37 lines)
+│   │   ├── apps/v1.Deployment -> my-app
+│   │   ├── v1.Service -> my-service
+│   │   └── v1.ConfigMap -> my-config
+│   ├── 📄 Makefile (714 tokens, 84 lines)
+│   │   ├── include dotenv/dev.env
+│   │   ├── .PHONY: dev
+│   │   ├── dev
+│   │   ├── services-down
+│   │   ├── services-stop: services-down
+│   │   ├── define CHECK_POSTGRES
+│   │   ├── damage-report
+│   │   ├── tail-logs
+│   │   └── cloud
+│   ├── 📄 requirements_test.txt (29 tokens, 10 lines)
+│   │   ├── psycopg2-binary
+│   │   ├── pytest
+│   │   ├── coverage
+│   │   ├── flask[async]
+│   │   ├── flask_cors
+│   │   ├── stripe
+│   │   ├── pyjwt[crypto]
+│   │   ├── cognitojwt[async]
+│   │   └── flask-lambda
+│   ├── 📄 rust_todo_test.rs (92 tokens, 26 lines)
+│   │   ├── TODO: This todo tests parse_todo
+│   │   ├── enum Color {
+│   │   │       Red,
+│   │   │       Blue,
+│   │   │       Green,
+│   │   │   }
+│   │   ├── struct Point
+│   │   ├── trait Drawable
+│   │   ├──     fn draw(&self)
+│   │   ├── impl Drawable for Point
+│   │   ├──     fn draw(&self)
+│   │   └── fn main()
+│   ├── 📄 sql_test.sql (270 tokens, 51 lines)
+│   │   ├── CREATE TABLE promoters
+│   │   ├──    user_id serial PRIMARY KEY,
+│   │   ├──    type varchar(20) NOT NULL,
+│   │   ├──    username varchar(20) NOT NULL,
+│   │   ├──    password varchar(20) NOT NULL,
+│   │   ├──    email varchar(30) NOT NULL,
+│   │   ├──    phone varchar(20) NOT NULL,
+│   │   ├──    promocode varchar(20),
+│   │   ├──    info json,
+│   │   ├──    going text[],
+│   │   ├──    invites text[],
+│   │   ├──    balance integer NOT NULL,
+│   │   ├──    rewards text[],
+│   │   ├──    created timestamp
+│   │   ├── CREATE TABLE events
+│   │   ├──    event_id serial PRIMARY KEY,
+│   │   ├──    name varchar(64) NOT NULL,
+│   │   ├──    date varchar(64) NOT NULL,
+│   │   ├──    location varchar(64) NOT NULL,
+│   │   ├──    performer varchar(64) NOT NULL,
+│   │   ├──    rewards json,
+│   │   └──    created timestamp
+│   ├── 📄 standard-app-routing.module.ts (100 tokens, 16 lines)
+│   │   └── const routes: Routes = [
+│   │         { path: '', component: HomeComponent },
+│   │         {
+│   │           path: 'heroes',
+│   │           component: HeroesListComponent,
+│   │           children: [
+│   │             { path: ':id', component: HeroDetailComponent },
+│   │             { path: 'new', component: HeroFormComponent },
+│   │           ],
+│   │         },
+│   │         { path: '**', component: PageNotFoundComponent },
+│   │       ];
+│   ├── 📄 test.env (190 tokens, 25 lines)
+│   │   ├── PROMO_PATH
+│   │   ├── PRODUCTION
+│   │   ├── SQL_SCHEMA_PATH
+│   │   ├── DB_LOGS
+│   │   ├── DB_LOG
+│   │   ├── PGPASSWORD
+│   │   ├── PGDATABASE
+│   │   ├── PGHOST
+│   │   ├── PGPORT
+│   │   ├── PGUSER
+│   │   ├── SERVER_LOG
+│   │   ├── SERVER_LOGS
+│   │   ├── API_URL
+│   │   ├── APP_LOGS
+│   │   ├── APP_LOG
+│   │   ├── APP_URL
+│   │   ├── COGNITO_USER_POOL_ID
+│   │   ├── COGNITO_APP_CLIENT_ID
+│   │   ├── AWS_REGION
+│   │   └── STRIPE_SECRET_KEY
+│   ├── 📄 testJsonSchema.json (421 tokens, 48 lines)
+│   │   ├── $schema: http://json-schema.org/draft-07/schema#
+│   │   ├── type: object
+│   │   ├── title: random_test
+│   │   └── description: A promoter's activites related to events
+│   ├── 📄 testPackage.json (349 tokens, 43 lines)
+│   │   ├── name: 'promo-app'
+│   │   ├── version: 0.0.0
+│   │   ├── scripts:
+│   │   ├──     ng: 'ng'
+│   │   ├──     start: 'ng serve'
+│   │   ├──     build: 'ng build'
+│   │   ├──     watch: 'ng build --watch --configuration development'
+│   │   └──     test: 'ng test'
+│   └── 📄 tickets.component.ts (7,160 tokens, 903 lines)
+│       ├── interface EnrichedTicket extends Ticket
+│       ├── interface SpinConfig
+│       ├── interface RotationState
+│       ├── interface SpeakInput
+│       ├── const formatSpeakInput = (input: SpeakInput): string =>
+│       ├── function hourToSpeech(hour: number, minute: number, period: string):
+│       │   string
+│       ├── export class TicketsComponent implements AfterViewInit
+│       ├──   speak(input: SpeakInput)
+│       ├──   speakEvent(ticket: EnrichedTicket): void
+│       ├──   formatEvent(ticket: EnrichedTicket): string
+│       ├──   speakVenue(ticket: EnrichedTicket): void
+│       ├──   formatDate(date: Date, oneLiner: boolean = false): string
+│       ├──   formatDateForSpeech(date: Date): string
+│       ├──   async spinQRCode(
+│       │       event: PointerEvent,
+│       │       config: SpinConfig = DEFAULT_SPIN_CONFIG
+│       │     )
+│       ├──   private animateRotation(
+│       │       imgElement: HTMLElement,
+│       │       targetRotation: number,
+│       │       config: SpinConfig,
+│       │       cleanup: () => void
+│       │     )
+│       ├──     const animate = (currentTime: number) =>
+│       ├──         requestAnimationFrame(animate)
+│       ├──         cleanup()
+│       ├──     requestAnimationFrame(animate)
+│       ├──   private getNext90Degree(currentRotation: number): number
+│       ├──   private getCurrentRotation(matrix: string): number
+│       ├──   ngAfterViewInit()
+│       ├──       const mouseEnterListener = () =>
+│       ├──       const mouseLeaveListener = () =>
+│       ├──   ngOnDestroy()
+│       ├──   toggleColumn(event: MatOptionSelectionChange, column: string)
+│       ├──   adjustColumns(event?: Event)
+│       ├──   onResize(event: Event)
+│       ├──   async ngOnInit()
+│       ├──   async loadTickets(): Promise<void>
+│       ├──   onDateRangeChange(
+│       │       type: "start" | "end",
+│       │       event: MatDatepickerInputEvent<Date>
+│       │     )
+│       ├──   applyFilter(column: string): void
+│       ├──   formatDateForComparison(date: Date): string
+│       ├──   constructor(private renderer: Renderer2)
+│       ├──   onFilterChange(event: Event, column: string)
+│       ├──   onLatitudeChange(event: Event)
+│       ├──   onLongitudeChange(event: Event)
+│       ├──   onRadiusChange(event: Event)
+│       ├──   sortData(sort: Sort): void
+│       ├──   onRowClick(event: Event, row: any)
+│       ├── function isDate(value: Date | undefined | null): value is Date
+│       ├── function isNonNullNumber(value: number | null): value is number
+│       ├── function hasLocation(
+│       │     ticket: any
+│       │   ): ticket is
+│       ├── const create_faker_ticket = async () =>
+│       ├── function compare(a: number | string, b: number | string, isAsc:
+│       │   boolean)
+│       ├── function compare_dates(a: Date, b: Date, isAsc: boolean)
+│       ├── async function mockMoreTickets(): Promise<Ticket[]>
+│       ├── const mockTickets = async () =>
+│       └── const renderQRCode = async (text: String): Promise<string> =>
+├── 📁 group6 (1 folder, 14 files)
+│   ├── 📄 catastrophic.c (5,339 tokens, 754 lines)
+│   │   ├── TODO: technically we should use a proper parser
+│   │   ├── struct Point
+│   │   ├──     int x;
+│   │   ├──     int y;
+│   │   ├── struct Point getOrigin()
+│   │   ├── float mul_two_floats(float x1, float x2)
+│   │   ├── enum days
+│   │   ├──     SUN,
+│   │   ├──     MON,
+│   │   ├──     TUE,
+│   │   ├──     WED,
+│   │   ├──     THU,
+│   │   ├──     FRI,
+│   │   ├──     SAT
+│   │   ├── enum worker_pool_flags
+│   │   ├──         POOL_BH                 = 1 << 0,
+│   │   ├──         POOL_MANAGER_ACTIVE     = 1 << 1,
+│   │   ├──         POOL_DISASSOCIATED      = 1 << 2,
+│   │   ├──         POOL_BH_DRAINING        = 1 << 3,
+│   │   ├── enum worker_flags
+│   │   ├──         WORKER_DIE              = 1 << 1,
+│   │   ├──         WORKER_IDLE             = 1 << 2,
+│   │   ├──         WORKER_PREP             = 1 << 3,
+│   │   ├──         WORKER_CPU_INTENSIVE    = 1 << 6,
+│   │   ├──         WORKER_UNBOUND          = 1 << 7,
+│   │   ├──         WORKER_REBOUND          = 1 << 8,
+│   │   ├──         WORKER_NOT_RUNNING      = WORKER_PREP | WORKER_CPU_INTENSIVE
+│   │   │   |
+│   │   │                                     WORKER_UNBOUND | WORKER_REBOUND,
+│   │   ├── struct worker_pool
+│   │   ├──   raw_spinlock_t    lock;
+│   │   ├──   int      cpu;
+│   │   ├──   int      node;
+│   │   ├──   int      id;
+│   │   ├──   unsigned int    flags;
+│   │   ├──   unsigned long    watchdog_ts;
+│   │   ├──   bool      cpu_stall;
+│   │   ├──   int      nr_running;
+│   │   ├──   struct list_head  worklist;
+│   │   ├──   int      nr_workers;
+│   │   ├──   int      nr_idle;
+│   │   ├──   struct list_head  idle_list;
+│   │   ├──   struct timer_list  idle_timer;
+│   │   ├──   struct work_struct      idle_cull_work;
+│   │   ├──   struct timer_list  mayday_timer;
+│   │   ├──   struct worker    *manager;
+│   │   ├──   struct list_head  workers;
+│   │   ├──   struct ida    worker_ida;
+│   │   ├──   struct workqueue_attrs  *attrs;
+│   │   ├──   struct hlist_node  hash_node;
+│   │   ├──   int      refcnt;
+│   │   ├──   struct rcu_head    rcu;
+│   │   ├── long add_two_longs(long x1, long x2)
+│   │   ├── double multiplyByTwo(double num)
+│   │   ├── char getFirstCharacter(char *str)
+│   │   ├── void greet(Person p)
+│   │   ├── typedef struct
+│   │   ├──     char name[50];
+│   │   ├── } Person;
+│   │   ├── typedef struct PersonA
+│   │   ├──     char name[50];
+│   │   ├── } PersonB;
+│   │   ├── int main()
+│   │   ├── int* getArrayStart(int arr[], int size)
+│   │   ├── long complexFunctionWithMultipleArguments(
+│   │   │       int param1,
+│   │   │       double param2,
+│   │   │       char *param3,
+│   │   │       struct Point point
+│   │   │   )
+│   │   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   │   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   │   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   │   *channel, int channellen, int is_pattern)
+│   │   ├──     while((ln = listNext(&li)))
+│   │   ├── static struct config
+│   │   ├──     aeEventLoop *el;
+│   │   ├──     cliConnInfo conn_info;
+│   │   ├──     const char *hostsocket;
+│   │   ├──     int tls;
+│   │   ├──     struct cliSSLconfig sslconfig;
+│   │   ├── } config;
+│   │   ├── class Person
+│   │   ├──     std::string name;
+│   │   ├── public:
+│   │   ├──     Person(std::string n) : name(n)
+│   │   ├──     void greet()
+│   │   ├── void globalGreet()
+│   │   ├── int main()
+│   │   ├── void printMessage(const std::string &message)
+│   │   ├── template<typename T>
+│   │   │   void printVector(const std::vector<T>& vec)
+│   │   ├── struct foo
+│   │   ├──     char x;
+│   │   ├──     struct foo_in
+│   │   ├──           char* y;
+│   │   ├──           short z;
+│   │   ├──     } inner;
+│   │   ├── struct Point
+│   │   ├──     int x, y;
+│   │   ├──     Point(int x, int y) : x(x), y(y)
+│   │   ├── class Animal
+│   │   ├──   public:
+│   │   ├──     Animal(const std::string &name) : name(name)
+│   │   ├──     virtual void speak() const
+│   │   ├──     virtual ~Animal()
+│   │   ├── protected:
+│   │   ├──     std::string name;
+│   │   ├── class Dog : public Animal
+│   │   ├──   public:
+│   │   ├──     Dog(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── class Cat : public Animal
+│   │   ├──   public:
+│   │   ├──     Cat(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── class CatDog: public Animal, public Cat, public Dog
+│   │   ├──   public:
+│   │   ├──       CatDog(const std::string &name) : Animal(name)
+│   │   ├──       int meow_bark()
+│   │   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   │   num_layers,
+│   │   │                                int batch_size, int max_seq_length,
+│   │   │   float dropout,
+│   │   │                                bool bidirectional, bool
+│   │   │   cudnn_allow_tf32,
+│   │   │                                int workspace_size, int
+│   │   │   reserve_space_size)
+│   │   ├── int main()
+│   │   ├── enum ECarTypes
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV,
+│   │   ├──   Wagon
+│   │   ├── ECarTypes GetPreferredCarType()
+│   │   ├── enum ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── enum class ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── void myFunction(string fname, int age)
+│   │   ├── template <typename T> T cos(T)
+│   │   ├── template <typename T> T sin(T)
+│   │   ├── template <typename T> T sqrt(T)
+│   │   ├── template<typename T> struct VLEN
+│   │   ├── template<typename T> class arr
+│   │   ├──   private:
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──   public:
+│   │   ├──     arr() : p(0), sz(0)
+│   │   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   │   ├──     arr(arr &&other)
+│   │   │         : p(other.p), sz(other.sz)
+│   │   ├──     ~arr()
+│   │   ├──     void resize(size_t n)
+│   │   ├──     T &operator[](size_t idx)
+│   │   ├──     T *data()
+│   │   ├──     size_t size() const
+│   │   ├── class Buffer
+│   │   ├──  private:
+│   │   ├──   void* ptr_;
+│   │   ├── std::tuple<array, array, array> quantize(
+│   │   │       const array& w,
+│   │   │       int group_size,
+│   │   │       int bits,
+│   │   │       StreamOrDevice s)
+│   │   ├── #define PY_SSIZE_T_CLEAN
+│   │   ├── #define PLATFORM_IS_X86
+│   │   ├── #define PLATFORM_WINDOWS
+│   │   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   │   ├── static int GetXCR0EAX()
+│   │   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   │   ├── static int GetXCR0EAX()
+│   │   ├──   asm("XGETBV" : "=a"(eax), "=d"(edx) : "c"(0))
+│   │   ├── static void ReportMissingCpuFeature(const char* name)
+│   │   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   │   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   │   ├── static PyMethodDef cpu_feature_guard_methods[]
+│   │   ├── static struct PyModuleDef cpu_feature_guard_module
+│   │   ├── #define EXPORT_SYMBOL __declspec(dllexport)
+│   │   ├── #define EXPORT_SYMBOL __attribute__ ((visibility("default")))
+│   │   ├── EXPORT_SYMBOL PyMODINIT_FUNC PyInit_cpu_feature_guard(void)
+│   │   ├── typedef struct
+│   │   ├──     GPT2Config config;
+│   │   ├──     ParameterTensors params;
+│   │   ├──     size_t param_sizes[NUM_PARAMETER_TENSORS];
+│   │   ├──     float* params_memory;
+│   │   ├──     size_t num_parameters;
+│   │   ├──     ParameterTensors grads;
+│   │   ├──     float* grads_memory;
+│   │   ├──     float* m_memory;
+│   │   ├──     float* v_memory;
+│   │   ├──     ActivationTensors acts;
+│   │   ├──     size_t act_sizes[NUM_ACTIVATION_TENSORS];
+│   │   ├──     float* acts_memory;
+│   │   ├──     size_t num_activations;
+│   │   ├──     ActivationTensors grads_acts;
+│   │   ├──     float* grads_acts_memory;
+│   │   ├──     int batch_size;
+│   │   ├──     int seq_len;
+│   │   ├──     int* inputs;
+│   │   ├──     int* targets;
+│   │   ├──     float mean_loss;
+│   │   └── } GPT2;
+│   ├── 📄 cpp_examples_impl.cc (60 tokens, 10 lines)
+│   │   ├── PYBIND11_MODULE(cpp_examples, m)
+│   │   └──     m.def("add", &add<int>, "An example function to add two
+│   │       numbers.")
+│   ├── 📄 cpp_examples_impl.cu (37 tokens, 10 lines)
+│   │   ├── template <typename T>
+│   │   │   T add(T a, T b)
+│   │   └── template <>
+│   │       int add<int>(int a, int b)
+│   ├── 📄 cpp_examples_impl.h (22 tokens, 6 lines)
+│   │   ├── template <typename T>
+│   │   │   T add(T a, T b)
+│   │   └── template <>
+│   │       int add<int>(int, int)
+│   ├── 📄 edge_case.hpp (426 tokens, 28 lines)
+│   ├── 📄 fractal.thy (1,712 tokens, 147 lines)
+│   │   ├── Title:      fractal.thy
+│   │   ├── Author:     Isabelle/HOL Contributors!
+│   │   ├── Author:     edge cases r us
+│   │   ├── theory Simplified_Ring
+│   │   ├── section ‹Basic Algebraic Structures›
+│   │   ├── class everything = nothing + itself
+│   │   ├── subsection ‹Monoids›
+│   │   ├── definition ring_hom :: "[('a, 'm) ring_scheme, ('b, 'n) ring_scheme]
+│   │   │   => ('a => 'b) set"
+│   │   ├── fun example_fun :: "nat ⇒ nat"
+│   │   ├── locale monoid =
+│   │   │     fixes G (structure)
+│   │   │     assumes m_closed: "⟦x ∈ carrier G; y ∈ carrier G⟧ ⟹  x ⊗ y ∈
+│   │   │   carrier G"
+│   │   │       and m_assoc: "⟦x ∈ carrier G; y ∈ carrier G; z ∈ carrier G⟧ ⟹
+│   │   │   (x ⊗ y) ⊗ z = x ⊗ (y ⊗ z)"
+│   │   │       and one_closed: "𝟭 ∈ carrier G"
+│   │   │       and l_one: "x ∈ carrier G ⟹  𝟭 ⊗ x = x"
+│   │   │       and r_one: "x ∈ carrier G ⟹  x ⊗ 𝟭 = x"
+│   │   ├── subsection ‹Groups›
+│   │   ├── locale group = monoid +
+│   │   │     assumes Units_closed: "x ∈ Units G ⟹  x ∈ carrier G"
+│   │   │       and l_inv_ex: "x ∈ carrier G ⟹  ∃ y ∈ carrier G. y ⊗ x = 𝟭"
+│   │   │       and r_inv_ex: "x ∈ carrier G ⟹  ∃ y ∈ carrier G. x ⊗ y = 𝟭"
+│   │   ├── subsection ‹Rings›
+│   │   ├── locale ring = abelian_group R + monoid R +
+│   │   │     assumes l_distr: "⟦x ∈ carrier R; y ∈ carrier R; z ∈ carrier R⟧ ⟹
+│   │   │   (x ⊕ y) ⊗ z = x ⊗ z ⊕ y ⊗ z"
+│   │   │       and r_distr: "⟦x ∈ carrier R; y ∈ carrier R; z ∈ carrier R⟧ ⟹  z
+│   │   │   ⊗ (x ⊕ y) = z ⊗ x ⊕ z ⊗ y"
+│   │   ├── locale commutative_ring = ring +
+│   │   │     assumes m_commutative: "⟦x ∈ carrier R; y ∈ carrier R⟧ ⟹  x ⊗ y =
+│   │   │   y ⊗ x"
+│   │   ├── locale domain = commutative_ring +
+│   │   │     assumes no_zero_divisors: "⟦a ⊗ b = 𝟬; a ∈ carrier R; b ∈ carrier
+│   │   │   R⟧ ⟹  a = 𝟬 ∨ b = 𝟬"
+│   │   ├── locale field = domain +
+│   │   │     assumes inv_ex: "x ∈ carrier R - {𝟬} ⟹  inv x ∈ carrier R"
+│   │   ├── subsection ‹Morphisms›
+│   │   ├── lemma example_lemma: "example_fun n = n"
+│   │   ├── qualified lemma gcd_0:
+│   │   │     "gcd a 0 = normalize a"
+│   │   ├── lemma abelian_monoidI:
+│   │   │     fixes R (structure)
+│   │   │         and f :: "'edge::{} ⇒ 'case::{}"
+│   │   │     assumes "⋀x y. ⟦ x ∈ carrier R; y ∈ carrier R ⟧ ⟹  x ⊕ y ∈ carrier
+│   │   │   R"
+│   │   │         and "𝟬 ∈ carrier R"
+│   │   │         and "⋀x y z. ⟦ x ∈ carrier R; y ∈ carrier R; z ∈ carrier R ⟧ ⟹
+│   │   │   (x ⊕ y) ⊕ z = x ⊕ (y ⊕ z)"
+│   │   │     shows "abelian_monoid R"
+│   │   ├── lemma euclidean_size_gcd_le1 [simp]:
+│   │   │     assumes "a ≠ 0"
+│   │   │     shows "euclidean_size (gcd a b) ≤ euclidean_size a"
+│   │   ├── theorem Residue_theorem:
+│   │   │     fixes S pts::"complex set" and f::"complex ⇒ complex"
+│   │   │       and g::"real ⇒ complex"
+│   │   │     assumes "open S" "connected S" "finite pts" and
+│   │   │             holo:"f holomorphic_on S-pts" and
+│   │   │             "valid_path g" and
+│   │   │             loop:"pathfinish g = pathstart g" and
+│   │   │             "path_image g ⊆ S-pts" and
+│   │   │             homo:"∀z. (z ∉ S) ⟶  winding_number g z  = 0"
+│   │   │     shows "contour_integral g f = 2 * pi * 𝗂 *(∑p ∈ pts.
+│   │   │   winding_number g p * residue f p)"
+│   │   ├── corollary fps_coeff_residues_bigo':
+│   │   │     fixes f :: "complex ⇒ complex" and r :: real
+│   │   │     assumes exp: "f has_fps_expansion F"
+│   │   │     assumes "open A" "connected A" "cball 0 r ⊆ A" "r > 0"
+│   │   │     assumes "f holomorphic_on A - S" "S ⊆ ball 0 r" "finite S" "0 ∉ S"
+│   │   │     assumes "eventually (λn. g n = -(∑z ∈ S. residue (λz. f z / z ^
+│   │   │   Suc n) z)) sequentially"
+│   │   │                (is "eventually (λn. _ = -?g' n) _")
+│   │   │     shows   "(λn. fps_nth F n - g n) ∈ O(λn. 1 / r ^ n)" (is "(λn. ?c
+│   │   │   n - _) ∈ O(_)")
+│   │   └── end
+│   ├── 📄 Microsoft.PowerShell_profile.ps1 (3,346 tokens, 497 lines)
+│   │   ├── function Log($message)
+│   │   ├── function Remove-ChocolateyFromPath
+│   │   ├── function Show-Profiles
+│   │   ├── function Show-Path
+│   │   ├── function Show-Error($err)
+│   │   ├── function Get-ScoopPackagePath
+│   │   ├──         param(
+│   │   │                   [Parameter(Mandatory = $true)]
+│   │   │                   [string]$PackageName)
+│   │   ├── function Check-Command
+│   │   ├──         param(
+│   │   │                   [Parameter(Mandatory = $true)]
+│   │   │                   [string]$Name)
+│   │   ├── function Add-ToPath
+│   │   ├──         param(
+│   │   │                   [Parameter(Mandatory = $true)]
+│   │   │                   [string]$PathToAdd)
+│   │   ├── function Install-Scoop
+│   │   ├── function Scoop-Install
+│   │   ├──         param(
+│   │   │                   [Parameter(Mandatory = $true)]
+│   │   │                   [string]$Name,
+│   │   │                   [string]$PathToAdd)
+│   │   ├── function Start-CondaEnv
+│   │   ├── function Install-PipPackage
+│   │   ├──         param(
+│   │   │           [Parameter(Mandatory = $true)]
+│   │   │                   [string]$PackageName)
+│   │   ├── function Install-VSBuildTools
+│   │   ├── function Install-Crate
+│   │   ├──         param(
+│   │   │           [Parameter(Mandatory = $true)]
+│   │   │                   [string]$CrateName)
+│   │   ├── function Get-ScoopVersion
+│   │   ├── function Get-Version
+│   │   ├──     param(
+│   │   │           [Parameter(Mandatory = $true)]
+│   │   │           [string]$ExecutablePath,
+│   │   │           [string]$ExecutableName)
+│   │   ├── function Show-Requirements
+│   │   ├──         function Measure-Status
+│   │   ├──                 param(
+│   │   │                           [Parameter(Mandatory = $true)]
+│   │   │                           [string]$Name)
+│   │   ├── function Find-Profile
+│   │   ├── function Edit-Profile
+│   │   ├── function Set-Profile
+│   │   └── function Show-Profile
+│   ├── 📄 python_complex_class.py (10 tokens, 2 lines)
+│   │   └── class Box(Space[NDArray[Any]])
+│   ├── 📄 ramda__cloneRegExp.js (173 tokens, 9 lines)
+│   │   └── export default function _cloneRegExp(pattern)
+│   ├── 📄 ramda_prop.js (646 tokens, 85 lines)
+│   │   ├── /**
+│   │   │    * Returns a function that when supplied an object returns the
+│   │   │   indicated
+│   │   │    * property of that object, if it exists.
+│   │   │    * @category Object
+│   │   │    * @typedefn Idx = String | Int | Symbol
+│   │   │    * @sig Idx -> {s: a} -> a | Undefined
+│   │   │    * @param {String|Number} p The property name or array index
+│   │   │    * @param {Object} obj The object to query
+│   │   │    * @return {*} The value at `obj.p`.
+│   │   │    */
+│   │   │   var prop = _curry2(function prop(p, obj)
+│   │   ├── /**
+│   │   │    * Solves equations of the form a * x = b
+│   │   │    * @param {{
+│   │   │    *  z: number
+│   │   │    * }} x
+│   │   │    */
+│   │   │   function foo(x)
+│   │   ├── /**
+│   │   │    * Deconstructs an array field from the input documents to output a
+│   │   │   document for each element.
+│   │   │    * Each output document is the input document with the value of the
+│   │   │   array field replaced by the element.
+│   │   │    * @category Object
+│   │   │    * @sig String -> {k: [v]} -> [{k: v}]
+│   │   │    * @param {String} key The key to determine which property of the
+│   │   │   object should be unwound.
+│   │   │    * @param {Object} object The object containing the list to unwind
+│   │   │   at the property named by the key.
+│   │   │    * @return {List} A list of new objects, each having the given key
+│   │   │   associated to an item from the unwound list.
+│   │   │    */
+│   │   │   var unwind = _curry2(function(key, object)
+│   │   └──   return _map(function(item)
+│   ├── 📄 tensorflow_flags.h (7,628 tokens, 668 lines)
+│   │   ├── TF_DECLARE_FLAG('test_only_experiment_1')
+│   │   ├── TF_DECLARE_FLAG('test_only_experiment_2')
+│   │   ├── TF_DECLARE_FLAG('enable_nested_function_shape_inference'):
+│   │   │           Allow ops such as tf.cond to invoke the ShapeRefiner on
+│   │   │   their nested functions.
+│   │   ├── TF_DECLARE_FLAG('enable_quantized_dtypes_training'):
+│   │   │           Set quantized dtypes, like tf.qint8, to be trainable.
+│   │   ├── TF_DECLARE_FLAG('graph_building_optimization'):
+│   │   │           Optimize graph building for faster tf.function tracing.
+│   │   ├── TF_DECLARE_FLAG('saved_model_fingerprinting'):
+│   │   │           Add fingerprint to SavedModels.
+│   │   ├── TF_DECLARE_FLAG('more_stack_traces'):
+│   │   │           Enable experimental code that preserves and propagates graph
+│   │   │   node stack traces in C++.
+│   │   ├── TF_DECLARE_FLAG('publish_function_graphs'):
+│   │   │           Enables the publication of partitioned function graphs via
+│   │   │   StatsPublisherInterface. Disabling this flag can reduce memory
+│   │   │   consumption.
+│   │   ├── TF_DECLARE_FLAG('enable_aggressive_constant_replication'):
+│   │   │           Replicate constants across CPU devices and even for local
+│   │   │   CPUs within the same task if available.
+│   │   ├── TF_DECLARE_FLAG('enable_colocation_key_propagation_in_while_op_lower
+│   │   │   ing'):
+│   │   │           If true, colocation key attributes for the ops will be
+│   │   │   propagated during while op lowering to switch/merge ops.
+│   │   ├── Flag('tf_xla_auto_jit'):
+│   │   │           Control compilation of operators into XLA computations on
+│   │   │   CPU and GPU devices.  0 = use ConfigProto setting; -1 = off; 1 = on
+│   │   │   for things very likely to be improved; 2 = on for everything;
+│   │   │   (experimental) fusible = only for Tensorflow operations that XLA
+│   │   │   knows how to fuse. If set to single-gpu(<N>) then this resolves to
+│   │   │   <N> for single-GPU graphs (graphs that have at least one node placed
+│   │   │   on a GPU and no more than one GPU is in use through the entire
+│   │   │   graph) and 0 otherwise.  Experimental.
+│   │   ├── Flag('tf_xla_min_cluster_size'):
+│   │   │           Minimum number of operators in an XLA compilation. Ignored
+│   │   │   for operators placed on an XLA device or operators explicitly marked
+│   │   │   for compilation.
+│   │   ├── Flag('tf_xla_max_cluster_size'):
+│   │   │           Maximum number of operators in an XLA compilation.
+│   │   ├── Flag('tf_xla_cluster_exclude_ops'):
+│   │   │           (experimental) Exclude the operations from auto-clustering.
+│   │   │   If multiple, separate them with commas. Where, Some_other_ops.
+│   │   ├── Flag('tf_xla_clustering_debug'):
+│   │   │           Dump graphs during XLA compilation.
+│   │   ├── Flag('tf_xla_cpu_global_jit'):
+│   │   │           Enables global JIT compilation for CPU via SessionOptions.
+│   │   ├── Flag('tf_xla_clustering_fuel'):
+│   │   │           Places an artificial limit on the number of ops marked as
+│   │   │   eligible for clustering.
+│   │   ├── Flag('tf_xla_disable_deadness_safety_checks_for_debugging'):
+│   │   │           Disable deadness related safety checks when clustering (this
+│   │   │   is unsound).
+│   │   ├── Flag('tf_xla_disable_resource_variable_safety_checks_for_debugging')
+│   │   │   :
+│   │   │           Disable resource variables related safety checks when
+│   │   │   clustering (this is unsound).
+│   │   ├── Flag('tf_xla_deterministic_cluster_names'):
+│   │   │           Causes the function names assigned by auto clustering to be
+│   │   │   deterministic from run to run.
+│   │   ├── Flag('tf_xla_persistent_cache_directory'):
+│   │   │           If non-empty, JIT-compiled executables are saved to and
+│   │   │   loaded from the specified file system directory path. Empty by
+│   │   │   default.
+│   │   ├── Flag('tf_xla_persistent_cache_device_types'):
+│   │   │           If non-empty, the persistent cache will only be used for the
+│   │   │   specified devices (comma separated). Each device type should be able
+│   │   │   to be converted to.
+│   │   ├── Flag('tf_xla_persistent_cache_read_only'):
+│   │   │           If true, the persistent cache will be read-only.
+│   │   ├── Flag('tf_xla_disable_strict_signature_checks'):
+│   │   │           If true, entires loaded into the XLA compile cache will not
+│   │   │   have their signatures checked strictly. Defaults to false.
+│   │   ├── Flag('tf_xla_persistent_cache_prefix'):
+│   │   │           Specifies the persistance cache prefix. Default is.
+│   │   ├── Flag('tf_xla_sparse_core_disable_table_stacking'):
+│   │   │           Disable table stacking for all the tables passed to the
+│   │   │   SparseCore mid level API.
+│   │   ├── Flag('tf_xla_sparse_core_minibatch_max_division_level'):
+│   │   │           Max level of division to split input data into minibatches.
+│   │   ├── Flag('tf_xla_sparse_core_stacking_mem_limit_bytes'):
+│   │   │           If non-zero, limits the size of the activations for a given
+│   │   │   table to be below these many bytes.
+│   │   ├── Flag('tf_xla_sparse_core_stacking_table_shard_limit_bytes'):
+│   │   │           If non-zero, limits the size of any table shard to be below
+│   │   │   these many bytes.
+│   │   ├── Flag('always_specialize')
+│   │   ├── Flag('cost_driven_async_parallel_for')
+│   │   ├── Flag('enable_crash_reproducer')
+│   │   ├── Flag('log_query_of_death')
+│   │   ├── Flag('vectorize')
+│   │   ├── Flag('tf_xla_enable_lazy_compilation')
+│   │   ├── Flag('tf_xla_print_cluster_outputs'):
+│   │   │           If true then insert Print nodes to print out values produced
+│   │   │   by XLA clusters.
+│   │   ├── Flag('tf_xla_check_cluster_input_numerics'):
+│   │   │           If true then insert CheckNumerics nodes to check all cluster
+│   │   │   inputs.
+│   │   ├── Flag('tf_xla_check_cluster_output_numerics'):
+│   │   │           If true then insert CheckNumerics nodes to check all cluster
+│   │   │   outputs.
+│   │   ├── Flag('tf_xla_disable_constant_folding'):
+│   │   │           If true then disables constant folding on TF graph before
+│   │   │   XLA compilation.
+│   │   ├── Flag('tf_xla_disable_full_embedding_pipelining'):
+│   │   │           If true then disables full embedding pipelining and instead
+│   │   │   use strict SparseCore / TensorCore sequencing.
+│   │   ├── Flag('tf_xla_embedding_parallel_iterations'):
+│   │   │           If >0 then use this many parallel iterations in
+│   │   │   embedding_pipelining and embedding_sequency. By default, use the
+│   │   │   parallel_iterations on the original model WhileOp.
+│   │   ├── Flag('tf_xla_compile_on_demand'):
+│   │   │           Switch a device into 'on-demand' mode, where instead of
+│   │   │   autoclustering ops are compiled one by one just-in-time.
+│   │   ├── Flag('tf_xla_enable_xla_devices'):
+│   │   │           Generate XLA_* devices, where placing a computation on such
+│   │   │   a device forces compilation by XLA. Deprecated.
+│   │   ├── Flag('tf_xla_always_defer_compilation')
+│   │   ├── Flag('tf_xla_async_compilation'):
+│   │   │           When lazy compilation is enabled, asynchronous compilation
+│   │   │   starts the cluster compilation in the background, and the fallback
+│   │   │   path is executed until the compilation has finished.
+│   │   ├── Flag('tf_xla_use_device_api_for_xla_launch'):
+│   │   │           If true, uses Device API (PjRt) for single device
+│   │   │   compilation and execution of functions marked for JIT compilation
+│   │   │   i.e. jit_compile=True. Defaults to false.
+│   │   ├── Flag('tf_xla_use_device_api_for_compile_on_demand'):
+│   │   │           If true, uses Device API (PjRt) for compiling and executing
+│   │   │   ops one by one in 'on-demand' mode. Defaults to false.
+│   │   ├── Flag('tf_xla_use_device_api_for_auto_jit'):
+│   │   │           If true, uses Device API (PjRt) for compilation and
+│   │   │   execution when auto-clustering is enabled. Defaults to false.
+│   │   ├── Flag('tf_xla_use_device_api'):
+│   │   │           If true, uses Device API (PjRt) for compilation and
+│   │   │   execution of ops one-by-one in 'on-demand' mode, for functions
+│   │   │   marked for JIT compilation, or when auto-clustering is enabled.
+│   │   │   Defaults to false.
+│   │   ├── Flag('tf_xla_enable_device_api_for_gpu'):
+│   │   │           If true, uses Device API (PjRt) for TF GPU device. This is a
+│   │   │   helper flag so that individual tests can turn on PjRt for GPU
+│   │   │   specifically.
+│   │   ├── Flag('tf_xla_call_module_disabled_checks'):
+│   │   │           A comma-sepated list of directives specifying the safety
+│   │   │   checks to be skipped when compiling XlaCallModuleOp. See the op
+│   │   │   documentation for the recognized values.
+│   │   ├── Flag('tf_mlir_enable_mlir_bridge'):
+│   │   │           Enables experimental MLIR-Based TensorFlow Compiler Bridge.
+│   │   ├── Flag('tf_mlir_enable_merge_control_flow_pass'):
+│   │   │           Enables MergeControlFlow pass for MLIR-Based TensorFlow
+│   │   │   Compiler Bridge.
+│   │   ├── Flag('tf_mlir_enable_convert_control_to_data_outputs_pass'):
+│   │   │           Enables MLIR-Based TensorFlow Compiler Bridge.
+│   │   ├── Flag('tf_mlir_enable_strict_clusters'):
+│   │   │           Do not allow clusters that have cyclic control dependencies.
+│   │   ├── Flag('tf_mlir_enable_multiple_local_cpu_devices'):
+│   │   │           Enable multiple local CPU devices. CPU ops which are outside
+│   │   │   compiled inside the tpu cluster will also be replicated across
+│   │   │   multiple cpu devices.
+│   │   ├── Flag('tf_dump_graphs_in_tfg'):
+│   │   │           When tf_dump_graphs_in_tfg is true, graphs after
+│   │   │   transformations are dumped in MLIR TFG dialect and not in GraphDef.
+│   │   ├── Flag('tf_mlir_enable_generic_outside_compilation'):
+│   │   │           Enables OutsideCompilation passes for MLIR-Based TensorFlow
+│   │   │   Generic Compiler Bridge.
+│   │   ├── Flag('tf_mlir_enable_tpu_variable_runtime_reformatting_pass'):
+│   │   │           Enables TPUVariableRuntimeReformatting pass for MLIR-Based
+│   │   │   TensorFlow Compiler Bridge. This enables weight update sharding and
+│   │   │   creates TPUReshardVariables ops.
+│   │   ├── TF_PY_DECLARE_FLAG('test_only_experiment_1')
+│   │   ├── TF_PY_DECLARE_FLAG('test_only_experiment_2')
+│   │   ├── TF_PY_DECLARE_FLAG('enable_nested_function_shape_inference')
+│   │   ├── TF_PY_DECLARE_FLAG('enable_quantized_dtypes_training')
+│   │   ├── TF_PY_DECLARE_FLAG('graph_building_optimization')
+│   │   ├── TF_PY_DECLARE_FLAG('op_building_optimization')
+│   │   ├── TF_PY_DECLARE_FLAG('saved_model_fingerprinting')
+│   │   ├── TF_PY_DECLARE_FLAG('tf_shape_default_int64')
+│   │   ├── TF_PY_DECLARE_FLAG('more_stack_traces')
+│   │   ├── TF_PY_DECLARE_FLAG('publish_function_graphs')
+│   │   ├── TF_PY_DECLARE_FLAG('enable_aggressive_constant_replication')
+│   │   ├── TF_PY_DECLARE_FLAG('enable_colocation_key_propagation_in_while_op_lo
+│   │   │   wering')
+│   │   ├── #define TENSORFLOW_CORE_CONFIG_FLAG_DEFS_H_
+│   │   ├── class Flags
+│   │   ├──  public:
+│   │   ├── bool SetterForXlaAutoJitFlag(const string& value)
+│   │   ├── bool SetterForXlaCallModuleDisabledChecks(const string& value)
+│   │   ├── void AppendMarkForCompilationPassFlagsInternal(std::vector<Flag>*
+│   │   │   flag_list)
+│   │   ├── void AllocateAndParseJitRtFlags()
+│   │   ├── void AllocateAndParseFlags()
+│   │   ├── void ResetFlags()
+│   │   ├── bool SetXlaAutoJitFlagFromFlagString(const string& value)
+│   │   ├── BuildXlaOpsPassFlags* GetBuildXlaOpsPassFlags()
+│   │   ├── MarkForCompilationPassFlags* GetMarkForCompilationPassFlags()
+│   │   ├── XlaSparseCoreFlags* GetXlaSparseCoreFlags()
+│   │   ├── XlaDeviceFlags* GetXlaDeviceFlags()
+│   │   ├── XlaOpsCommonFlags* GetXlaOpsCommonFlags()
+│   │   ├── XlaCallModuleFlags* GetXlaCallModuleFlags()
+│   │   ├── MlirCommonFlags* GetMlirCommonFlags()
+│   │   ├── void ResetJitCompilerFlags()
+│   │   ├── const JitRtFlags& GetJitRtFlags()
+│   │   ├── ConfigProto::Experimental::MlirBridgeRollout
+│   │   │   GetMlirBridgeRolloutState(
+│   │   │       std::optional<const ConfigProto> config_proto)
+│   │   ├── void AppendMarkForCompilationPassFlags(std::vector<Flag>* flag_list)
+│   │   ├── void DisableXlaCompilation()
+│   │   ├── void EnableXlaCompilation()
+│   │   ├── bool FailOnXlaCompilation()
+│   │   ├── #define TF_PY_DECLARE_FLAG(flag_name)
+│   │   └── PYBIND11_MODULE(flags_pybind, m)
+│   ├── 📄 test.f (181 tokens, 30 lines)
+│   │   ├── MODULE basic_mod
+│   │   ├──     TYPE :: person
+│   │   │           CHARACTER(LEN=50) :: name
+│   │   │           INTEGER :: age
+│   │   │       END TYPE person
+│   │   ├──     SUBROUTINE short_hello(happy, path)
+│   │   │       END SUBROUTINE short_hello
+│   │   ├──     SUBROUTINE long_hello(
+│   │   │           p,
+│   │   │           message
+│   │   │       )
+│   │   │       END SUBROUTINE long_hello
+│   │   ├── END MODULE basic_mod
+│   │   └── PROGRAM HelloFortran
+│   │       END PROGRAM HelloFortran
+│   ├── 📄 torch.rst (60 tokens, 8 lines)
+│   │   ├── # libtorch (C++-only)
+│   │   └── - Building libtorch using Python
+│   └── 📄 yc.html (9,063 tokens, 169 lines)
+├── 📁 group7 (1 folder, 5 files)
+│   ├── 📄 absurdly_huge.jsonl (8,347 tokens, 126 lines)
+│   │   ├── SMILES: str
+│   │   ├── Yield: float
+│   │   ├── Temperature: int
+│   │   ├── Pressure: float
+│   │   ├── Solvent: str
+│   │   ├── Success: bool
+│   │   ├── Reaction_Conditions: dict
+│   │   ├── Products: list
+│   │   └── EdgeCasesMissed: None
+│   ├── 📄 angular_crud.ts (1,192 tokens, 148 lines)
+│   │   ├── interface DBCommand<T = any>
+│   │   ├── export class IndexedDbService
+│   │   ├──     constructor()
+│   │   ├──     async create_connection({ db_name = 'client_db', table_name }:
+│   │   │   DBCommand)
+│   │   ├──                 upgrade(db)
+│   │   ├──     async create_model({ db_name, table_name, model }: DBCommand)
+│   │   ├──         verify_matching({ table_name, model })
+│   │   ├──     async read_key({ db_name, table_name, key }: DBCommand)
+│   │   ├──     async update_model({ db_name, table_name, model }: DBCommand)
+│   │   ├──         verify_matching({ table_name, model })
+│   │   ├──     async delete_key({ db_name, table_name, key }: DBCommand)
+│   │   ├──     async list_table({
+│   │   │           db_name,
+│   │   │           table_name,
+│   │   │           where,
+│   │   │       }: DBCommand & { where?: { [key: string]: string | number } })
+│   │   └──     async search_table(criteria: SearchCriteria)
+│   ├── 📄 structure.py (400 tokens, 92 lines)
+│   │   ├── @runtime_checkable
+│   │   │   class DataClass(Protocol)
+│   │   ├──     __dataclass_fields__: dict
+│   │   ├── class MyInteger(Enum)
+│   │   ├──     ONE = 1
+│   │   ├──     TWO = 2
+│   │   ├──     THREE = 42
+│   │   ├── class MyString(Enum)
+│   │   ├──     AAA1 = "aaa"
+│   │   ├──     BB_B = """edge
+│   │   │   case"""
+│   │   ├── @dataclass(frozen=True, slots=True, kw_only=True)
+│   │   │   class Tool
+│   │   ├──     name: str
+│   │   ├──     description: str
+│   │   ├──     input_model: DataClass
+│   │   ├──     output_model: DataClass
+│   │   ├──     def execute(self, *args, **kwargs)
+│   │   ├──     @property
+│   │   │       def edge_case(self) -> str
+│   │   ├──     def should_still_see_me(self, x: bool = True) -> "Tool"
+│   │   ├── @dataclass
+│   │   │   class MyInput[T]
+│   │   ├──     name: str
+│   │   ├──     rank: MyInteger
+│   │   ├──     serial_n: int
+│   │   ├── @dataclass
+│   │   │   class Thingy
+│   │   ├──     is_edge_case: bool
+│   │   ├── @dataclass
+│   │   │   class MyOutput
+│   │   ├──     orders: str
+│   │   ├── class MyTools(Enum)
+│   │   ├──     TOOL_A = Tool(
+│   │   │           name="complicated",
+│   │   │           description="edge case!",
+│   │   │           input_model=MyInput[Thingy],
+│   │   │           output_model=MyOutput,
+│   │   │       )
+│   │   ├──     TOOL_B = Tool(
+│   │   │           name="""super
+│   │   │   complicated
+│   │   │   """,
+│   │   │           description="edge case!",
+│   │   │           input_model=MyInput,
+│   │   │           output_model=MyOutput,
+│   │   │       )
+│   │   ├── @final
+│   │   │   class dtype(Generic[_DTypeScalar_co])
+│   │   └──     names: None | tuple[builtins.str, ...]
+│   ├── 📄 test.wgsl (528 tokens, 87 lines)
+│   │   ├── alias MyVec = vec4<f32>
+│   │   ├── alias AnotherVec = vec2<f32>
+│   │   ├── struct VertexInput
+│   │   ├── struct VertexOutput
+│   │   ├── struct MyUniforms
+│   │   ├── @group(0) @binding(0) var<uniform> u_mvp: mat4x4<f32>
+│   │   ├── @group(0) @binding(1) var<uniform> u_color: MyVec
+│   │   ├── @group(1) @binding(0) var my_texture: texture_2d<f32>
+│   │   ├── @group(1) @binding(1) var my_sampler: sampler
+│   │   ├── @vertex
+│   │   │   fn vs_main(in: VertexInput) -> VertexOutput
+│   │   ├── @fragment
+│   │   │   fn fs_main(in: VertexOutput) -> @location(0) vec4<f32>
+│   │   ├── @compute @workgroup_size(8, 8, 1)
+│   │   │   fn cs_main(@builtin(global_invocation_id) global_id: vec3<u32>)
+│   │   ├── fn helper_function(val: f32) -> f32
+│   │   ├── struct AnotherStruct
+│   │   └── @compute
+│   │       @workgroup_size(8, 8, 1)
+│   │       fn multi_line_edge_case(
+│   │           @builtin(global_invocation_id)
+│   │           globalId       : vec3<u32>,
+│   │           @group(1)
+│   │           @binding(0)
+│   │           srcTexture     : texture_2d<f32>,
+│   │           @group(1)
+│   │           @binding(1)
+│   │           srcSampler     : sampler,
+│   │           @group(0)
+│   │           @binding(0)
+│   │           uniformsPtr    : ptr<uniform, MyUniforms>,
+│   │           storageBuffer  : ptr<storage, array<vec4<f32>, 64>, read_write>,
+│   │       )
+│   └── 📄 test.metal (272 tokens, 34 lines)
+│       ├── struct MyData
+│       ├── kernel void myKernel(device MyData* data [[buffer(0)]],
+│       │                        uint id [[thread_position_in_grid]])
+│       ├── float myHelperFunction(float x, float y)
+│       ├── vertex float4 vertexShader(const device packed_float3* vertex_array
+│       │   [[buffer(0)]],
+│       │                              unsigned int vid [[vertex_id]])
+│       ├── fragment half4 fragmentShader(float4 P [[position]])
+│       └── float3 computeNormalMap(ColorInOut in, texture2d<float>
+│           normalMapTexture)
+├── 📁 group_lisp (1 folder, 4 files)
+│   ├── 📄 clojure_test.clj (682 tokens, 85 lines)
+│   │   ├── defprotocol P
+│   │   ├── defrecord Person
+│   │   ├── defn -main
+│   │   ├── ns bion.likes_trees
+│   │   ├── def repo-url
+│   │   ├── defn config
+│   │   ├── defmacro with-os
+│   │   └── defrecord SetFullElement
+│   ├── 📄 LispTest.lisp (25 tokens, 6 lines)
+│   │   ├── defstruct person
+│   │   └── defun greet
+│   ├── 📄 racket_struct.rkt (14 tokens, 1 line)
+│   │   └── struct point
+│   └── 📄 test_scheme.scm (360 tokens, 44 lines)
+│       ├── define topological-sort
+│       ├──   define table
+│       ├──   define queue
+│       ├──   define result
+│       ├──   define set-up
+│       └──   define traverse
+└── 📁 group_todo (1 folder, 12 files)
+    ├── 📄 AAPLShaders.metal (5,780 tokens, 566 lines)
+    │   ├── struct LightingParameters
+    │   ├── float Geometry(float Ndotv, float alphaG)
+    │   ├── float3 computeNormalMap(ColorInOut in, texture2d<float>
+    │   │   normalMapTexture)
+    │   ├── float3 computeDiffuse(LightingParameters parameters)
+    │   ├── float Distribution(float NdotH, float roughness)
+    │   ├── float3 computeSpecular(LightingParameters parameters)
+    │   ├── float4 equirectangularSample(float3 direction, sampler s,
+    │   │   texture2d<float> image)
+    │   ├── LightingParameters calculateParameters(ColorInOut in,
+    │   │                                          AAPLCameraData cameraData,
+    │   │                                          constant AAPLLightData&
+    │   │   lightData,
+    │   │                                          texture2d<float>
+    │   │   baseColorMap,
+    │   │                                          texture2d<float>   normalMap,
+    │   │                                          texture2d<float>
+    │   │   metallicMap,
+    │   │                                          texture2d<float>
+    │   │   roughnessMap,
+    │   │                                          texture2d<float>
+    │   │   ambientOcclusionMap,
+    │   │                                          texture2d<float>
+    │   │   skydomeMap)
+    │   ├── struct SkyboxVertex
+    │   ├── struct SkyboxV2F
+    │   ├── vertex SkyboxV2F skyboxVertex(SkyboxVertex in [[stage_in]],
+    │   │                                    constant AAPLCameraData& cameraData
+    │   │   [[buffer(BufferIndexCameraData)]])
+    │   ├── fragment float4 skyboxFragment(SkyboxV2F v [[stage_in]],
+    │   │   texture2d<float> skytexture [[texture(0)]])
+    │   ├── vertex ColorInOut vertexShader(Vertex in [[stage_in]],
+    │   │                                  constant AAPLInstanceTransform&
+    │   │   instanceTransform [[ buffer(BufferIndexInstanceTransforms) ]],
+    │   │                                  constant AAPLCameraData& cameraData
+    │   │   [[ buffer(BufferIndexCameraData) ]])
+    │   ├── float2 calculateScreenCoord( float3 ndcpos )
+    │   ├── fragment float4 fragmentShader(
+    │   │                       ColorInOut                  in
+    │   │   [[stage_in]],
+    │   │                       constant AAPLCameraData&    cameraData
+    │   │   [[ buffer(BufferIndexCameraData) ]],
+    │   │                       constant AAPLLightData&     lightData
+    │   │   [[ buffer(BufferIndexLightData) ]],
+    │   │                       constant AAPLSubmeshKeypath&submeshKeypath
+    │   │   [[ buffer(BufferIndexSubmeshKeypath)]],
+    │   │                       constant Scene*             pScene
+    │   │   [[ buffer(SceneIndex)]],
+    │   │                       texture2d<float>            skydomeMap
+    │   │   [[ texture(AAPLSkyDomeTexture) ]],
+    │   │                       texture2d<float>            rtReflections
+    │   │   [[ texture(AAPLTextureIndexReflections),
+    │   │   function_constant(is_raytracing_enabled)]])
+    │   ├── fragment float4 reflectionShader(ColorInOut in [[stage_in]],
+    │   │                                    texture2d<float> rtReflections
+    │   │   [[texture(AAPLTextureIndexReflections)]])
+    │   ├── struct ThinGBufferOut
+    │   ├── fragment ThinGBufferOut gBufferFragmentShader(ColorInOut in
+    │   │   [[stage_in]])
+    │   ├── kernel void rtReflection(
+    │   │                texture2d< float, access::write >      outImage
+    │   │   [[texture(OutImageIndex)]],
+    │   │                texture2d< float >                     positions
+    │   │   [[texture(ThinGBufferPositionIndex)]],
+    │   │                texture2d< float >                     directions
+    │   │   [[texture(ThinGBufferDirectionIndex)]],
+    │   │                texture2d< float >                     skydomeMap
+    │   │   [[texture(AAPLSkyDomeTexture)]],
+    │   │                constant AAPLInstanceTransform*
+    │   │   instanceTransforms      [[buffer(BufferIndexInstanceTransforms)]],
+    │   │                constant AAPLCameraData&               cameraData
+    │   │   [[buffer(BufferIndexCameraData)]],
+    │   │                constant AAPLLightData&                lightData
+    │   │   [[buffer(BufferIndexLightData)]],
+    │   │                constant Scene*                        pScene
+    │   │   [[buffer(SceneIndex)]],
+    │   │                instance_acceleration_structure
+    │   │   accelerationStructure   [[buffer(AccelerationStructureIndex)]],
+    │   │                uint2 tid [[thread_position_in_grid]])
+    │   ├── else if ( intersection.type == raytracing::intersection_type::none )
+    │   ├── struct VertexInOut
+    │   ├── vertex VertexInOut vertexPassthrough( uint vid [[vertex_id]] )
+    │   ├── fragment float4 fragmentPassthrough( VertexInOut in [[stage_in]],
+    │   │   texture2d< float > tin )
+    │   ├── fragment float4 fragmentBloomThreshold( VertexInOut in [[stage_in]],
+    │   │                                          texture2d< float > tin
+    │   │   [[texture(0)]],
+    │   │                                          constant float* threshold
+    │   │   [[buffer(0)]] )
+    │   └── fragment float4 fragmentPostprocessMerge( VertexInOut in
+    │       [[stage_in]],
+    │                                                constant float& exposure
+    │       [[buffer(0)]],
+    │                                                texture2d< float > texture0
+    │       [[texture(0)]],
+    │                                                texture2d< float > texture1
+    │       [[texture(1)]])
+    ├── 📄 crystal_test.cr (48 tokens, 15 lines)
+    ├── 📄 dart_test.dart (108 tokens, 24 lines)
+    ├── 📄 elixir_test.exs (39 tokens, 10 lines)
+    ├── 📄 forward.frag (739 tokens, 87 lines)
+    ├── 📄 forward.vert (359 tokens, 48 lines)
+    ├── 📄 nodemon.json (118 tokens, 20 lines)
+    ├── 📄 sas_test.sas (97 tokens, 22 lines)
+    ├── 📄 test_setup_py.test (133 tokens, 24 lines)
+    ├── 📄 testTypings.d.ts (158 tokens, 23 lines)
+    ├── 📄 vba_test.bas (67 tokens, 16 lines)
+    └── 📄 wgsl_test.wgsl (94 tokens, 17 lines)
+        ├── @binding(0) @group(0) var<uniform> frame : u32
+        ├── @vertex
+        │   fn vtx_main(@builtin(vertex_index) vertex_index : u32) ->
+        │   @builtin(position) vec4f
+        └── @fragment
+            fn frag_main() -> @location(0) vec4f
diff --git a/tests/golden/legacy/trees/more_languages_group1.txt b/tests/golden/legacy/trees/more_languages_group1.txt
new file mode 100644
index 0000000..b57238e
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group1.txt
@@ -0,0 +1,374 @@
+📁 group1 (1 folder, 11 files)
+├── 📄 addamt.cobol (441 tokens, 40 lines)
+│   ├── IDENTIFICATION DIVISION.
+│   ├── PROGRAM-ID.
+│   │              ADDAMT.
+│   ├── DATA DIVISION.
+│   ├── WORKING-STORAGE SECTION.
+│   ├── 01  KEYED-INPUT.
+│   ├──     05  CUST-NO-IN.
+│   ├──     05  AMT1-IN.
+│   ├──     05  AMT2-IN.
+│   ├──     05  AMT3-IN.
+│   ├── 01  DISPLAYED-OUTPUT.
+│   ├──     05  CUST-NO-OUT.
+│   ├──     05  TOTAL-OUT.
+│   ├── 01  MORE-DATA.
+│   ├── PROCEDURE DIVISION.
+│   └── 100-MAIN.
+├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+│   ├── IDENTIFICATION DIVISION.
+│   ├── PROGRAM-ID. CUSTOMER-INVOICE.
+│   ├── AUTHOR. JANE DOE.
+│   ├── DATE. 2023-12-30.
+│   ├──   DATE-COMPILED. 06/30/10.
+│   ├──     DATE-WRITTEN. 12/34/56.
+│   ├── ENVIRONMENT DIVISION.
+│   ├── INPUT-OUTPUT SECTION.
+│   ├── FILE-CONTROL.
+│   ├──     SELECT CUSTOMER-FILE.
+│   ├──     SELECT INVOICE-FILE.
+│   ├──     SELECT REPORT-FILE.
+│   ├── DATA DIVISION.
+│   ├── FILE SECTION.
+│   ├── FD CUSTOMER-FILE.
+│   ├── 01 CUSTOMER-RECORD.
+│   ├──    05 CUSTOMER-ID.
+│   ├──    05 CUSTOMER-NAME.
+│   ├──    05 CUSTOMER-BALANCE.
+│   ├── FD INVOICE-FILE.
+│   ├── 01 INVOICE-RECORD.
+│   ├──    05 INVOICE-ID.
+│   ├──    05 CUSTOMER-ID.
+│   ├──    05 INVOICE-AMOUNT.
+│   ├── FD REPORT-FILE.
+│   ├── 01 REPORT-RECORD.
+│   ├── WORKING-STORAGE SECTION.
+│   ├── 01 WS-CUSTOMER-FOUND.
+│   ├── 01 WS-END-OF-FILE.
+│   ├── 01 WS-TOTAL-BALANCE.
+│   ├── PROCEDURE DIVISION.
+│   ├── 0000-MAIN-ROUTINE.
+│   ├── 1000-PROCESS-RECORDS.
+│   ├── 1100-UPDATE-CUSTOMER-BALANCE.
+│   └── END PROGRAM CUSTOMER-INVOICE.
+├── 📄 JavaTest.java (578 tokens, 86 lines)
+│   ├── abstract class LivingBeing
+│   ├──     abstract void breathe()
+│   ├── interface Communicator
+│   ├──     String communicate()
+│   ├── @Log
+│   ├── @Getter
+│   ├── @Setter
+│   ├── class Person extends LivingBeing implements Communicator
+│   ├──     Person(String name, int age)
+│   ├──     @Override
+│   ├──     void breathe()
+│   ├──     @Override
+│   ├──     public String communicate()
+│   ├──     void greet()
+│   ├──     String personalizedGreeting(String greeting, Optional<Boolean>
+│   │   includeAge)
+│   ├── @Singleton
+│   ├── @RestController
+│   ├── @SpringBootApplication
+│   ├── public class Example
+│   ├──     @Inject
+│   ├──     public Example(Person person)
+│   ├──     @RequestMapping("/greet")
+│   ├──     String home(@RequestParam(value = "name", defaultValue = "World")
+│   │   String name,
+│   │                   @RequestParam(value = "age", defaultValue = "30") int
+│   │   age)
+│   └──     public static void main(String[] args)
+├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+│   ├── module JuliaTest_EdgeCase
+│   ├── struct Location
+│   │       name::String
+│   │       lat::Float32
+│   │       lon::Float32
+│   │   end
+│   ├── mutable struct mPerson
+│   │       name::String
+│   │       age::Int
+│   │   end
+│   ├── Base.@kwdef mutable struct Param
+│   │       Δt::Float64 = 0.1
+│   │       n::Int64
+│   │       m::Int64
+│   │   end
+│   ├──     sic(x,y)
+│   ├── welcome(l::Location)
+│   ├── ∑(α, Ω)
+│   ├── function noob()
+│   │   end
+│   ├── function ye_olde(hello::String, world::Location)
+│   │   end
+│   ├── function multiline_greet(
+│   │           p::mPerson,
+│   │           greeting::String
+│   │       )
+│   │   end
+│   ├── function julia_is_awesome(prob::DiffEqBase.AbstractDAEProblem{uType,
+│   │   duType, tType,
+│   │           isinplace};
+│   │       kwargs...) where {uType, duType, tType, isinplace}
+│   │   end
+│   └── end
+├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+│   ├── data class Person(val name: String)
+│   ├── fun greet(person: Person)
+│   ├── fun <T> processItems(items: List<T>, processor: (T) -> Unit)
+│   ├── interface Source<out T>
+│   ├──     fun nextT(): T
+│   ├── fun MutableList<Int>.swap(index1: Int, index2: Int)
+│   ├── fun Any?.toString(): String
+│   ├── tailrec fun findFixPoint(x: Double = 1.0): Double
+│   ├── class GenericRepository<T>
+│   ├──     fun getItem(id: Int): T?
+│   ├── sealed interface Error
+│   ├── sealed class IOError(): Error
+│   ├── object Runner
+│   ├──     inline fun <reified S: SomeClass<T>, T> run() : T
+│   ├── infix fun Int.shl(x: Int): Int
+│   ├── class MyStringCollection
+│   ├──     infix fun add(s: String)
+│   ├──     fun build()
+│   ├── open class Base(p: Int)
+│   ├── class Derived(p: Int) : Base(p)
+│   ├── open class Shape
+│   ├──     open fun draw()
+│   ├──     fun fill()
+│   ├──     open fun edge(case: Int)
+│   ├── interface Thingy
+│   ├──     fun edge()
+│   ├── class Circle() : Shape(), Thingy
+│   ├──     override fun draw()
+│   ├──     final override fun edge(case: Int)
+│   ├── interface Base
+│   ├──     fun print()
+│   ├── class BaseImpl(val x: Int) : Base
+│   ├──     override fun print()
+│   ├── internal class Derived(b: Base) : Base by b
+│   ├── class Person constructor(firstName: String)
+│   ├── class People(
+│   │       firstNames: Array<String>,
+│   │       ages: Array<Int>(42),
+│   │   )
+│   ├──     fun edgeCases(): Boolean
+│   ├── class Alien public @Inject constructor(
+│   │       val firstName: String,
+│   │       val lastName: String,
+│   │       var age: Int,
+│   │       val pets: MutableList<Pet> = mutableListOf(),
+│   │   )
+│   ├──     fun objectOriented(): String
+│   ├──  enum class IntArithmetics : BinaryOperator<Int>, IntBinaryOperator
+│   ├──     PLUS {
+│   │           override fun apply(t: Int, u: Int): Int
+│   ├──     TIMES {
+│   │           override fun apply(t: Int, u: Int): Int
+│   ├──     override fun applyAsInt(t: Int, u: Int)
+│   ├── fun reformat(
+│   │       str: String,
+│   │       normalizeCase: Boolean = true,
+│   │       upperCaseFirstLetter: Boolean = true,
+│   │       divideByCamelHumps: Boolean = false,
+│   │       wordSeparator: Char = ' ',
+│   │   )
+│   ├── operator fun Point.unaryMinus()
+│   ├── abstract class Polygon
+│   └──     abstract fun draw()
+├── 📄 lesson.cbl (635 tokens, 78 lines)
+│   ├── IDENTIFICATION DIVISION.
+│   ├── PROGRAM-ID.    CBL0002.
+│   ├── AUTHOR.        Otto B. Fun.
+│   ├── ENVIRONMENT DIVISION.
+│   ├── INPUT-OUTPUT SECTION.
+│   ├── FILE-CONTROL.
+│   ├──     SELECT PRINT-LINE.
+│   ├──     SELECT ACCT-REC.
+│   ├── DATA DIVISION.
+│   ├── FILE SECTION.
+│   ├── FD  PRINT-LINE.
+│   ├── 01  PRINT-REC.
+│   ├──     05  ACCT-NO-O.
+│   ├──     05  ACCT-LIMIT-O.
+│   ├──     05  ACCT-BALANCE-O.
+│   ├──     05  LAST-NAME-O.
+│   ├──     05  FIRST-NAME-O.
+│   ├──     05  COMMENTS-O.
+│   ├── FD  ACCT-REC.
+│   ├── 01  ACCT-FIELDS.
+│   ├──     05  ACCT-NO.
+│   ├──     05  ACCT-LIMIT.
+│   ├──     05  ACCT-BALANCE.
+│   ├──     05  LAST-NAME.
+│   ├──     05  FIRST-NAME.
+│   ├──     05  CLIENT-ADDR.
+│   ├──         10  STREET-ADDR.
+│   ├──         10  CITY-COUNTY.
+│   ├──         10  USA-STATE.
+│   ├──     05  RESERVED.
+│   ├──     05  COMMENTS.
+│   ├── WORKING-STORAGE SECTION.
+│   ├── 01 FLAGS.
+│   ├──   05 LASTREC.
+│   ├── PROCEDURE DIVISION.
+│   ├── OPEN-FILES.
+│   ├── READ-NEXT-RECORD.
+│   ├── CLOSE-STOP.
+│   ├── READ-RECORD.
+│   └── WRITE-RECORD.
+├── 📄 LuaTest.lua (83 tokens, 16 lines)
+│   ├── function HelloWorld.new
+│   ├── function HelloWorld.greet
+│   └── function say_hello
+├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+│   ├── @interface HelloWorld
+│   ├── @interface HelloWorld -> (void) sayHello
+│   ├── @implementation HelloWorld
+│   ├── @implementation HelloWorld -> (void) sayHello
+│   └── void sayHelloWorld()
+├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+│   ├── type color
+│   ├── class hello
+│   ├── class hello -> method say_hello
+│   └── let main ()
+├── 📄 test.js (757 tokens, 154 lines)
+│   ├── class MyClass
+│   ├──   myMethod()
+│   ├──   async asyncMethod(a, b)
+│   ├──   methodWithDefaultParameters(a = 5, b = 10)
+│   ├──   multilineMethod(
+│   │       c,
+│   │       d
+│   │     )
+│   ├──   multilineMethodWithDefaults(
+│   │       t = "tree",
+│   │       p = "plus"
+│   │     )
+│   ├── function myFunction(param1, param2)
+│   ├── function multilineFunction(
+│   │     param1,
+│   │     param2
+│   │   )
+│   ├── const arrowFunction = () =>
+│   ├── const parametricArrow = (a, b) =>
+│   ├── function ()
+│   ├── function outerFunction(outerParam)
+│   ├──   function innerFunction(innerParam)
+│   ├──   innerFunction("inner")
+│   ├── const myObject = {
+│   ├──   myMethod: function (stuff)
+│   ├── let myArrowObject = {
+│   ├──   myArrow: ({
+│   │       a,
+│   │       b,
+│   │       c,
+│   │     }) =>
+│   ├── const myAsyncArrowFunction = async () =>
+│   ├── function functionWithRestParameters(...args)
+│   ├── const namedFunctionExpression = function myNamedFunction()
+│   ├── const multilineArrowFunction = (
+│   │     a,
+│   │     b
+│   │   ) =>
+│   ├── function functionReturningFunction()
+│   ├──   return function ()
+│   ├── function destructuringOnMultipleLines({
+│   │     a,
+│   │     b,
+│   │   })
+│   ├── const arrowFunctionWithDestructuring = ({ a, b }) =>
+│   ├── const multilineDestructuringArrow = ({
+│   │     a,
+│   │     b,
+│   │   }) =>
+│   ├── async function asyncFunctionWithErrorHandling()
+│   ├── class Car
+│   ├──   constructor(brand)
+│   ├──   present()
+│   ├── class Model extends Car
+│   ├──   constructor(brand, mod)
+│   ├──     super(brand)
+│   └──   show()
+└── 📄 test.ts (832 tokens, 165 lines)
+    ├── type MyType
+    ├── interface MyInterface
+    ├── class TsClass
+    ├──   myMethod()
+    ├──   myMethodWithArgs(param1: string, param2: number): void
+    ├──   static myStaticMethod<T>(param: T): T
+    ├──   multilineMethod(
+    │       c: number,
+    │       d: number
+    │     ): number
+    ├──   multilineMethodWithDefaults(
+    │       t: string = "tree",
+    │       p: string = "plus"
+    │     ): string
+    ├── export class AdvancedComponent implements MyInterface
+    ├──   async myAsyncMethod(
+    │       a: string,
+    │       b: number,
+    │       c: string
+    │     ): Promise<void>
+    ├──   genericMethod<T, U>(
+    │       arg1: T,
+    │       arg2: U
+    │     ): [T, U]
+    ├── export class TicketsComponent implements MyInterface
+    ├──   async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String })
+    ├── function tsFunction()
+    ├── function tsFunctionSigned(
+    │     param1: number,
+    │     param2: number
+    │   ): void
+    ├── export default async function tsFunctionComplicated<A, B, C>({
+    │     a = 1 | 2,
+    │     b = "bob",
+    │     c = async () => "charlie",
+    │   }: {
+    │     a: number;
+    │     b: string;
+    │     c: () => Promise<string>;
+    │   }): Promise<string>
+    ├──   return("Standalone function with parameters")
+    ├── const tsArrowFunctionSigned = ({
+    │     a,
+    │     b,
+    │   }: {
+    │     a: number;
+    │     b: string;
+    │   }) =>
+    ├── export const tsComplicatedArrow = async ({
+    │     a = 1 | 2,
+    │     b = "bob",
+    │     c = async () => "charlie",
+    │   }: {
+    │     a: number;
+    │     b: string;
+    │     c: () => Promise<string>;
+    │   }): Promise<string> =>
+    ├── const arrowFunction = () =>
+    ├── const arrow = (a: String, b: Number) =>
+    ├── const asyncArrowFunction = async () =>
+    ├── const asyncArrow = async (a: String, b: Number) =>
+    ├── let weirdArrow = () =>
+    ├── const asyncPromiseArrow = async (): Promise<void> =>
+    ├── let myWeirdArrowSigned = (x: number): number =>
+    ├── class Person
+    ├──   constructor(private firstName: string, private lastName: string)
+    ├──   getFullName(): string
+    ├──   describe(): string
+    ├── class Employee extends Person
+    ├──   constructor(
+    │       firstName: string,
+    │       lastName: string,
+    │       private jobTitle: string
+    │     )
+    ├──     super(firstName, lastName)
+    ├──   describe(): string
+    ├── interface Shape
+    └── interface Square extends Shape
diff --git a/tests/golden/legacy/trees/more_languages_group2.txt b/tests/golden/legacy/trees/more_languages_group2.txt
new file mode 100644
index 0000000..0eba148
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group2.txt
@@ -0,0 +1,135 @@
+📁 group2 (1 folder, 8 files)
+├── 📄 apl_test.apl (28 tokens, 5 lines)
+│   ├── :Namespace HelloWorld
+│   ├── :Namespace HelloWorld -> hello ← 'Hello, World!'
+│   └── :Namespace HelloWorld -> plus ← {⍺+⍵}
+├── 📄 c_test.c (837 tokens, 142 lines)
+│   ├── struct Point
+│   ├──     int x;
+│   ├──     int y;
+│   ├── struct Point getOrigin()
+│   ├── float mul_two_floats(float x1, float x2)
+│   ├── enum days
+│   ├──     SUN,
+│   ├──     MON,
+│   ├──     TUE,
+│   ├──     WED,
+│   ├──     THU,
+│   ├──     FRI,
+│   ├──     SAT
+│   ├── long add_two_longs(long x1, long x2)
+│   ├── double multiplyByTwo(double num)
+│   ├── char getFirstCharacter(char *str)
+│   ├── void greet(Person p)
+│   ├── typedef struct
+│   ├──     char name[50];
+│   ├── } Person;
+│   ├── int main()
+│   ├── int* getArrayStart(int arr[], int size)
+│   ├── long complexFunctionWithMultipleArguments(
+│   │       int param1,
+│   │       double param2,
+│   │       char *param3,
+│   │       struct Point point
+│   │   )
+│   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   *channel, int channellen, int is_pattern)
+│   ├──     while((ln = listNext(&li)))
+│   ├── static struct config
+│   ├──     aeEventLoop *el;
+│   ├──     cliConnInfo conn_info;
+│   ├──     const char *hostsocket;
+│   ├──     int tls;
+│   ├──     struct cliSSLconfig sslconfig;
+│   └── } config;
+├── 📄 go_test.go (179 tokens, 46 lines)
+│   ├── type Greeting struct
+│   ├── func (g Greeting) sayHello()
+│   ├── func createGreeting(m string) Greeting
+│   ├── type SomethingLong struct
+│   ├── func (s *SomethingLong) WithAReasonableName(
+│   │           ctx context.Context,
+│   │           param1 string,
+│   │           param2 int,
+│   │           param3 map[string]interface{},
+│   │           callback func(int) error,
+│   │   ) (resultType, error)
+│   ├── type resultType struct
+│   └── func main()
+├── 📄 PerlTest.pl (63 tokens, 20 lines)
+│   ├── package PerlTest
+│   ├── package PerlTest -> sub new
+│   ├── package PerlTest -> sub hello
+│   └── package PerlTest -> sub say_hello
+├── 📄 PhpTest.php (70 tokens, 19 lines)
+│   ├── class HelloWorld
+│   ├── class HelloWorld -> function sayHello
+│   ├── function greet
+│   ├── class Person
+│   └── class Person -> function __construct
+├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)
+│   ├── function Say-Nothing()
+│   ├── class Person
+│   ├──     Person([string]$name)
+│   ├──     [string]Greet()
+│   ├──     [string]GreetMany([int]$times)
+│   ├──     [string]GreetWithDetails([string]$greeting, [int]$times)
+│   ├──     [string]GreetMultiline(
+│   │           [string]$greeting,
+│   │           [int]$times
+│   │       )
+│   ├──     NoReturn([int]$times)
+│   ├──     NoReturnNoArgs()
+│   ├── function Say-Hello([Person]$person)
+│   ├── function Multi-Hello([Person]$personA, [Person]$personB)
+│   ├── function Switch-Item
+│   ├──   param ([switch]$on)
+│   ├── function Get-SmallFiles
+│   ├──   param (
+│   │         [PSDefaultValue(Help = '100')]
+│   │         $Size = 100)
+│   ├── function Get-User
+│   ├──   [CmdletBinding(DefaultParameterSetName="ID")]
+│   ├──   [OutputType("System.Int32", ParameterSetName="ID")]
+│   ├──   [OutputType([String], ParameterSetName="Name")]
+│   ├──   Param (
+│   │       [parameter(Mandatory=$true, ParameterSetName="ID")]
+│   │       [Int[]]
+│   │       $UserID,
+│   │       [parameter(Mandatory=$true, ParameterSetName="Name")]
+│   │       [String[]]
+│   │       $UserName)
+│   ├── filter Get-ErrorLog ([switch]$Message)
+│   └── function global:MultilineSignature(
+│         [string]$param1,
+│         [int]$param2,
+│         [Parameter(Mandatory=$true)]
+│         [string]$param3
+│       )
+├── 📄 ScalaTest.scala (171 tokens, 40 lines)
+│   ├── def sumOfSquares(x: Int, y: Int): Int
+│   ├── trait Bark
+│   ├──   def bark: String
+│   ├── case class Person(name: String)
+│   ├── class GenericClass[T](
+│   │       val data: T,
+│   │       val count: Int
+│   │   )
+│   ├──   def getData: T
+│   ├── object HelloWorld
+│   ├──   def greet(person: Person): Unit
+│   ├──   def main(args: Array[String]): Unit
+│   ├── def complexFunction(
+│   │       a: Int,
+│   │       b: String,
+│   │       c: Float
+│   │   ): (Int, String) Option
+│   └── def sumOfSquaresShort(x: Int, y: Int): Int
+└── 📄 test.csv (0 tokens, 0 lines)
+    ├── Name
+    ├── Age
+    ├── Country
+    ├── City
+    └── Email
diff --git a/tests/golden/legacy/trees/more_languages_group3.txt b/tests/golden/legacy/trees/more_languages_group3.txt
new file mode 100644
index 0000000..efc9ce0
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group3.txt
@@ -0,0 +1,346 @@
+📁 group3 (1 folder, 16 files)
+├── 📄 bash_test.sh (127 tokens, 22 lines)
+│   ├── echo_hello_world()
+│   ├── function fun_echo_hello_world()
+│   ├── export SECRET
+│   ├── alias md='make debug'
+│   ├── add_alias()
+│   └── create_conda_env()
+├── 📄 cpp_test.cpp (1,670 tokens, 259 lines)
+│   ├── class Person
+│   ├──     std::string name;
+│   ├── public:
+│   ├──     Person(std::string n) : name(n)
+│   ├──     void greet()
+│   ├── void globalGreet()
+│   ├── int main()
+│   ├── void printMessage(const std::string &message)
+│   ├── template<typename T>
+│   │   void printVector(const std::vector<T>& vec)
+│   ├── struct Point
+│   ├──     int x, y;
+│   ├──     Point(int x, int y) : x(x), y(y)
+│   ├── class Animal
+│   ├── public:
+│   ├──     Animal(const std::string &name) : name(name)
+│   ├──     virtual void speak() const
+│   ├──     virtual ~Animal()
+│   ├── protected:
+│   ├──     std::string name;
+│   ├── class Dog : public Animal
+│   ├── public:
+│   ├──     Dog(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── class Cat : public Animal
+│   ├── public:
+│   ├──     Cat(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   num_layers,
+│   │                                int batch_size, int max_seq_length, float
+│   │   dropout,
+│   │                                bool bidirectional, bool cudnn_allow_tf32,
+│   │                                int workspace_size, int reserve_space_size)
+│   ├── int main()
+│   ├── enum ECarTypes
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV,
+│   ├──   Wagon
+│   ├── ECarTypes GetPreferredCarType()
+│   ├── enum ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── enum class ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── void myFunction(string fname, int age)
+│   ├── template <typename T> T cos(T)
+│   ├── template <typename T> T sin(T)
+│   ├── template <typename T> T sqrt(T)
+│   ├── template<typename T> struct VLEN
+│   ├── template<typename T> class arr
+│   ├──   private:
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──   public:
+│   ├──     arr() : p(0), sz(0)
+│   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   ├──     arr(arr &&other)
+│   │         : p(other.p), sz(other.sz)
+│   ├──     ~arr()
+│   ├──     void resize(size_t n)
+│   ├──     T &operator[](size_t idx)
+│   ├──     T *data()
+│   ├──     size_t size() const
+│   ├── class Buffer
+│   ├──  private:
+│   ├──   void* ptr_;
+│   └── std::tuple<array, array, array> quantize(
+│           const array& w,
+│           int group_size,
+│           int bits,
+│           StreamOrDevice s)
+├── 📄 csharp_test.cs (957 tokens, 146 lines)
+│   ├── public interface IExcelTemplate
+│   ├──     void LoadTemplate(string templateFilePath)
+│   ├──     void LoadData(Dictionary<string, string> data)
+│   ├──     void ModifyCell(string cellName, string value)
+│   ├──     void SaveToFile(string filePath)
+│   ├── public interface IGreet
+│   ├──     void Greet()
+│   ├── public enum WeekDays
+│   ├── public delegate void DisplayMessage(string message)
+│   ├── public struct Address
+│   ├── public static class HelperFunctions
+│   ├──     public static void PrintMessage(string message)
+│   ├──     public static int AddNumbers(int a, int b)
+│   ├── namespace HelloWorldApp
+│   ├──     class Person : IGreet
+│   ├──         public Person(string name, int age)
+│   ├──         public void Greet()
+│   ├──     class HelloWorld
+│   ├──         static void Main(string[] args)
+│   ├── namespace TemplateToExcelServer.Template
+│   ├──     public interface ITemplateObject
+│   ├──         string[,] GetContent()
+│   ├──         string[] GetContentArray()
+│   ├──         string[] GetFormat()
+│   ├──         int? GetFormatLength()
+│   ├──         TemplateObject SetContent(string[,] Content)
+│   ├──         TemplateObject SetContentArray(string[] value)
+│   ├──         TemplateObject SetFormat(string[] Header)
+│   ├──         TemplateObject SetNameOfReport(
+│   │               ReadOnlyMemory<byte> ReportName,
+│   │               int[] EdgeCase)
+│   ├──         TemplateObject SetSheetName(ReadOnlyMemory<byte> SheetName)
+│   ├── public class BankAccount(string accountID, string owner)
+│   ├──     public override string ToString() =>
+│   ├── var IncrementBy = (int source, int increment = 1) =>
+│   ├── Func<int, int, int> add = (x, y) =>
+│   ├── button.Click += (sender, args) =>
+│   ├── public Func<int, int> GetMultiplier(int factor)
+│   ├── public void Method(
+│   │           int param1,
+│   │           int param2,
+│   │           int param3,
+│   │           int param4,
+│   │           int param5,
+│   │           int param6,
+│   │       )
+│   ├── System.Net.ServicePointManager.ServerCertificateValidationCallback +=
+│   │       (se, cert, chain, sslerror) =>
+│   ├── class ServerCertificateValidation
+│   ├──     public bool OnRemoteCertificateValidation(
+│   │           object se,
+│   │           X509Certificate cert,
+│   │           X509Chain chain,
+│   │           SslPolicyErrors sslerror
+│   │       )
+│   ├── s_downloadButton.Clicked += async (o, e) =>
+│   ├── [HttpGet, Route("DotNetCount")]
+│   └── static public async Task<int> GetDotNetCount(string URL)
+├── 📄 hallucination.tex (1,633 tokens, 126 lines)
+│   ├── Harnessing the Master Algorithm: Strategies for AI LLMs to Mitigate
+│   │   Hallucinations
+│   ├── Hallucinated Pedro Domingos et al.
+│   ├── Christmas Eve 2023
+│   ├── 1 Introduction
+│   ├── 2 Representation in LLMs
+│   ├──   2.1 Current Representational Models
+│   ├──   2.2 Incorporating Cognitive Structures
+│   ├──   2.3 Conceptual Diagrams of Advanced Representational Models
+│   ├── 3 Evaluation Strategies
+│   ├──   3.1 Existing Evaluation Metrics for LLMs
+│   ├──   3.2 Integrating Contextual and Ethical Considerations
+│   ├──   3.3 Case Studies: Evaluation in Practice
+│   ├── 4 Optimization Techniques
+│   ├──   4.1 Continuous Learning Models
+│   ├──   4.2 Adaptive Algorithms for Real-time Adjustments
+│   ├──   4.3 Performance Metrics Pre- and Post-Optimization
+│   ├── 5 Interdisciplinary Insights
+│   ├──   5.1 Cognitive Science and AI: A Symbiotic Relationship
+│   ├──   5.2 Learning from Human Cognitive Processes
+│   ├── 6 Challenges and Future Directions
+│   ├──   6.1 Addressing Current Limitations
+│   ├──   6.2 The Road Ahead: Ethical and Practical Considerations
+│   ├── 7 Conclusion
+│   ├──   7.1 Summarizing Key Findings
+│   └──   7.2 The Next Steps in AI Development
+├── 📄 ruby_test.rb (138 tokens, 37 lines)
+│   ├── module Greeter
+│   ├──   def self.say_hello
+│   ├── class HelloWorld
+│   ├──   def say_hello
+│   ├── class Human
+│   ├──   def self.bar
+│   ├──   def self.bar=(value)
+│   ├── class Doctor < Human
+│   └──   def brachial_plexus(
+│             roots,
+│             trunks,
+│             divisions: true,
+│             cords: [],
+│             branches: Time.now
+│           )
+├── 📄 swift_test.swift (469 tokens, 110 lines)
+│   ├── class Person
+│   ├──     init(name: String)
+│   ├──     func greet()
+│   ├──     func yEdgeCase(
+│   │           fname: String,
+│   │           lname: String,
+│   │           age: Int,
+│   │           address: String,
+│   │           phoneNumber: String
+│   │       )
+│   ├── func globalGreet()
+│   ├── struct Point
+│   ├── protocol Animal
+│   ├──     func speak()
+│   ├── struct Dog: Animal
+│   ├── class Cat: Animal
+│   ├──     init(name: String)
+│   ├──     func speak()
+│   ├── enum CarType
+│   ├── func getPreferredCarType() -> CarType
+│   ├── enum CarType: UInt8
+│   ├── enum class CarType: UInt8
+│   ├── func myFunction(fname: String, age: Int)
+│   └── func myFunctionWithMultipleParameters(
+│           fname: String,
+│           lname: String,
+│           age: Int,
+│           address: String,
+│           phoneNumber: String
+│       )
+├── 📄 test.lean (289 tokens, 42 lines)
+│   ├── # Advanced Topics in Group Theory
+│   ├── section GroupDynamics
+│   ├── lemma group_stability (G : Type*) [Group G] (H : Subgroup G)
+│   ├── theorem subgroup_closure {G : Type*} [Group G] (S : Set G)
+│   ├── axiom group_homomorphism_preservation {G H : Type*} [Group G] [Group H]
+│   │   (f : G → H)
+│   ├── end GroupDynamics
+│   ├── section ConstructiveApproach
+│   ├── lemma finite_group_order (G : Type*) [Group G] [Fintype G]
+│   ├── lemma complex_lemma {X Y : Type*} [SomeClass X] [AnotherClass Y]
+│   │     (f : X → Y) (g : Y → X)
+│   └── end ConstructiveApproach
+├── 📄 test.capnp (117 tokens, 30 lines)
+│   ├── struct Employee
+│   ├──   id @0 :Int32
+│   ├──   name @1 :Text
+│   ├──   role @2 :Text
+│   ├──   skills @3 :List(Skill)
+│   ├──   struct Skill
+│   ├──     name @0 :Text
+│   ├──     level @1 :Level
+│   ├──     enum Level
+│   ├──       beginner @0
+│   ├──       intermediate @1
+│   ├──       expert @2
+│   ├──   status :union
+│   ├──     active @4 :Void
+│   ├──     onLeave @5 :Void
+│   ├──     retired @6 :Void
+│   ├── struct Company
+│   └──   employees @0 :List(Employee)
+├── 📄 test.graphql (66 tokens, 21 lines)
+│   ├── type Query
+│   ├──     getBooks: [Book]
+│   ├──     getAuthors: [Author]
+│   ├── type Mutation
+│   ├──     addBook(title: String, author: String): Book
+│   ├──     removeBook(id: ID): Book
+│   ├── type Book
+│   ├──     id: ID
+│   ├──     title: String
+│   ├──     author: Author
+│   ├── type Author
+│   ├──     id: ID
+│   ├──     name: String
+│   └──     books: [Book]
+├── 📄 test.proto (142 tokens, 34 lines)
+│   ├── syntax = "proto3"
+│   ├── service EmployeeService
+│   ├──     rpc GetEmployee(EmployeeId) returns (EmployeeInfo)
+│   ├──     rpc AddEmployee(EmployeeData) returns (EmployeeInfo)
+│   ├──     rpc UpdateEmployee(EmployeeUpdate) returns (EmployeeInfo)
+│   ├── message EmployeeId
+│   ├──     int32 id = 1
+│   ├── message EmployeeInfo
+│   ├──     int32 id = 1
+│   ├──     string name = 2
+│   ├──     string role = 3
+│   ├── message EmployeeData
+│   ├──     string name = 1
+│   ├──     string role = 2
+│   ├── message EmployeeUpdate
+│   ├──     int32 id = 1
+│   ├──     string name = 2
+│   └──     string role = 3
+├── 📄 test.sqlite (0 tokens, 0 lines)
+│   ├── students table:
+│   ├──    id integer primary key
+│   ├──    name text not null
+│   ├──    age integer not null
+│   ├── courses table:
+│   ├──    id integer primary key
+│   ├──    title text not null
+│   └──    credits integer not null
+├── 📄 test_Cargo.toml (119 tokens, 18 lines)
+│   ├── name: test_cargo
+│   ├── version: 0.1.0
+│   ├── description: A test Cargo.toml
+│   ├── license: MIT OR Apache-2.0
+│   ├── dependencies:
+│   ├──   clap 4.4
+│   └──   sqlx 0.7 (features: runtime-tokio, tls-rustls)
+├── 📄 test_json_rpc_2_0.json (26 tokens, 6 lines)
+│   ├── jsonrpc: 2.0
+│   ├── method: subtract
+│   ├── params:
+│   ├──     minuend: 42
+│   ├──     subtrahend: 23
+│   └── id: 1
+├── 📄 test_openapi.yaml (753 tokens, 92 lines)
+│   ├── openapi: 3.0.1
+│   ├──     title: TODO Plugin
+│   ├──     description: A plugin to create and manage TODO lists using ChatGPT.
+│   ├──     version: v1
+│   ├── servers:
+│   ├──     - url: PLUGIN_HOSTNAME
+│   ├── paths:
+│   ├──     '/todos/{username}':
+│   ├──         GET (getTodos): Get the list of todos
+│   ├──         POST (addTodo): Add a todo to the list
+│   └──         DELETE (deleteTodo): Delete a todo from the list
+├── 📄 test_openrpc.json (225 tokens, 44 lines)
+│   ├── openrpc: 1.2.1
+│   ├── info:
+│   ├──     title: Demo Petstore
+│   ├──     version: 1.0.0
+│   ├── methods:
+│   ├──     listPets: List all pets
+│   ├──         params:
+│   ├──             - limit: integer
+│   └──         result: pets = An array of pets
+└── 📄 test_pyproject.toml (304 tokens, 39 lines)
+    ├── name: tree_plus
+    ├── version: 1.0.8
+    ├── description: A `tree` util enhanced with tokens, lines, and components.
+    ├── License :: OSI Approved :: Apache Software License
+    ├── License :: OSI Approved :: MIT License
+    ├── dependencies:
+    ├──     tiktoken
+    ├──     PyYAML
+    ├──     click
+    ├──     rich
+    └──     tomli
diff --git a/tests/golden/legacy/trees/more_languages_group4.txt b/tests/golden/legacy/trees/more_languages_group4.txt
new file mode 100644
index 0000000..3b55b2c
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group4.txt
@@ -0,0 +1,214 @@
+📁 group4 (1 folder, 10 files)
+├── 📄 erl_test.erl (480 tokens, 68 lines)
+│   ├── -module(erl_test).
+│   ├── -record(person).
+│   ├── -type ra_peer_status().
+│   ├── -type ra_membership().
+│   ├── -opaque my_opaq_type().
+│   ├── -type orddict(Key, Val).
+│   ├── -type edge(
+│   │           Cases,
+│   │           Pwn,
+│   │       ).
+│   ├── -spec guarded(X) -> X when X :: tuple().
+│   ├── -spec edge_case(
+│   │           {integer(), any()} | [any()]
+│   │       ) -> processed, integer(), any()} | [{item, any()}].
+│   ├── -spec complex_function({integer(), any()} | [any()]) ->
+│   │       {processed, integer(), any()} | [{item, any()}].
+│   ├── -spec list_manipulation([integer()]) -> [integer()].
+│   ├── -spec overload(T1, T2) -> T3
+│   │           ; (T4, T5) -> T6.
+│   ├── -spec multiguard({X, integer()}) -> X when X :: atom()
+│   │           ; ([Y]) -> Y when Y :: number().
+│   ├── -record(multiline).
+│   └── -record(maybe_undefined).
+├── 📄 haskell_test.hs (414 tokens, 41 lines)
+│   ├── data Person
+│   ├── greet :: Person -> String
+│   └── resolveVariables ::
+│         forall m fragments.
+│         (MonadError QErr m, Traversable fragments) =>
+│         Options.BackwardsCompatibleNullInNonNullableVariables ->
+│         [G.VariableDefinition] ->
+│         GH.VariableValues ->
+│         [G.Directive G.Name] ->
+│         G.SelectionSet fragments G.Name ->
+│         m
+│           ( [G.Directive Variable],
+│             G.SelectionSet fragments Variable
+│           )
+├── 📄 mathematica_test.nb (133 tokens, 21 lines)
+│   ├── person[name_]
+│   ├── sayHello[]
+│   └── sumList[list_List]
+├── 📄 matlab_test.m (48 tokens, 12 lines)
+│   ├── classdef HelloWorld -> function greet
+│   └── function loneFun
+├── 📄 RTest.R (367 tokens, 46 lines)
+│   ├── class(person)
+│   ├── greet.Person <- function
+│   ├── ensure_between = function
+│   └── run_intermediate_annealing_process = function
+├── 📄 rust_test.rs (1,368 tokens, 259 lines)
+│   ├── fn at_beginning<'a>(&'a str)
+│   ├── pub enum Days<E: EdgeCase> {
+│   │       #\[default]
+│   │       Sun,
+│   │       Mon,
+│   │       #\[error("edge case {idx}, expected at least {} and at most {}",
+│   │   .limits.lo, .limits.hi)]
+│   │       Tue,
+│   │       Wed,
+│   │       Thu(i16, bool),
+│   │       Fri { day: u8 },
+│   │       Sat {
+│   │           urday: String,
+│   │           edge_case: E,
+│   │       },
+│   │   }
+│   ├── struct Point
+│   ├── impl Point
+│   ├──     fn get_origin() -> Point
+│   ├── struct Person
+│   ├── impl Person
+│   ├──     fn greet(&self)
+│   ├── fn add_two_longs(x1: i64, x2: i64) -> i64
+│   ├── fn add_two_longs_longer(
+│   │       x1: i64,
+│   │       x2: i64,
+│   │   ) -> i64
+│   ├── const fn multiply_by_two(num: f64) -> f64
+│   ├── fn get_first_character(s: &str) -> Option<char>
+│   ├── trait Drawable
+│   ├──     fn draw(&self)
+│   ├── impl Drawable for Point
+│   ├──     fn draw(&self)
+│   ├── fn with_generic<D: Drawable>(d: D)
+│   ├── fn with_generic<D>(d: D)
+│   │   where
+│   │       D: Drawable
+│   ├── fn main()
+│   ├── pub struct VisibleStruct
+│   ├── mod my_module
+│   ├──     pub struct AlsoVisibleStruct<T>(T, T)
+│   ├── macro_rules! say_hello
+│   ├── #[macro_export]
+│   │   macro_rules! hello_tree_plus
+│   ├── pub mod lib
+│   ├──     pub mod interfaces
+│   ├──     mod engine
+│   ├── pub fn flow<S1, S2, S3, S4, E, T, L>(
+│   │       source: S1,
+│   │       extractor: E,
+│   │       inbox: S2,
+│   │       transformer: T,
+│   │       outbox: S3,
+│   │       loader: L,
+│   │       sink: &mut S4,
+│   │   ) -> Result<(), Box<dyn Error>>
+│   │   where
+│   │       S1: Extractable,
+│   │       S2: Extractable + Loadable,
+│   │       S3: Extractable + Loadable,
+│   │       S4: Loadable,
+│   │       E: Extractor<S1, S2>,
+│   │       T: Transformer<S2, S3>,
+│   │       L: Loader<S3, S4>
+│   ├── trait Container
+│   ├──     fn items(&self) -> impl Iterator<Item = Widget>
+│   ├── trait HttpService
+│   ├──     async fn fetch(&self, url: Url) -> HtmlBody
+│   ├── struct Pair<T, U>
+│   ├── trait Transformer<T>
+│   ├──     fn transform(&self, input: T) -> T
+│   ├── impl<T: std::ops::Add<Output = T> + Copy> Transformer<T> for Pair<T, T>
+│   ├──     fn transform(&self, input: T) -> T
+│   ├── fn main()
+│   ├── async fn handle_get(State(pool): State<PgPool>) -> Result<Html<String>,
+│   │   (StatusCode, String)>
+│   │   where
+│   │       Bion: Cool
+│   ├── #[macro_export]
+│   │   macro_rules! unit
+│   ├──             fn insert(
+│   │                   &mut self,
+│   │                   key: (),
+│   │                   value: $unit_dtype,
+│   │               ) -> Result<Option<$unit_dtype>, ETLError>
+│   ├── pub async fn handle_get_axum_route(
+│   │       Session { maybe_claims }: Session,
+│   │       Path(RouteParams {
+│   │           alpha,
+│   │           bravo,
+│   │           charlie,
+│   │           edge_case
+│   │       }): Path<RouteParams>,
+│   │   ) -> ServerResult<Response>
+│   ├── fn encode_pipeline(cmds: &[Cmd], atomic: bool) -> Vec<u8>
+│   ├── pub async fn handle_post_yeet(
+│   │       State(auth_backend): State<AuthBackend>,
+│   │       Session { maybe_claims }: Session,
+│   │       Form(yeet_form): Form<YeetForm>,
+│   │   ) -> Result<Response, AuthError>
+│   └── pub async fn handle_get_thingy(
+│           session: Session,
+│           State(ApiBackend {
+│               page_cache,
+│               auth_backend,
+│               library_sql,
+│               some_data_cache,
+│               metadata_cache,
+│               thingy_client,
+│               ..
+│           }): State<ApiBackend>,
+│       ) -> ServerResult<Response>
+├── 📄 test.zig (397 tokens, 60 lines)
+│   ├── pub fn add(a: i32, b: i32) i32
+│   ├── test "add function"
+│   ├── const BunBuildOptions = struct
+│   ├──     pub fn updateRuntime(this: *BunBuildOptions) anyerror!void
+│   ├──     pub fn step(this: BunBuildOptions, b: anytype)
+│   │   *std.build.OptionsStep
+│   └── pub fn sgemv(
+│           order: Order,
+│           trans: Trans,
+│           m: usize,
+│           n: usize,
+│           alpha: f32,
+│           a: []const f32,
+│           lda: usize,
+│           x: []const f32,
+│           x_add: usize,
+│           beta: f32,
+│           y: []f32,
+│           y_add: usize,
+│       ) void
+├── 📄 test_fsharp.fs (92 tokens, 27 lines)
+│   ├── module TestFSharp
+│   ├── type Person = {
+│   ├── let add x y =
+│   ├── let multiply
+│   │       (x: int)
+│   │       (y: int): int =
+│   ├── let complexFunction
+│   │       (a: int)
+│   │       (b: string)
+│   │       (c: float)
+│   │       : (int * string) option =
+│   └── type Result<'T> =
+├── 📄 test_tcl_tk.tcl (54 tokens, 16 lines)
+│   ├── proc sayHello {}
+│   ├── proc arrg { input }
+│   └── proc multiLine {
+│           x,
+│           y
+│       }
+└── 📄 tf_test.tf (202 tokens, 38 lines)
+    ├── provider "aws"
+    ├── resource "aws_instance" "example"
+    ├── data "aws_ami" "ubuntu"
+    ├── variable "instance_type"
+    ├── output "instance_public_ip"
+    ├── locals
+    └── module "vpc"
diff --git a/tests/golden/legacy/trees/more_languages_group5.txt b/tests/golden/legacy/trees/more_languages_group5.txt
new file mode 100644
index 0000000..9acc6f2
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group5.txt
@@ -0,0 +1,257 @@
+📁 group5 (1 folder, 19 files)
+├── 📄 ansible_test.yml (55 tokens, 14 lines)
+│   ├── Install package
+│   ├── Start service
+│   └── Create user
+├── 📄 app-routing.module.ts (287 tokens, 28 lines)
+│   ├── const routes: Routes = [
+│   │       { path: '', redirectTo: 'login', pathMatch: 'full' },
+│   │       { path: '*', redirectTo: 'login' },
+│   │       { path: 'home', component: HomeComponent },
+│   │       { path: 'login', component: LoginComponent },
+│   │       { path: 'register', component: RegisterComponent },
+│   │       { path: 'events', component: EventsComponent },
+│   │       { path: 'invites', component: InvitesComponent },
+│   │       { path: 'rewards', component: RewardsComponent },
+│   │       { path: 'profile', component: ProfileComponent },
+│   │   ];
+│   └── export class AppRoutingModule
+├── 📄 app.component.spec.ts (410 tokens, 47 lines)
+│   ├── describe 'AppComponent'
+│   ├──     it should create the app
+│   ├──     it should welcome the user
+│   ├──     it should welcome 'Jimbo'
+│   └──     it should request login if not logged in
+├── 📄 app.component.ts (271 tokens, 45 lines)
+│   ├── export class AppComponent
+│   ├──   constructor(
+│   │       private http: HttpClient,
+│   │       private loginService: LoginService,
+│   │       private stripeService: StripeService
+│   │     )
+│   ├──   constructor(private loginService: LoginService)
+│   ├──   checkSession()
+│   ├──   async goToEvent(event_id: string)
+│   └──   valInvitedBy(event: any, event_id: string)
+├── 📄 app.module.ts (374 tokens, 43 lines)
+│   ├── @NgModule({
+│   │       declarations: [
+│   │           AppComponent,
+│   │           HomeComponent,
+│   │           LoginComponent,
+│   │           RegisterComponent,
+│   │           EventsComponent,
+│   │           InvitesComponent,
+│   │           RewardsComponent,
+│   │           ProfileComponent
+│   └── export class AppModule
+├── 📄 checkbox_test.md (191 tokens, 29 lines)
+│   ├── # My Checkbox Test
+│   ├── ## My No Parens Test
+│   ├── ## My Empty href Test
+│   ├── ## My other url Test [Q&A]
+│   ├── ## My other other url Test [Q&A]
+│   ├── ## My 2nd other url Test [Q&A]
+│   ├── ## My 3rd other url Test [Q&A]
+│   ├── - [ ] Task 1
+│   ├──     - [ ] No Space Task 1.1
+│   ├──     - [ ] Two Spaces Task 1.2
+│   ├──         - [ ] Subtask 1.2.1
+│   ├── - [ ] Task 2
+│   ├── - [x] Task 3
+│   ├──     - [ ] Subtask 3.1
+│   ├── - [x] Task 6
+│   ├──     - [x] Subtask 6.1
+│   ├──         - [ ] Handle edge cases
+│   └── # My Codeblock Test
+├── 📄 checkbox_test.txt (257 tokens, 33 lines)
+│   ├── - [ ] fix phone number format +1
+│   ├── - [ ] add forgot password
+│   ├── - [ ] ? add email verification
+│   ├── - [ ] store token the right way
+│   ├── - [ ] test nesting of checkboxes
+│   ├── - [ ] user can use option to buy ticket at 2-referred price
+│   ├── - [ ] CTA refer 2 people to get instant lower price
+│   └── - [ ] form to send referrals
+├── 📄 environment.test.ts (197 tokens, 19 lines)
+│   ├── environment:
+│   ├──    production
+│   ├──    cognitoUserPoolId
+│   ├──    cognitoAppClientId
+│   └──    apiurl
+├── 📄 hello_world.pyi (22 tokens, 3 lines)
+│   ├── @final
+│   │   class dtype(Generic[_DTypeScalar_co])
+│   └──     names: None | tuple[builtins.str, ...]
+├── 📄 k8s_test.yaml (140 tokens, 37 lines)
+│   ├── apps/v1.Deployment -> my-app
+│   ├── v1.Service -> my-service
+│   └── v1.ConfigMap -> my-config
+├── 📄 Makefile (714 tokens, 84 lines)
+│   ├── include dotenv/dev.env
+│   ├── .PHONY: dev
+│   ├── dev
+│   ├── services-down
+│   ├── services-stop: services-down
+│   ├── define CHECK_POSTGRES
+│   ├── damage-report
+│   ├── tail-logs
+│   └── cloud
+├── 📄 requirements_test.txt (29 tokens, 10 lines)
+│   ├── psycopg2-binary
+│   ├── pytest
+│   ├── coverage
+│   ├── flask[async]
+│   ├── flask_cors
+│   ├── stripe
+│   ├── pyjwt[crypto]
+│   ├── cognitojwt[async]
+│   └── flask-lambda
+├── 📄 rust_todo_test.rs (92 tokens, 26 lines)
+│   ├── TODO: This todo tests parse_todo
+│   ├── enum Color {
+│   │       Red,
+│   │       Blue,
+│   │       Green,
+│   │   }
+│   ├── struct Point
+│   ├── trait Drawable
+│   ├──     fn draw(&self)
+│   ├── impl Drawable for Point
+│   ├──     fn draw(&self)
+│   └── fn main()
+├── 📄 sql_test.sql (270 tokens, 51 lines)
+│   ├── CREATE TABLE promoters
+│   ├──    user_id serial PRIMARY KEY,
+│   ├──    type varchar(20) NOT NULL,
+│   ├──    username varchar(20) NOT NULL,
+│   ├──    password varchar(20) NOT NULL,
+│   ├──    email varchar(30) NOT NULL,
+│   ├──    phone varchar(20) NOT NULL,
+│   ├──    promocode varchar(20),
+│   ├──    info json,
+│   ├──    going text[],
+│   ├──    invites text[],
+│   ├──    balance integer NOT NULL,
+│   ├──    rewards text[],
+│   ├──    created timestamp
+│   ├── CREATE TABLE events
+│   ├──    event_id serial PRIMARY KEY,
+│   ├──    name varchar(64) NOT NULL,
+│   ├──    date varchar(64) NOT NULL,
+│   ├──    location varchar(64) NOT NULL,
+│   ├──    performer varchar(64) NOT NULL,
+│   ├──    rewards json,
+│   └──    created timestamp
+├── 📄 standard-app-routing.module.ts (100 tokens, 16 lines)
+│   └── const routes: Routes = [
+│         { path: '', component: HomeComponent },
+│         {
+│           path: 'heroes',
+│           component: HeroesListComponent,
+│           children: [
+│             { path: ':id', component: HeroDetailComponent },
+│             { path: 'new', component: HeroFormComponent },
+│           ],
+│         },
+│         { path: '**', component: PageNotFoundComponent },
+│       ];
+├── 📄 test.env (190 tokens, 25 lines)
+│   ├── PROMO_PATH
+│   ├── PRODUCTION
+│   ├── SQL_SCHEMA_PATH
+│   ├── DB_LOGS
+│   ├── DB_LOG
+│   ├── PGPASSWORD
+│   ├── PGDATABASE
+│   ├── PGHOST
+│   ├── PGPORT
+│   ├── PGUSER
+│   ├── SERVER_LOG
+│   ├── SERVER_LOGS
+│   ├── API_URL
+│   ├── APP_LOGS
+│   ├── APP_LOG
+│   ├── APP_URL
+│   ├── COGNITO_USER_POOL_ID
+│   ├── COGNITO_APP_CLIENT_ID
+│   ├── AWS_REGION
+│   └── STRIPE_SECRET_KEY
+├── 📄 testJsonSchema.json (421 tokens, 48 lines)
+│   ├── $schema: http://json-schema.org/draft-07/schema#
+│   ├── type: object
+│   ├── title: random_test
+│   └── description: A promoter's activites related to events
+├── 📄 testPackage.json (349 tokens, 43 lines)
+│   ├── name: 'promo-app'
+│   ├── version: 0.0.0
+│   ├── scripts:
+│   ├──     ng: 'ng'
+│   ├──     start: 'ng serve'
+│   ├──     build: 'ng build'
+│   ├──     watch: 'ng build --watch --configuration development'
+│   └──     test: 'ng test'
+└── 📄 tickets.component.ts (7,160 tokens, 903 lines)
+    ├── interface EnrichedTicket extends Ticket
+    ├── interface SpinConfig
+    ├── interface RotationState
+    ├── interface SpeakInput
+    ├── const formatSpeakInput = (input: SpeakInput): string =>
+    ├── function hourToSpeech(hour: number, minute: number, period: string):
+    │   string
+    ├── export class TicketsComponent implements AfterViewInit
+    ├──   speak(input: SpeakInput)
+    ├──   speakEvent(ticket: EnrichedTicket): void
+    ├──   formatEvent(ticket: EnrichedTicket): string
+    ├──   speakVenue(ticket: EnrichedTicket): void
+    ├──   formatDate(date: Date, oneLiner: boolean = false): string
+    ├──   formatDateForSpeech(date: Date): string
+    ├──   async spinQRCode(
+    │       event: PointerEvent,
+    │       config: SpinConfig = DEFAULT_SPIN_CONFIG
+    │     )
+    ├──   private animateRotation(
+    │       imgElement: HTMLElement,
+    │       targetRotation: number,
+    │       config: SpinConfig,
+    │       cleanup: () => void
+    │     )
+    ├──     const animate = (currentTime: number) =>
+    ├──         requestAnimationFrame(animate)
+    ├──         cleanup()
+    ├──     requestAnimationFrame(animate)
+    ├──   private getNext90Degree(currentRotation: number): number
+    ├──   private getCurrentRotation(matrix: string): number
+    ├──   ngAfterViewInit()
+    ├──       const mouseEnterListener = () =>
+    ├──       const mouseLeaveListener = () =>
+    ├──   ngOnDestroy()
+    ├──   toggleColumn(event: MatOptionSelectionChange, column: string)
+    ├──   adjustColumns(event?: Event)
+    ├──   onResize(event: Event)
+    ├──   async ngOnInit()
+    ├──   async loadTickets(): Promise<void>
+    ├──   onDateRangeChange(
+    │       type: "start" | "end",
+    │       event: MatDatepickerInputEvent<Date>
+    │     )
+    ├──   applyFilter(column: string): void
+    ├──   formatDateForComparison(date: Date): string
+    ├──   constructor(private renderer: Renderer2)
+    ├──   onFilterChange(event: Event, column: string)
+    ├──   onLatitudeChange(event: Event)
+    ├──   onLongitudeChange(event: Event)
+    ├──   onRadiusChange(event: Event)
+    ├──   sortData(sort: Sort): void
+    ├──   onRowClick(event: Event, row: any)
+    ├── function isDate(value: Date | undefined | null): value is Date
+    ├── function isNonNullNumber(value: number | null): value is number
+    ├── function hasLocation(
+    │     ticket: any
+    │   ): ticket is
+    ├── const create_faker_ticket = async () =>
+    ├── function compare(a: number | string, b: number | string, isAsc: boolean)
+    ├── function compare_dates(a: Date, b: Date, isAsc: boolean)
+    ├── async function mockMoreTickets(): Promise<Ticket[]>
+    ├── const mockTickets = async () =>
+    └── const renderQRCode = async (text: String): Promise<string> =>
diff --git a/tests/golden/legacy/trees/more_languages_group6.txt b/tests/golden/legacy/trees/more_languages_group6.txt
new file mode 100644
index 0000000..63e4015
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group6.txt
@@ -0,0 +1,615 @@
+📁 group6 (1 folder, 14 files)
+├── 📄 catastrophic.c (5,339 tokens, 754 lines)
+│   ├── TODO: technically we should use a proper parser
+│   ├── struct Point
+│   ├──     int x;
+│   ├──     int y;
+│   ├── struct Point getOrigin()
+│   ├── float mul_two_floats(float x1, float x2)
+│   ├── enum days
+│   ├──     SUN,
+│   ├──     MON,
+│   ├──     TUE,
+│   ├──     WED,
+│   ├──     THU,
+│   ├──     FRI,
+│   ├──     SAT
+│   ├── enum worker_pool_flags
+│   ├──         POOL_BH                 = 1 << 0,
+│   ├──         POOL_MANAGER_ACTIVE     = 1 << 1,
+│   ├──         POOL_DISASSOCIATED      = 1 << 2,
+│   ├──         POOL_BH_DRAINING        = 1 << 3,
+│   ├── enum worker_flags
+│   ├──         WORKER_DIE              = 1 << 1,
+│   ├──         WORKER_IDLE             = 1 << 2,
+│   ├──         WORKER_PREP             = 1 << 3,
+│   ├──         WORKER_CPU_INTENSIVE    = 1 << 6,
+│   ├──         WORKER_UNBOUND          = 1 << 7,
+│   ├──         WORKER_REBOUND          = 1 << 8,
+│   ├──         WORKER_NOT_RUNNING      = WORKER_PREP | WORKER_CPU_INTENSIVE |
+│   │                                     WORKER_UNBOUND | WORKER_REBOUND,
+│   ├── struct worker_pool
+│   ├──   raw_spinlock_t    lock;
+│   ├──   int      cpu;
+│   ├──   int      node;
+│   ├──   int      id;
+│   ├──   unsigned int    flags;
+│   ├──   unsigned long    watchdog_ts;
+│   ├──   bool      cpu_stall;
+│   ├──   int      nr_running;
+│   ├──   struct list_head  worklist;
+│   ├──   int      nr_workers;
+│   ├──   int      nr_idle;
+│   ├──   struct list_head  idle_list;
+│   ├──   struct timer_list  idle_timer;
+│   ├──   struct work_struct      idle_cull_work;
+│   ├──   struct timer_list  mayday_timer;
+│   ├──   struct worker    *manager;
+│   ├──   struct list_head  workers;
+│   ├──   struct ida    worker_ida;
+│   ├──   struct workqueue_attrs  *attrs;
+│   ├──   struct hlist_node  hash_node;
+│   ├──   int      refcnt;
+│   ├──   struct rcu_head    rcu;
+│   ├── long add_two_longs(long x1, long x2)
+│   ├── double multiplyByTwo(double num)
+│   ├── char getFirstCharacter(char *str)
+│   ├── void greet(Person p)
+│   ├── typedef struct
+│   ├──     char name[50];
+│   ├── } Person;
+│   ├── typedef struct PersonA
+│   ├──     char name[50];
+│   ├── } PersonB;
+│   ├── int main()
+│   ├── int* getArrayStart(int arr[], int size)
+│   ├── long complexFunctionWithMultipleArguments(
+│   │       int param1,
+│   │       double param2,
+│   │       char *param3,
+│   │       struct Point point
+│   │   )
+│   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   *channel, int channellen, int is_pattern)
+│   ├──     while((ln = listNext(&li)))
+│   ├── static struct config
+│   ├──     aeEventLoop *el;
+│   ├──     cliConnInfo conn_info;
+│   ├──     const char *hostsocket;
+│   ├──     int tls;
+│   ├──     struct cliSSLconfig sslconfig;
+│   ├── } config;
+│   ├── class Person
+│   ├──     std::string name;
+│   ├── public:
+│   ├──     Person(std::string n) : name(n)
+│   ├──     void greet()
+│   ├── void globalGreet()
+│   ├── int main()
+│   ├── void printMessage(const std::string &message)
+│   ├── template<typename T>
+│   │   void printVector(const std::vector<T>& vec)
+│   ├── struct foo
+│   ├──     char x;
+│   ├──     struct foo_in
+│   ├──           char* y;
+│   ├──           short z;
+│   ├──     } inner;
+│   ├── struct Point
+│   ├──     int x, y;
+│   ├──     Point(int x, int y) : x(x), y(y)
+│   ├── class Animal
+│   ├──   public:
+│   ├──     Animal(const std::string &name) : name(name)
+│   ├──     virtual void speak() const
+│   ├──     virtual ~Animal()
+│   ├── protected:
+│   ├──     std::string name;
+│   ├── class Dog : public Animal
+│   ├──   public:
+│   ├──     Dog(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── class Cat : public Animal
+│   ├──   public:
+│   ├──     Cat(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── class CatDog: public Animal, public Cat, public Dog
+│   ├──   public:
+│   ├──       CatDog(const std::string &name) : Animal(name)
+│   ├──       int meow_bark()
+│   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   num_layers,
+│   │                                int batch_size, int max_seq_length, float
+│   │   dropout,
+│   │                                bool bidirectional, bool cudnn_allow_tf32,
+│   │                                int workspace_size, int reserve_space_size)
+│   ├── int main()
+│   ├── enum ECarTypes
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV,
+│   ├──   Wagon
+│   ├── ECarTypes GetPreferredCarType()
+│   ├── enum ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── enum class ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── void myFunction(string fname, int age)
+│   ├── template <typename T> T cos(T)
+│   ├── template <typename T> T sin(T)
+│   ├── template <typename T> T sqrt(T)
+│   ├── template<typename T> struct VLEN
+│   ├── template<typename T> class arr
+│   ├──   private:
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──   public:
+│   ├──     arr() : p(0), sz(0)
+│   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   ├──     arr(arr &&other)
+│   │         : p(other.p), sz(other.sz)
+│   ├──     ~arr()
+│   ├──     void resize(size_t n)
+│   ├──     T &operator[](size_t idx)
+│   ├──     T *data()
+│   ├──     size_t size() const
+│   ├── class Buffer
+│   ├──  private:
+│   ├──   void* ptr_;
+│   ├── std::tuple<array, array, array> quantize(
+│   │       const array& w,
+│   │       int group_size,
+│   │       int bits,
+│   │       StreamOrDevice s)
+│   ├── #define PY_SSIZE_T_CLEAN
+│   ├── #define PLATFORM_IS_X86
+│   ├── #define PLATFORM_WINDOWS
+│   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   ├── static int GetXCR0EAX()
+│   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   ├── static int GetXCR0EAX()
+│   ├──   asm("XGETBV" : "=a"(eax), "=d"(edx) : "c"(0))
+│   ├── static void ReportMissingCpuFeature(const char* name)
+│   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   ├── static PyMethodDef cpu_feature_guard_methods[]
+│   ├── static struct PyModuleDef cpu_feature_guard_module
+│   ├── #define EXPORT_SYMBOL __declspec(dllexport)
+│   ├── #define EXPORT_SYMBOL __attribute__ ((visibility("default")))
+│   ├── EXPORT_SYMBOL PyMODINIT_FUNC PyInit_cpu_feature_guard(void)
+│   ├── typedef struct
+│   ├──     GPT2Config config;
+│   ├──     ParameterTensors params;
+│   ├──     size_t param_sizes[NUM_PARAMETER_TENSORS];
+│   ├──     float* params_memory;
+│   ├──     size_t num_parameters;
+│   ├──     ParameterTensors grads;
+│   ├──     float* grads_memory;
+│   ├──     float* m_memory;
+│   ├──     float* v_memory;
+│   ├──     ActivationTensors acts;
+│   ├──     size_t act_sizes[NUM_ACTIVATION_TENSORS];
+│   ├──     float* acts_memory;
+│   ├──     size_t num_activations;
+│   ├──     ActivationTensors grads_acts;
+│   ├──     float* grads_acts_memory;
+│   ├──     int batch_size;
+│   ├──     int seq_len;
+│   ├──     int* inputs;
+│   ├──     int* targets;
+│   ├──     float mean_loss;
+│   └── } GPT2;
+├── 📄 cpp_examples_impl.cc (60 tokens, 10 lines)
+│   ├── PYBIND11_MODULE(cpp_examples, m)
+│   └──     m.def("add", &add<int>, "An example function to add two numbers.")
+├── 📄 cpp_examples_impl.cu (37 tokens, 10 lines)
+│   ├── template <typename T>
+│   │   T add(T a, T b)
+│   └── template <>
+│       int add<int>(int a, int b)
+├── 📄 cpp_examples_impl.h (22 tokens, 6 lines)
+│   ├── template <typename T>
+│   │   T add(T a, T b)
+│   └── template <>
+│       int add<int>(int, int)
+├── 📄 edge_case.hpp (426 tokens, 28 lines)
+├── 📄 fractal.thy (1,712 tokens, 147 lines)
+│   ├── Title:      fractal.thy
+│   ├── Author:     Isabelle/HOL Contributors!
+│   ├── Author:     edge cases r us
+│   ├── theory Simplified_Ring
+│   ├── section ‹Basic Algebraic Structures›
+│   ├── class everything = nothing + itself
+│   ├── subsection ‹Monoids›
+│   ├── definition ring_hom :: "[('a, 'm) ring_scheme, ('b, 'n) ring_scheme] =>
+│   │   ('a => 'b) set"
+│   ├── fun example_fun :: "nat ⇒ nat"
+│   ├── locale monoid =
+│   │     fixes G (structure)
+│   │     assumes m_closed: "⟦x ∈ carrier G; y ∈ carrier G⟧ ⟹  x ⊗ y ∈ carrier
+│   │   G"
+│   │       and m_assoc: "⟦x ∈ carrier G; y ∈ carrier G; z ∈ carrier G⟧ ⟹  (x ⊗
+│   │   y) ⊗ z = x ⊗ (y ⊗ z)"
+│   │       and one_closed: "𝟭 ∈ carrier G"
+│   │       and l_one: "x ∈ carrier G ⟹  𝟭 ⊗ x = x"
+│   │       and r_one: "x ∈ carrier G ⟹  x ⊗ 𝟭 = x"
+│   ├── subsection ‹Groups›
+│   ├── locale group = monoid +
+│   │     assumes Units_closed: "x ∈ Units G ⟹  x ∈ carrier G"
+│   │       and l_inv_ex: "x ∈ carrier G ⟹  ∃ y ∈ carrier G. y ⊗ x = 𝟭"
+│   │       and r_inv_ex: "x ∈ carrier G ⟹  ∃ y ∈ carrier G. x ⊗ y = 𝟭"
+│   ├── subsection ‹Rings›
+│   ├── locale ring = abelian_group R + monoid R +
+│   │     assumes l_distr: "⟦x ∈ carrier R; y ∈ carrier R; z ∈ carrier R⟧ ⟹  (x
+│   │   ⊕ y) ⊗ z = x ⊗ z ⊕ y ⊗ z"
+│   │       and r_distr: "⟦x ∈ carrier R; y ∈ carrier R; z ∈ carrier R⟧ ⟹  z ⊗
+│   │   (x ⊕ y) = z ⊗ x ⊕ z ⊗ y"
+│   ├── locale commutative_ring = ring +
+│   │     assumes m_commutative: "⟦x ∈ carrier R; y ∈ carrier R⟧ ⟹  x ⊗ y = y ⊗
+│   │   x"
+│   ├── locale domain = commutative_ring +
+│   │     assumes no_zero_divisors: "⟦a ⊗ b = 𝟬; a ∈ carrier R; b ∈ carrier R⟧ ⟹
+│   │   a = 𝟬 ∨ b = 𝟬"
+│   ├── locale field = domain +
+│   │     assumes inv_ex: "x ∈ carrier R - {𝟬} ⟹  inv x ∈ carrier R"
+│   ├── subsection ‹Morphisms›
+│   ├── lemma example_lemma: "example_fun n = n"
+│   ├── qualified lemma gcd_0:
+│   │     "gcd a 0 = normalize a"
+│   ├── lemma abelian_monoidI:
+│   │     fixes R (structure)
+│   │         and f :: "'edge::{} ⇒ 'case::{}"
+│   │     assumes "⋀x y. ⟦ x ∈ carrier R; y ∈ carrier R ⟧ ⟹  x ⊕ y ∈ carrier R"
+│   │         and "𝟬 ∈ carrier R"
+│   │         and "⋀x y z. ⟦ x ∈ carrier R; y ∈ carrier R; z ∈ carrier R ⟧ ⟹  (x
+│   │   ⊕ y) ⊕ z = x ⊕ (y ⊕ z)"
+│   │     shows "abelian_monoid R"
+│   ├── lemma euclidean_size_gcd_le1 [simp]:
+│   │     assumes "a ≠ 0"
+│   │     shows "euclidean_size (gcd a b) ≤ euclidean_size a"
+│   ├── theorem Residue_theorem:
+│   │     fixes S pts::"complex set" and f::"complex ⇒ complex"
+│   │       and g::"real ⇒ complex"
+│   │     assumes "open S" "connected S" "finite pts" and
+│   │             holo:"f holomorphic_on S-pts" and
+│   │             "valid_path g" and
+│   │             loop:"pathfinish g = pathstart g" and
+│   │             "path_image g ⊆ S-pts" and
+│   │             homo:"∀z. (z ∉ S) ⟶  winding_number g z  = 0"
+│   │     shows "contour_integral g f = 2 * pi * 𝗂 *(∑p ∈ pts. winding_number g
+│   │   p * residue f p)"
+│   ├── corollary fps_coeff_residues_bigo':
+│   │     fixes f :: "complex ⇒ complex" and r :: real
+│   │     assumes exp: "f has_fps_expansion F"
+│   │     assumes "open A" "connected A" "cball 0 r ⊆ A" "r > 0"
+│   │     assumes "f holomorphic_on A - S" "S ⊆ ball 0 r" "finite S" "0 ∉ S"
+│   │     assumes "eventually (λn. g n = -(∑z ∈ S. residue (λz. f z / z ^ Suc n)
+│   │   z)) sequentially"
+│   │                (is "eventually (λn. _ = -?g' n) _")
+│   │     shows   "(λn. fps_nth F n - g n) ∈ O(λn. 1 / r ^ n)" (is "(λn. ?c n -
+│   │   _) ∈ O(_)")
+│   └── end
+├── 📄 Microsoft.PowerShell_profile.ps1 (3,346 tokens, 497 lines)
+│   ├── function Log($message)
+│   ├── function Remove-ChocolateyFromPath
+│   ├── function Show-Profiles
+│   ├── function Show-Path
+│   ├── function Show-Error($err)
+│   ├── function Get-ScoopPackagePath
+│   ├──         param(
+│   │                   [Parameter(Mandatory = $true)]
+│   │                   [string]$PackageName)
+│   ├── function Check-Command
+│   ├──         param(
+│   │                   [Parameter(Mandatory = $true)]
+│   │                   [string]$Name)
+│   ├── function Add-ToPath
+│   ├──         param(
+│   │                   [Parameter(Mandatory = $true)]
+│   │                   [string]$PathToAdd)
+│   ├── function Install-Scoop
+│   ├── function Scoop-Install
+│   ├──         param(
+│   │                   [Parameter(Mandatory = $true)]
+│   │                   [string]$Name,
+│   │                   [string]$PathToAdd)
+│   ├── function Start-CondaEnv
+│   ├── function Install-PipPackage
+│   ├──         param(
+│   │           [Parameter(Mandatory = $true)]
+│   │                   [string]$PackageName)
+│   ├── function Install-VSBuildTools
+│   ├── function Install-Crate
+│   ├──         param(
+│   │           [Parameter(Mandatory = $true)]
+│   │                   [string]$CrateName)
+│   ├── function Get-ScoopVersion
+│   ├── function Get-Version
+│   ├──     param(
+│   │           [Parameter(Mandatory = $true)]
+│   │           [string]$ExecutablePath,
+│   │           [string]$ExecutableName)
+│   ├── function Show-Requirements
+│   ├──         function Measure-Status
+│   ├──                 param(
+│   │                           [Parameter(Mandatory = $true)]
+│   │                           [string]$Name)
+│   ├── function Find-Profile
+│   ├── function Edit-Profile
+│   ├── function Set-Profile
+│   └── function Show-Profile
+├── 📄 python_complex_class.py (10 tokens, 2 lines)
+│   └── class Box(Space[NDArray[Any]])
+├── 📄 ramda__cloneRegExp.js (173 tokens, 9 lines)
+│   └── export default function _cloneRegExp(pattern)
+├── 📄 ramda_prop.js (646 tokens, 85 lines)
+│   ├── /**
+│   │    * Returns a function that when supplied an object returns the indicated
+│   │    * property of that object, if it exists.
+│   │    * @category Object
+│   │    * @typedefn Idx = String | Int | Symbol
+│   │    * @sig Idx -> {s: a} -> a | Undefined
+│   │    * @param {String|Number} p The property name or array index
+│   │    * @param {Object} obj The object to query
+│   │    * @return {*} The value at `obj.p`.
+│   │    */
+│   │   var prop = _curry2(function prop(p, obj)
+│   ├── /**
+│   │    * Solves equations of the form a * x = b
+│   │    * @param {{
+│   │    *  z: number
+│   │    * }} x
+│   │    */
+│   │   function foo(x)
+│   ├── /**
+│   │    * Deconstructs an array field from the input documents to output a
+│   │   document for each element.
+│   │    * Each output document is the input document with the value of the
+│   │   array field replaced by the element.
+│   │    * @category Object
+│   │    * @sig String -> {k: [v]} -> [{k: v}]
+│   │    * @param {String} key The key to determine which property of the object
+│   │   should be unwound.
+│   │    * @param {Object} object The object containing the list to unwind at
+│   │   the property named by the key.
+│   │    * @return {List} A list of new objects, each having the given key
+│   │   associated to an item from the unwound list.
+│   │    */
+│   │   var unwind = _curry2(function(key, object)
+│   └──   return _map(function(item)
+├── 📄 tensorflow_flags.h (7,628 tokens, 668 lines)
+│   ├── TF_DECLARE_FLAG('test_only_experiment_1')
+│   ├── TF_DECLARE_FLAG('test_only_experiment_2')
+│   ├── TF_DECLARE_FLAG('enable_nested_function_shape_inference'):
+│   │           Allow ops such as tf.cond to invoke the ShapeRefiner on their
+│   │   nested functions.
+│   ├── TF_DECLARE_FLAG('enable_quantized_dtypes_training'):
+│   │           Set quantized dtypes, like tf.qint8, to be trainable.
+│   ├── TF_DECLARE_FLAG('graph_building_optimization'):
+│   │           Optimize graph building for faster tf.function tracing.
+│   ├── TF_DECLARE_FLAG('saved_model_fingerprinting'):
+│   │           Add fingerprint to SavedModels.
+│   ├── TF_DECLARE_FLAG('more_stack_traces'):
+│   │           Enable experimental code that preserves and propagates graph
+│   │   node stack traces in C++.
+│   ├── TF_DECLARE_FLAG('publish_function_graphs'):
+│   │           Enables the publication of partitioned function graphs via
+│   │   StatsPublisherInterface. Disabling this flag can reduce memory
+│   │   consumption.
+│   ├── TF_DECLARE_FLAG('enable_aggressive_constant_replication'):
+│   │           Replicate constants across CPU devices and even for local CPUs
+│   │   within the same task if available.
+│   ├── TF_DECLARE_FLAG('enable_colocation_key_propagation_in_while_op_lowering'
+│   │   ):
+│   │           If true, colocation key attributes for the ops will be
+│   │   propagated during while op lowering to switch/merge ops.
+│   ├── Flag('tf_xla_auto_jit'):
+│   │           Control compilation of operators into XLA computations on CPU
+│   │   and GPU devices.  0 = use ConfigProto setting; -1 = off; 1 = on for
+│   │   things very likely to be improved; 2 = on for everything; (experimental)
+│   │   fusible = only for Tensorflow operations that XLA knows how to fuse. If
+│   │   set to single-gpu(<N>) then this resolves to <N> for single-GPU graphs
+│   │   (graphs that have at least one node placed on a GPU and no more than one
+│   │   GPU is in use through the entire graph) and 0 otherwise.  Experimental.
+│   ├── Flag('tf_xla_min_cluster_size'):
+│   │           Minimum number of operators in an XLA compilation. Ignored for
+│   │   operators placed on an XLA device or operators explicitly marked for
+│   │   compilation.
+│   ├── Flag('tf_xla_max_cluster_size'):
+│   │           Maximum number of operators in an XLA compilation.
+│   ├── Flag('tf_xla_cluster_exclude_ops'):
+│   │           (experimental) Exclude the operations from auto-clustering. If
+│   │   multiple, separate them with commas. Where, Some_other_ops.
+│   ├── Flag('tf_xla_clustering_debug'):
+│   │           Dump graphs during XLA compilation.
+│   ├── Flag('tf_xla_cpu_global_jit'):
+│   │           Enables global JIT compilation for CPU via SessionOptions.
+│   ├── Flag('tf_xla_clustering_fuel'):
+│   │           Places an artificial limit on the number of ops marked as
+│   │   eligible for clustering.
+│   ├── Flag('tf_xla_disable_deadness_safety_checks_for_debugging'):
+│   │           Disable deadness related safety checks when clustering (this is
+│   │   unsound).
+│   ├── Flag('tf_xla_disable_resource_variable_safety_checks_for_debugging'):
+│   │           Disable resource variables related safety checks when clustering
+│   │   (this is unsound).
+│   ├── Flag('tf_xla_deterministic_cluster_names'):
+│   │           Causes the function names assigned by auto clustering to be
+│   │   deterministic from run to run.
+│   ├── Flag('tf_xla_persistent_cache_directory'):
+│   │           If non-empty, JIT-compiled executables are saved to and loaded
+│   │   from the specified file system directory path. Empty by default.
+│   ├── Flag('tf_xla_persistent_cache_device_types'):
+│   │           If non-empty, the persistent cache will only be used for the
+│   │   specified devices (comma separated). Each device type should be able to
+│   │   be converted to.
+│   ├── Flag('tf_xla_persistent_cache_read_only'):
+│   │           If true, the persistent cache will be read-only.
+│   ├── Flag('tf_xla_disable_strict_signature_checks'):
+│   │           If true, entires loaded into the XLA compile cache will not have
+│   │   their signatures checked strictly. Defaults to false.
+│   ├── Flag('tf_xla_persistent_cache_prefix'):
+│   │           Specifies the persistance cache prefix. Default is.
+│   ├── Flag('tf_xla_sparse_core_disable_table_stacking'):
+│   │           Disable table stacking for all the tables passed to the
+│   │   SparseCore mid level API.
+│   ├── Flag('tf_xla_sparse_core_minibatch_max_division_level'):
+│   │           Max level of division to split input data into minibatches.
+│   ├── Flag('tf_xla_sparse_core_stacking_mem_limit_bytes'):
+│   │           If non-zero, limits the size of the activations for a given
+│   │   table to be below these many bytes.
+│   ├── Flag('tf_xla_sparse_core_stacking_table_shard_limit_bytes'):
+│   │           If non-zero, limits the size of any table shard to be below
+│   │   these many bytes.
+│   ├── Flag('always_specialize')
+│   ├── Flag('cost_driven_async_parallel_for')
+│   ├── Flag('enable_crash_reproducer')
+│   ├── Flag('log_query_of_death')
+│   ├── Flag('vectorize')
+│   ├── Flag('tf_xla_enable_lazy_compilation')
+│   ├── Flag('tf_xla_print_cluster_outputs'):
+│   │           If true then insert Print nodes to print out values produced by
+│   │   XLA clusters.
+│   ├── Flag('tf_xla_check_cluster_input_numerics'):
+│   │           If true then insert CheckNumerics nodes to check all cluster
+│   │   inputs.
+│   ├── Flag('tf_xla_check_cluster_output_numerics'):
+│   │           If true then insert CheckNumerics nodes to check all cluster
+│   │   outputs.
+│   ├── Flag('tf_xla_disable_constant_folding'):
+│   │           If true then disables constant folding on TF graph before XLA
+│   │   compilation.
+│   ├── Flag('tf_xla_disable_full_embedding_pipelining'):
+│   │           If true then disables full embedding pipelining and instead use
+│   │   strict SparseCore / TensorCore sequencing.
+│   ├── Flag('tf_xla_embedding_parallel_iterations'):
+│   │           If >0 then use this many parallel iterations in
+│   │   embedding_pipelining and embedding_sequency. By default, use the
+│   │   parallel_iterations on the original model WhileOp.
+│   ├── Flag('tf_xla_compile_on_demand'):
+│   │           Switch a device into 'on-demand' mode, where instead of
+│   │   autoclustering ops are compiled one by one just-in-time.
+│   ├── Flag('tf_xla_enable_xla_devices'):
+│   │           Generate XLA_* devices, where placing a computation on such a
+│   │   device forces compilation by XLA. Deprecated.
+│   ├── Flag('tf_xla_always_defer_compilation')
+│   ├── Flag('tf_xla_async_compilation'):
+│   │           When lazy compilation is enabled, asynchronous compilation
+│   │   starts the cluster compilation in the background, and the fallback path
+│   │   is executed until the compilation has finished.
+│   ├── Flag('tf_xla_use_device_api_for_xla_launch'):
+│   │           If true, uses Device API (PjRt) for single device compilation
+│   │   and execution of functions marked for JIT compilation i.e.
+│   │   jit_compile=True. Defaults to false.
+│   ├── Flag('tf_xla_use_device_api_for_compile_on_demand'):
+│   │           If true, uses Device API (PjRt) for compiling and executing ops
+│   │   one by one in 'on-demand' mode. Defaults to false.
+│   ├── Flag('tf_xla_use_device_api_for_auto_jit'):
+│   │           If true, uses Device API (PjRt) for compilation and execution
+│   │   when auto-clustering is enabled. Defaults to false.
+│   ├── Flag('tf_xla_use_device_api'):
+│   │           If true, uses Device API (PjRt) for compilation and execution of
+│   │   ops one-by-one in 'on-demand' mode, for functions marked for JIT
+│   │   compilation, or when auto-clustering is enabled. Defaults to false.
+│   ├── Flag('tf_xla_enable_device_api_for_gpu'):
+│   │           If true, uses Device API (PjRt) for TF GPU device. This is a
+│   │   helper flag so that individual tests can turn on PjRt for GPU
+│   │   specifically.
+│   ├── Flag('tf_xla_call_module_disabled_checks'):
+│   │           A comma-sepated list of directives specifying the safety checks
+│   │   to be skipped when compiling XlaCallModuleOp. See the op documentation
+│   │   for the recognized values.
+│   ├── Flag('tf_mlir_enable_mlir_bridge'):
+│   │           Enables experimental MLIR-Based TensorFlow Compiler Bridge.
+│   ├── Flag('tf_mlir_enable_merge_control_flow_pass'):
+│   │           Enables MergeControlFlow pass for MLIR-Based TensorFlow Compiler
+│   │   Bridge.
+│   ├── Flag('tf_mlir_enable_convert_control_to_data_outputs_pass'):
+│   │           Enables MLIR-Based TensorFlow Compiler Bridge.
+│   ├── Flag('tf_mlir_enable_strict_clusters'):
+│   │           Do not allow clusters that have cyclic control dependencies.
+│   ├── Flag('tf_mlir_enable_multiple_local_cpu_devices'):
+│   │           Enable multiple local CPU devices. CPU ops which are outside
+│   │   compiled inside the tpu cluster will also be replicated across multiple
+│   │   cpu devices.
+│   ├── Flag('tf_dump_graphs_in_tfg'):
+│   │           When tf_dump_graphs_in_tfg is true, graphs after transformations
+│   │   are dumped in MLIR TFG dialect and not in GraphDef.
+│   ├── Flag('tf_mlir_enable_generic_outside_compilation'):
+│   │           Enables OutsideCompilation passes for MLIR-Based TensorFlow
+│   │   Generic Compiler Bridge.
+│   ├── Flag('tf_mlir_enable_tpu_variable_runtime_reformatting_pass'):
+│   │           Enables TPUVariableRuntimeReformatting pass for MLIR-Based
+│   │   TensorFlow Compiler Bridge. This enables weight update sharding and
+│   │   creates TPUReshardVariables ops.
+│   ├── TF_PY_DECLARE_FLAG('test_only_experiment_1')
+│   ├── TF_PY_DECLARE_FLAG('test_only_experiment_2')
+│   ├── TF_PY_DECLARE_FLAG('enable_nested_function_shape_inference')
+│   ├── TF_PY_DECLARE_FLAG('enable_quantized_dtypes_training')
+│   ├── TF_PY_DECLARE_FLAG('graph_building_optimization')
+│   ├── TF_PY_DECLARE_FLAG('op_building_optimization')
+│   ├── TF_PY_DECLARE_FLAG('saved_model_fingerprinting')
+│   ├── TF_PY_DECLARE_FLAG('tf_shape_default_int64')
+│   ├── TF_PY_DECLARE_FLAG('more_stack_traces')
+│   ├── TF_PY_DECLARE_FLAG('publish_function_graphs')
+│   ├── TF_PY_DECLARE_FLAG('enable_aggressive_constant_replication')
+│   ├── TF_PY_DECLARE_FLAG('enable_colocation_key_propagation_in_while_op_loweri
+│   │   ng')
+│   ├── #define TENSORFLOW_CORE_CONFIG_FLAG_DEFS_H_
+│   ├── class Flags
+│   ├──  public:
+│   ├── bool SetterForXlaAutoJitFlag(const string& value)
+│   ├── bool SetterForXlaCallModuleDisabledChecks(const string& value)
+│   ├── void AppendMarkForCompilationPassFlagsInternal(std::vector<Flag>*
+│   │   flag_list)
+│   ├── void AllocateAndParseJitRtFlags()
+│   ├── void AllocateAndParseFlags()
+│   ├── void ResetFlags()
+│   ├── bool SetXlaAutoJitFlagFromFlagString(const string& value)
+│   ├── BuildXlaOpsPassFlags* GetBuildXlaOpsPassFlags()
+│   ├── MarkForCompilationPassFlags* GetMarkForCompilationPassFlags()
+│   ├── XlaSparseCoreFlags* GetXlaSparseCoreFlags()
+│   ├── XlaDeviceFlags* GetXlaDeviceFlags()
+│   ├── XlaOpsCommonFlags* GetXlaOpsCommonFlags()
+│   ├── XlaCallModuleFlags* GetXlaCallModuleFlags()
+│   ├── MlirCommonFlags* GetMlirCommonFlags()
+│   ├── void ResetJitCompilerFlags()
+│   ├── const JitRtFlags& GetJitRtFlags()
+│   ├── ConfigProto::Experimental::MlirBridgeRollout GetMlirBridgeRolloutState(
+│   │       std::optional<const ConfigProto> config_proto)
+│   ├── void AppendMarkForCompilationPassFlags(std::vector<Flag>* flag_list)
+│   ├── void DisableXlaCompilation()
+│   ├── void EnableXlaCompilation()
+│   ├── bool FailOnXlaCompilation()
+│   ├── #define TF_PY_DECLARE_FLAG(flag_name)
+│   └── PYBIND11_MODULE(flags_pybind, m)
+├── 📄 test.f (181 tokens, 30 lines)
+│   ├── MODULE basic_mod
+│   ├──     TYPE :: person
+│   │           CHARACTER(LEN=50) :: name
+│   │           INTEGER :: age
+│   │       END TYPE person
+│   ├──     SUBROUTINE short_hello(happy, path)
+│   │       END SUBROUTINE short_hello
+│   ├──     SUBROUTINE long_hello(
+│   │           p,
+│   │           message
+│   │       )
+│   │       END SUBROUTINE long_hello
+│   ├── END MODULE basic_mod
+│   └── PROGRAM HelloFortran
+│       END PROGRAM HelloFortran
+├── 📄 torch.rst (60 tokens, 8 lines)
+│   ├── # libtorch (C++-only)
+│   └── - Building libtorch using Python
+└── 📄 yc.html (9,063 tokens, 169 lines)
diff --git a/tests/golden/legacy/trees/more_languages_group7.txt b/tests/golden/legacy/trees/more_languages_group7.txt
new file mode 100644
index 0000000..1de0ae4
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group7.txt
@@ -0,0 +1,126 @@
+📁 group7 (1 folder, 5 files)
+├── 📄 absurdly_huge.jsonl (8,347 tokens, 126 lines)
+│   ├── SMILES: str
+│   ├── Yield: float
+│   ├── Temperature: int
+│   ├── Pressure: float
+│   ├── Solvent: str
+│   ├── Success: bool
+│   ├── Reaction_Conditions: dict
+│   ├── Products: list
+│   └── EdgeCasesMissed: None
+├── 📄 angular_crud.ts (1,192 tokens, 148 lines)
+│   ├── interface DBCommand<T = any>
+│   ├── export class IndexedDbService
+│   ├──     constructor()
+│   ├──     async create_connection({ db_name = 'client_db', table_name }:
+│   │   DBCommand)
+│   ├──                 upgrade(db)
+│   ├──     async create_model({ db_name, table_name, model }: DBCommand)
+│   ├──         verify_matching({ table_name, model })
+│   ├──     async read_key({ db_name, table_name, key }: DBCommand)
+│   ├──     async update_model({ db_name, table_name, model }: DBCommand)
+│   ├──         verify_matching({ table_name, model })
+│   ├──     async delete_key({ db_name, table_name, key }: DBCommand)
+│   ├──     async list_table({
+│   │           db_name,
+│   │           table_name,
+│   │           where,
+│   │       }: DBCommand & { where?: { [key: string]: string | number } })
+│   └──     async search_table(criteria: SearchCriteria)
+├── 📄 structure.py (400 tokens, 92 lines)
+│   ├── @runtime_checkable
+│   │   class DataClass(Protocol)
+│   ├──     __dataclass_fields__: dict
+│   ├── class MyInteger(Enum)
+│   ├──     ONE = 1
+│   ├──     TWO = 2
+│   ├──     THREE = 42
+│   ├── class MyString(Enum)
+│   ├──     AAA1 = "aaa"
+│   ├──     BB_B = """edge
+│   │   case"""
+│   ├── @dataclass(frozen=True, slots=True, kw_only=True)
+│   │   class Tool
+│   ├──     name: str
+│   ├──     description: str
+│   ├──     input_model: DataClass
+│   ├──     output_model: DataClass
+│   ├──     def execute(self, *args, **kwargs)
+│   ├──     @property
+│   │       def edge_case(self) -> str
+│   ├──     def should_still_see_me(self, x: bool = True) -> "Tool"
+│   ├── @dataclass
+│   │   class MyInput[T]
+│   ├──     name: str
+│   ├──     rank: MyInteger
+│   ├──     serial_n: int
+│   ├── @dataclass
+│   │   class Thingy
+│   ├──     is_edge_case: bool
+│   ├── @dataclass
+│   │   class MyOutput
+│   ├──     orders: str
+│   ├── class MyTools(Enum)
+│   ├──     TOOL_A = Tool(
+│   │           name="complicated",
+│   │           description="edge case!",
+│   │           input_model=MyInput[Thingy],
+│   │           output_model=MyOutput,
+│   │       )
+│   ├──     TOOL_B = Tool(
+│   │           name="""super
+│   │   complicated
+│   │   """,
+│   │           description="edge case!",
+│   │           input_model=MyInput,
+│   │           output_model=MyOutput,
+│   │       )
+│   ├── @final
+│   │   class dtype(Generic[_DTypeScalar_co])
+│   └──     names: None | tuple[builtins.str, ...]
+├── 📄 test.wgsl (528 tokens, 87 lines)
+│   ├── alias MyVec = vec4<f32>
+│   ├── alias AnotherVec = vec2<f32>
+│   ├── struct VertexInput
+│   ├── struct VertexOutput
+│   ├── struct MyUniforms
+│   ├── @group(0) @binding(0) var<uniform> u_mvp: mat4x4<f32>
+│   ├── @group(0) @binding(1) var<uniform> u_color: MyVec
+│   ├── @group(1) @binding(0) var my_texture: texture_2d<f32>
+│   ├── @group(1) @binding(1) var my_sampler: sampler
+│   ├── @vertex
+│   │   fn vs_main(in: VertexInput) -> VertexOutput
+│   ├── @fragment
+│   │   fn fs_main(in: VertexOutput) -> @location(0) vec4<f32>
+│   ├── @compute @workgroup_size(8, 8, 1)
+│   │   fn cs_main(@builtin(global_invocation_id) global_id: vec3<u32>)
+│   ├── fn helper_function(val: f32) -> f32
+│   ├── struct AnotherStruct
+│   └── @compute
+│       @workgroup_size(8, 8, 1)
+│       fn multi_line_edge_case(
+│           @builtin(global_invocation_id)
+│           globalId       : vec3<u32>,
+│           @group(1)
+│           @binding(0)
+│           srcTexture     : texture_2d<f32>,
+│           @group(1)
+│           @binding(1)
+│           srcSampler     : sampler,
+│           @group(0)
+│           @binding(0)
+│           uniformsPtr    : ptr<uniform, MyUniforms>,
+│           storageBuffer  : ptr<storage, array<vec4<f32>, 64>, read_write>,
+│       )
+└── 📄 test.metal (272 tokens, 34 lines)
+    ├── struct MyData
+    ├── kernel void myKernel(device MyData* data [[buffer(0)]],
+    │                        uint id [[thread_position_in_grid]])
+    ├── float myHelperFunction(float x, float y)
+    ├── vertex float4 vertexShader(const device packed_float3* vertex_array
+    │   [[buffer(0)]],
+    │                              unsigned int vid [[vertex_id]])
+    ├── fragment half4 fragmentShader(float4 P [[position]])
+    └── float3 computeNormalMap(ColorInOut in, texture2d<float>
+        normalMapTexture)
diff --git a/tests/golden/legacy/trees/more_languages_group_lisp.txt b/tests/golden/legacy/trees/more_languages_group_lisp.txt
new file mode 100644
index 0000000..3aed61a
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group_lisp.txt
@@ -0,0 +1,22 @@
+📁 group_lisp (1 folder, 4 files)
+├── 📄 clojure_test.clj (682 tokens, 85 lines)
+│   ├── defprotocol P
+│   ├── defrecord Person
+│   ├── defn -main
+│   ├── ns bion.likes_trees
+│   ├── def repo-url
+│   ├── defn config
+│   ├── defmacro with-os
+│   └── defrecord SetFullElement
+├── 📄 LispTest.lisp (25 tokens, 6 lines)
+│   ├── defstruct person
+│   └── defun greet
+├── 📄 racket_struct.rkt (14 tokens, 1 line)
+│   └── struct point
+└── 📄 test_scheme.scm (360 tokens, 44 lines)
+    ├── define topological-sort
+    ├──   define table
+    ├──   define queue
+    ├──   define result
+    ├──   define set-up
+    └──   define traverse
diff --git a/tests/golden/legacy/trees/more_languages_group_todo.txt b/tests/golden/legacy/trees/more_languages_group_todo.txt
new file mode 100644
index 0000000..b2d6fe7
--- /dev/null
+++ b/tests/golden/legacy/trees/more_languages_group_todo.txt
@@ -0,0 +1,111 @@
+📁 group_todo (1 folder, 12 files)
+├── 📄 AAPLShaders.metal (5,780 tokens, 566 lines)
+│   ├── struct LightingParameters
+│   ├── float Geometry(float Ndotv, float alphaG)
+│   ├── float3 computeNormalMap(ColorInOut in, texture2d<float>
+│   │   normalMapTexture)
+│   ├── float3 computeDiffuse(LightingParameters parameters)
+│   ├── float Distribution(float NdotH, float roughness)
+│   ├── float3 computeSpecular(LightingParameters parameters)
+│   ├── float4 equirectangularSample(float3 direction, sampler s,
+│   │   texture2d<float> image)
+│   ├── LightingParameters calculateParameters(ColorInOut in,
+│   │                                          AAPLCameraData cameraData,
+│   │                                          constant AAPLLightData&
+│   │   lightData,
+│   │                                          texture2d<float>   baseColorMap,
+│   │                                          texture2d<float>   normalMap,
+│   │                                          texture2d<float>   metallicMap,
+│   │                                          texture2d<float>   roughnessMap,
+│   │                                          texture2d<float>
+│   │   ambientOcclusionMap,
+│   │                                          texture2d<float>   skydomeMap)
+│   ├── struct SkyboxVertex
+│   ├── struct SkyboxV2F
+│   ├── vertex SkyboxV2F skyboxVertex(SkyboxVertex in [[stage_in]],
+│   │                                    constant AAPLCameraData& cameraData
+│   │   [[buffer(BufferIndexCameraData)]])
+│   ├── fragment float4 skyboxFragment(SkyboxV2F v [[stage_in]],
+│   │   texture2d<float> skytexture [[texture(0)]])
+│   ├── vertex ColorInOut vertexShader(Vertex in [[stage_in]],
+│   │                                  constant AAPLInstanceTransform&
+│   │   instanceTransform [[ buffer(BufferIndexInstanceTransforms) ]],
+│   │                                  constant AAPLCameraData& cameraData [[
+│   │   buffer(BufferIndexCameraData) ]])
+│   ├── float2 calculateScreenCoord( float3 ndcpos )
+│   ├── fragment float4 fragmentShader(
+│   │                       ColorInOut                  in
+│   │   [[stage_in]],
+│   │                       constant AAPLCameraData&    cameraData            [[
+│   │   buffer(BufferIndexCameraData) ]],
+│   │                       constant AAPLLightData&     lightData             [[
+│   │   buffer(BufferIndexLightData) ]],
+│   │                       constant AAPLSubmeshKeypath&submeshKeypath        [[
+│   │   buffer(BufferIndexSubmeshKeypath)]],
+│   │                       constant Scene*             pScene                [[
+│   │   buffer(SceneIndex)]],
+│   │                       texture2d<float>            skydomeMap            [[
+│   │   texture(AAPLSkyDomeTexture) ]],
+│   │                       texture2d<float>            rtReflections         [[
+│   │   texture(AAPLTextureIndexReflections),
+│   │   function_constant(is_raytracing_enabled)]])
+│   ├── fragment float4 reflectionShader(ColorInOut in [[stage_in]],
+│   │                                    texture2d<float> rtReflections
+│   │   [[texture(AAPLTextureIndexReflections)]])
+│   ├── struct ThinGBufferOut
+│   ├── fragment ThinGBufferOut gBufferFragmentShader(ColorInOut in
+│   │   [[stage_in]])
+│   ├── kernel void rtReflection(
+│   │                texture2d< float, access::write >      outImage
+│   │   [[texture(OutImageIndex)]],
+│   │                texture2d< float >                     positions
+│   │   [[texture(ThinGBufferPositionIndex)]],
+│   │                texture2d< float >                     directions
+│   │   [[texture(ThinGBufferDirectionIndex)]],
+│   │                texture2d< float >                     skydomeMap
+│   │   [[texture(AAPLSkyDomeTexture)]],
+│   │                constant AAPLInstanceTransform*        instanceTransforms
+│   │   [[buffer(BufferIndexInstanceTransforms)]],
+│   │                constant AAPLCameraData&               cameraData
+│   │   [[buffer(BufferIndexCameraData)]],
+│   │                constant AAPLLightData&                lightData
+│   │   [[buffer(BufferIndexLightData)]],
+│   │                constant Scene*                        pScene
+│   │   [[buffer(SceneIndex)]],
+│   │                instance_acceleration_structure
+│   │   accelerationStructure   [[buffer(AccelerationStructureIndex)]],
+│   │                uint2 tid [[thread_position_in_grid]])
+│   ├── else if ( intersection.type == raytracing::intersection_type::none )
+│   ├── struct VertexInOut
+│   ├── vertex VertexInOut vertexPassthrough( uint vid [[vertex_id]] )
+│   ├── fragment float4 fragmentPassthrough( VertexInOut in [[stage_in]],
+│   │   texture2d< float > tin )
+│   ├── fragment float4 fragmentBloomThreshold( VertexInOut in [[stage_in]],
+│   │                                          texture2d< float > tin
+│   │   [[texture(0)]],
+│   │                                          constant float* threshold
+│   │   [[buffer(0)]] )
+│   └── fragment float4 fragmentPostprocessMerge( VertexInOut in [[stage_in]],
+│                                                constant float& exposure
+│       [[buffer(0)]],
+│                                                texture2d< float > texture0
+│       [[texture(0)]],
+│                                                texture2d< float > texture1
+│       [[texture(1)]])
+├── 📄 crystal_test.cr (48 tokens, 15 lines)
+├── 📄 dart_test.dart (108 tokens, 24 lines)
+├── 📄 elixir_test.exs (39 tokens, 10 lines)
+├── 📄 forward.frag (739 tokens, 87 lines)
+├── 📄 forward.vert (359 tokens, 48 lines)
+├── 📄 nodemon.json (118 tokens, 20 lines)
+├── 📄 sas_test.sas (97 tokens, 22 lines)
+├── 📄 test_setup_py.test (133 tokens, 24 lines)
+├── 📄 testTypings.d.ts (158 tokens, 23 lines)
+├── 📄 vba_test.bas (67 tokens, 16 lines)
+└── 📄 wgsl_test.wgsl (94 tokens, 17 lines)
+    ├── @binding(0) @group(0) var<uniform> frame : u32
+    ├── @vertex
+    │   fn vtx_main(@builtin(vertex_index) vertex_index : u32) ->
+    │   @builtin(position) vec4f
+    └── @fragment
+        fn frag_main() -> @location(0) vec4f
diff --git a/tests/golden/legacy/trees/multi_seed.txt b/tests/golden/legacy/trees/multi_seed.txt
new file mode 100644
index 0000000..e1fb21c
--- /dev/null
+++ b/tests/golden/legacy/trees/multi_seed.txt
@@ -0,0 +1,427 @@
+🌵 Root (2 folders, 17 files)
+├── 📁 group1 (1 folder, 11 files)
+│   ├── 📄 addamt.cobol (441 tokens, 40 lines)
+│   │   ├── IDENTIFICATION DIVISION.
+│   │   ├── PROGRAM-ID.
+│   │   │              ADDAMT.
+│   │   ├── DATA DIVISION.
+│   │   ├── WORKING-STORAGE SECTION.
+│   │   ├── 01  KEYED-INPUT.
+│   │   ├──     05  CUST-NO-IN.
+│   │   ├──     05  AMT1-IN.
+│   │   ├──     05  AMT2-IN.
+│   │   ├──     05  AMT3-IN.
+│   │   ├── 01  DISPLAYED-OUTPUT.
+│   │   ├──     05  CUST-NO-OUT.
+│   │   ├──     05  TOTAL-OUT.
+│   │   ├── 01  MORE-DATA.
+│   │   ├── PROCEDURE DIVISION.
+│   │   └── 100-MAIN.
+│   ├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+│   │   ├── IDENTIFICATION DIVISION.
+│   │   ├── PROGRAM-ID. CUSTOMER-INVOICE.
+│   │   ├── AUTHOR. JANE DOE.
+│   │   ├── DATE. 2023-12-30.
+│   │   ├──   DATE-COMPILED. 06/30/10.
+│   │   ├──     DATE-WRITTEN. 12/34/56.
+│   │   ├── ENVIRONMENT DIVISION.
+│   │   ├── INPUT-OUTPUT SECTION.
+│   │   ├── FILE-CONTROL.
+│   │   ├──     SELECT CUSTOMER-FILE.
+│   │   ├──     SELECT INVOICE-FILE.
+│   │   ├──     SELECT REPORT-FILE.
+│   │   ├── DATA DIVISION.
+│   │   ├── FILE SECTION.
+│   │   ├── FD CUSTOMER-FILE.
+│   │   ├── 01 CUSTOMER-RECORD.
+│   │   ├──    05 CUSTOMER-ID.
+│   │   ├──    05 CUSTOMER-NAME.
+│   │   ├──    05 CUSTOMER-BALANCE.
+│   │   ├── FD INVOICE-FILE.
+│   │   ├── 01 INVOICE-RECORD.
+│   │   ├──    05 INVOICE-ID.
+│   │   ├──    05 CUSTOMER-ID.
+│   │   ├──    05 INVOICE-AMOUNT.
+│   │   ├── FD REPORT-FILE.
+│   │   ├── 01 REPORT-RECORD.
+│   │   ├── WORKING-STORAGE SECTION.
+│   │   ├── 01 WS-CUSTOMER-FOUND.
+│   │   ├── 01 WS-END-OF-FILE.
+│   │   ├── 01 WS-TOTAL-BALANCE.
+│   │   ├── PROCEDURE DIVISION.
+│   │   ├── 0000-MAIN-ROUTINE.
+│   │   ├── 1000-PROCESS-RECORDS.
+│   │   ├── 1100-UPDATE-CUSTOMER-BALANCE.
+│   │   └── END PROGRAM CUSTOMER-INVOICE.
+│   ├── 📄 JavaTest.java (578 tokens, 86 lines)
+│   │   ├── abstract class LivingBeing
+│   │   ├──     abstract void breathe()
+│   │   ├── interface Communicator
+│   │   ├──     String communicate()
+│   │   ├── @Log
+│   │   ├── @Getter
+│   │   ├── @Setter
+│   │   ├── class Person extends LivingBeing implements Communicator
+│   │   ├──     Person(String name, int age)
+│   │   ├──     @Override
+│   │   ├──     void breathe()
+│   │   ├──     @Override
+│   │   ├──     public String communicate()
+│   │   ├──     void greet()
+│   │   ├──     String personalizedGreeting(String greeting, Optional<Boolean>
+│   │   │   includeAge)
+│   │   ├── @Singleton
+│   │   ├── @RestController
+│   │   ├── @SpringBootApplication
+│   │   ├── public class Example
+│   │   ├──     @Inject
+│   │   ├──     public Example(Person person)
+│   │   ├──     @RequestMapping("/greet")
+│   │   ├──     String home(@RequestParam(value = "name", defaultValue =
+│   │   │   "World") String name,
+│   │   │                   @RequestParam(value = "age", defaultValue = "30")
+│   │   │   int age)
+│   │   └──     public static void main(String[] args)
+│   ├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+│   │   ├── module JuliaTest_EdgeCase
+│   │   ├── struct Location
+│   │   │       name::String
+│   │   │       lat::Float32
+│   │   │       lon::Float32
+│   │   │   end
+│   │   ├── mutable struct mPerson
+│   │   │       name::String
+│   │   │       age::Int
+│   │   │   end
+│   │   ├── Base.@kwdef mutable struct Param
+│   │   │       Δt::Float64 = 0.1
+│   │   │       n::Int64
+│   │   │       m::Int64
+│   │   │   end
+│   │   ├──     sic(x,y)
+│   │   ├── welcome(l::Location)
+│   │   ├── ∑(α, Ω)
+│   │   ├── function noob()
+│   │   │   end
+│   │   ├── function ye_olde(hello::String, world::Location)
+│   │   │   end
+│   │   ├── function multiline_greet(
+│   │   │           p::mPerson,
+│   │   │           greeting::String
+│   │   │       )
+│   │   │   end
+│   │   ├── function julia_is_awesome(prob::DiffEqBase.AbstractDAEProblem{uType,
+│   │   │   duType, tType,
+│   │   │           isinplace};
+│   │   │       kwargs...) where {uType, duType, tType, isinplace}
+│   │   │   end
+│   │   └── end
+│   ├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+│   │   ├── data class Person(val name: String)
+│   │   ├── fun greet(person: Person)
+│   │   ├── fun <T> processItems(items: List<T>, processor: (T) -> Unit)
+│   │   ├── interface Source<out T>
+│   │   ├──     fun nextT(): T
+│   │   ├── fun MutableList<Int>.swap(index1: Int, index2: Int)
+│   │   ├── fun Any?.toString(): String
+│   │   ├── tailrec fun findFixPoint(x: Double = 1.0): Double
+│   │   ├── class GenericRepository<T>
+│   │   ├──     fun getItem(id: Int): T?
+│   │   ├── sealed interface Error
+│   │   ├── sealed class IOError(): Error
+│   │   ├── object Runner
+│   │   ├──     inline fun <reified S: SomeClass<T>, T> run() : T
+│   │   ├── infix fun Int.shl(x: Int): Int
+│   │   ├── class MyStringCollection
+│   │   ├──     infix fun add(s: String)
+│   │   ├──     fun build()
+│   │   ├── open class Base(p: Int)
+│   │   ├── class Derived(p: Int) : Base(p)
+│   │   ├── open class Shape
+│   │   ├──     open fun draw()
+│   │   ├──     fun fill()
+│   │   ├──     open fun edge(case: Int)
+│   │   ├── interface Thingy
+│   │   ├──     fun edge()
+│   │   ├── class Circle() : Shape(), Thingy
+│   │   ├──     override fun draw()
+│   │   ├──     final override fun edge(case: Int)
+│   │   ├── interface Base
+│   │   ├──     fun print()
+│   │   ├── class BaseImpl(val x: Int) : Base
+│   │   ├──     override fun print()
+│   │   ├── internal class Derived(b: Base) : Base by b
+│   │   ├── class Person constructor(firstName: String)
+│   │   ├── class People(
+│   │   │       firstNames: Array<String>,
+│   │   │       ages: Array<Int>(42),
+│   │   │   )
+│   │   ├──     fun edgeCases(): Boolean
+│   │   ├── class Alien public @Inject constructor(
+│   │   │       val firstName: String,
+│   │   │       val lastName: String,
+│   │   │       var age: Int,
+│   │   │       val pets: MutableList<Pet> = mutableListOf(),
+│   │   │   )
+│   │   ├──     fun objectOriented(): String
+│   │   ├──  enum class IntArithmetics : BinaryOperator<Int>, IntBinaryOperator
+│   │   ├──     PLUS {
+│   │   │           override fun apply(t: Int, u: Int): Int
+│   │   ├──     TIMES {
+│   │   │           override fun apply(t: Int, u: Int): Int
+│   │   ├──     override fun applyAsInt(t: Int, u: Int)
+│   │   ├── fun reformat(
+│   │   │       str: String,
+│   │   │       normalizeCase: Boolean = true,
+│   │   │       upperCaseFirstLetter: Boolean = true,
+│   │   │       divideByCamelHumps: Boolean = false,
+│   │   │       wordSeparator: Char = ' ',
+│   │   │   )
+│   │   ├── operator fun Point.unaryMinus()
+│   │   ├── abstract class Polygon
+│   │   └──     abstract fun draw()
+│   ├── 📄 lesson.cbl (635 tokens, 78 lines)
+│   │   ├── IDENTIFICATION DIVISION.
+│   │   ├── PROGRAM-ID.    CBL0002.
+│   │   ├── AUTHOR.        Otto B. Fun.
+│   │   ├── ENVIRONMENT DIVISION.
+│   │   ├── INPUT-OUTPUT SECTION.
+│   │   ├── FILE-CONTROL.
+│   │   ├──     SELECT PRINT-LINE.
+│   │   ├──     SELECT ACCT-REC.
+│   │   ├── DATA DIVISION.
+│   │   ├── FILE SECTION.
+│   │   ├── FD  PRINT-LINE.
+│   │   ├── 01  PRINT-REC.
+│   │   ├──     05  ACCT-NO-O.
+│   │   ├──     05  ACCT-LIMIT-O.
+│   │   ├──     05  ACCT-BALANCE-O.
+│   │   ├──     05  LAST-NAME-O.
+│   │   ├──     05  FIRST-NAME-O.
+│   │   ├──     05  COMMENTS-O.
+│   │   ├── FD  ACCT-REC.
+│   │   ├── 01  ACCT-FIELDS.
+│   │   ├──     05  ACCT-NO.
+│   │   ├──     05  ACCT-LIMIT.
+│   │   ├──     05  ACCT-BALANCE.
+│   │   ├──     05  LAST-NAME.
+│   │   ├──     05  FIRST-NAME.
+│   │   ├──     05  CLIENT-ADDR.
+│   │   ├──         10  STREET-ADDR.
+│   │   ├──         10  CITY-COUNTY.
+│   │   ├──         10  USA-STATE.
+│   │   ├──     05  RESERVED.
+│   │   ├──     05  COMMENTS.
+│   │   ├── WORKING-STORAGE SECTION.
+│   │   ├── 01 FLAGS.
+│   │   ├──   05 LASTREC.
+│   │   ├── PROCEDURE DIVISION.
+│   │   ├── OPEN-FILES.
+│   │   ├── READ-NEXT-RECORD.
+│   │   ├── CLOSE-STOP.
+│   │   ├── READ-RECORD.
+│   │   └── WRITE-RECORD.
+│   ├── 📄 LuaTest.lua (83 tokens, 16 lines)
+│   │   ├── function HelloWorld.new
+│   │   ├── function HelloWorld.greet
+│   │   └── function say_hello
+│   ├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+│   │   ├── @interface HelloWorld
+│   │   ├── @interface HelloWorld -> (void) sayHello
+│   │   ├── @implementation HelloWorld
+│   │   ├── @implementation HelloWorld -> (void) sayHello
+│   │   └── void sayHelloWorld()
+│   ├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+│   │   ├── type color
+│   │   ├── class hello
+│   │   ├── class hello -> method say_hello
+│   │   └── let main ()
+│   ├── 📄 test.js (757 tokens, 154 lines)
+│   │   ├── class MyClass
+│   │   ├──   myMethod()
+│   │   ├──   async asyncMethod(a, b)
+│   │   ├──   methodWithDefaultParameters(a = 5, b = 10)
+│   │   ├──   multilineMethod(
+│   │   │       c,
+│   │   │       d
+│   │   │     )
+│   │   ├──   multilineMethodWithDefaults(
+│   │   │       t = "tree",
+│   │   │       p = "plus"
+│   │   │     )
+│   │   ├── function myFunction(param1, param2)
+│   │   ├── function multilineFunction(
+│   │   │     param1,
+│   │   │     param2
+│   │   │   )
+│   │   ├── const arrowFunction = () =>
+│   │   ├── const parametricArrow = (a, b) =>
+│   │   ├── function ()
+│   │   ├── function outerFunction(outerParam)
+│   │   ├──   function innerFunction(innerParam)
+│   │   ├──   innerFunction("inner")
+│   │   ├── const myObject = {
+│   │   ├──   myMethod: function (stuff)
+│   │   ├── let myArrowObject = {
+│   │   ├──   myArrow: ({
+│   │   │       a,
+│   │   │       b,
+│   │   │       c,
+│   │   │     }) =>
+│   │   ├── const myAsyncArrowFunction = async () =>
+│   │   ├── function functionWithRestParameters(...args)
+│   │   ├── const namedFunctionExpression = function myNamedFunction()
+│   │   ├── const multilineArrowFunction = (
+│   │   │     a,
+│   │   │     b
+│   │   │   ) =>
+│   │   ├── function functionReturningFunction()
+│   │   ├──   return function ()
+│   │   ├── function destructuringOnMultipleLines({
+│   │   │     a,
+│   │   │     b,
+│   │   │   })
+│   │   ├── const arrowFunctionWithDestructuring = ({ a, b }) =>
+│   │   ├── const multilineDestructuringArrow = ({
+│   │   │     a,
+│   │   │     b,
+│   │   │   }) =>
+│   │   ├── async function asyncFunctionWithErrorHandling()
+│   │   ├── class Car
+│   │   ├──   constructor(brand)
+│   │   ├──   present()
+│   │   ├── class Model extends Car
+│   │   ├──   constructor(brand, mod)
+│   │   ├──     super(brand)
+│   │   └──   show()
+│   └── 📄 test.ts (832 tokens, 165 lines)
+│       ├── type MyType
+│       ├── interface MyInterface
+│       ├── class TsClass
+│       ├──   myMethod()
+│       ├──   myMethodWithArgs(param1: string, param2: number): void
+│       ├──   static myStaticMethod<T>(param: T): T
+│       ├──   multilineMethod(
+│       │       c: number,
+│       │       d: number
+│       │     ): number
+│       ├──   multilineMethodWithDefaults(
+│       │       t: string = "tree",
+│       │       p: string = "plus"
+│       │     ): string
+│       ├── export class AdvancedComponent implements MyInterface
+│       ├──   async myAsyncMethod(
+│       │       a: string,
+│       │       b: number,
+│       │       c: string
+│       │     ): Promise<void>
+│       ├──   genericMethod<T, U>(
+│       │       arg1: T,
+│       │       arg2: U
+│       │     ): [T, U]
+│       ├── export class TicketsComponent implements MyInterface
+│       ├──   async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String
+│       │   })
+│       ├── function tsFunction()
+│       ├── function tsFunctionSigned(
+│       │     param1: number,
+│       │     param2: number
+│       │   ): void
+│       ├── export default async function tsFunctionComplicated<A, B, C>({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string>
+│       ├──   return("Standalone function with parameters")
+│       ├── const tsArrowFunctionSigned = ({
+│       │     a,
+│       │     b,
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │   }) =>
+│       ├── export const tsComplicatedArrow = async ({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string> =>
+│       ├── const arrowFunction = () =>
+│       ├── const arrow = (a: String, b: Number) =>
+│       ├── const asyncArrowFunction = async () =>
+│       ├── const asyncArrow = async (a: String, b: Number) =>
+│       ├── let weirdArrow = () =>
+│       ├── const asyncPromiseArrow = async (): Promise<void> =>
+│       ├── let myWeirdArrowSigned = (x: number): number =>
+│       ├── class Person
+│       ├──   constructor(private firstName: string, private lastName: string)
+│       ├──   getFullName(): string
+│       ├──   describe(): string
+│       ├── class Employee extends Person
+│       ├──   constructor(
+│       │       firstName: string,
+│       │       lastName: string,
+│       │       private jobTitle: string
+│       │     )
+│       ├──     super(firstName, lastName)
+│       ├──   describe(): string
+│       ├── interface Shape
+│       └── interface Square extends Shape
+└── 📁 path_to_test (1 folder, 6 files)
+    ├── 📄 class_method_type.py (525 tokens, 101 lines)
+    │   ├── T = TypeVar("T")
+    │   ├── def parse_py(contents: str) -> List[str]
+    │   ├── class MyClass
+    │   ├──     @staticmethod
+    │   │       def physical_element_aval(dtype) -> core.ShapedArray
+    │   ├──     def my_method(self)
+    │   ├──     @staticmethod
+    │   │       def my_typed_method(obj: dict) -> int
+    │   ├──     def my_multiline_signature_method(
+    │   │           self,
+    │   │           alice: str = None,
+    │   │           bob: int = None,
+    │   │       ) -> tuple
+    │   ├── @lru_cache(maxsize=None)
+    │   │   def my_multiline_signature_function(
+    │   │       tree: tuple = (),
+    │   │       plus: str = "+",
+    │   │   ) -> tuple
+    │   ├── class LogLevelEnum(str, Enum)
+    │   ├──     CRITICAL = "CRITICAL"
+    │   ├──     GREETING = "GREETING"
+    │   ├──     WARNING = "WARNING"
+    │   ├──     ERROR = "ERROR"
+    │   ├──     DEBUG = "DEBUG"
+    │   ├──     INFO = "INFO"
+    │   ├──     OFF = "OFF"
+    │   ├── class Thingy(BaseModel)
+    │   ├──     metric: float
+    │   ├── @dataclass
+    │   │   class TestDataclass
+    │   ├──     tree: str
+    │   ├── A = TypeVar("A", str, bytes)
+    │   ├── def omega_yikes(file: str, expected: List[str]) -> bool
+    │   ├── def ice[T](args: Iterable[T] = ())
+    │   ├── class list[T]
+    │   ├──     def __getitem__(self, index: int, /) -> T
+    │   ├──     @classmethod
+    │   │       def from_code(cls, toolbox, code: bytes, score=None) -> "Thingy"
+    │   ├──     @classmethod
+    │   │       def from_str(cls, toolbox, string: str, score=None) -> "Thingy"
+    │   └── class Router(hk.Module)
+    ├── 📄 empty.py (0 tokens, 0 lines)
+    ├── 📄 file.md (11 tokens, 2 lines)
+    │   └── # Hello, world!
+    ├── 📄 file.py (18 tokens, 3 lines)
+    │   └── def hello_world()
+    ├── 📄 file.txt (10 tokens, 2 lines)
+    └── 📄 version.py (13 tokens, 2 lines)
+        └── __version__ = "1.2.3"
diff --git a/tests/golden/legacy/trees/path_to_test.txt b/tests/golden/legacy/trees/path_to_test.txt
new file mode 100644
index 0000000..c1f5bc7
--- /dev/null
+++ b/tests/golden/legacy/trees/path_to_test.txt
@@ -0,0 +1,51 @@
+📁 path_to_test (1 folder, 6 files)
+├── 📄 class_method_type.py (525 tokens, 101 lines)
+│   ├── T = TypeVar("T")
+│   ├── def parse_py(contents: str) -> List[str]
+│   ├── class MyClass
+│   ├──     @staticmethod
+│   │       def physical_element_aval(dtype) -> core.ShapedArray
+│   ├──     def my_method(self)
+│   ├──     @staticmethod
+│   │       def my_typed_method(obj: dict) -> int
+│   ├──     def my_multiline_signature_method(
+│   │           self,
+│   │           alice: str = None,
+│   │           bob: int = None,
+│   │       ) -> tuple
+│   ├── @lru_cache(maxsize=None)
+│   │   def my_multiline_signature_function(
+│   │       tree: tuple = (),
+│   │       plus: str = "+",
+│   │   ) -> tuple
+│   ├── class LogLevelEnum(str, Enum)
+│   ├──     CRITICAL = "CRITICAL"
+│   ├──     GREETING = "GREETING"
+│   ├──     WARNING = "WARNING"
+│   ├──     ERROR = "ERROR"
+│   ├──     DEBUG = "DEBUG"
+│   ├──     INFO = "INFO"
+│   ├──     OFF = "OFF"
+│   ├── class Thingy(BaseModel)
+│   ├──     metric: float
+│   ├── @dataclass
+│   │   class TestDataclass
+│   ├──     tree: str
+│   ├── A = TypeVar("A", str, bytes)
+│   ├── def omega_yikes(file: str, expected: List[str]) -> bool
+│   ├── def ice[T](args: Iterable[T] = ())
+│   ├── class list[T]
+│   ├──     def __getitem__(self, index: int, /) -> T
+│   ├──     @classmethod
+│   │       def from_code(cls, toolbox, code: bytes, score=None) -> "Thingy"
+│   ├──     @classmethod
+│   │       def from_str(cls, toolbox, string: str, score=None) -> "Thingy"
+│   └── class Router(hk.Module)
+├── 📄 empty.py (0 tokens, 0 lines)
+├── 📄 file.md (11 tokens, 2 lines)
+│   └── # Hello, world!
+├── 📄 file.py (18 tokens, 3 lines)
+│   └── def hello_world()
+├── 📄 file.txt (10 tokens, 2 lines)
+└── 📄 version.py (13 tokens, 2 lines)
+    └── __version__ = "1.2.3"
diff --git a/tests/golden/legacy/trees/repo_concise.txt b/tests/golden/legacy/trees/repo_concise.txt
new file mode 100644
index 0000000..17182cb
--- /dev/null
+++ b/tests/golden/legacy/trees/repo_concise.txt
@@ -0,0 +1,699 @@
+📁 tree_plus (44 folders, 439 files)
+├── 📄 .env.test (4 tokens, 0 lines)
+├── 📁 .github (2 folders, 3 files)
+│   ├── 📄 dependabot.yml (128 tokens, 11 lines)
+│   └── 📁 workflows (1 folder, 2 files)
+│       ├── 📄 microsoft.yml (284 tokens, 40 lines)
+│       └── 📄 unix.yml (713 tokens, 92 lines)
+├── 📄 .gitignore (219 tokens, 57 lines)
+├── 📄 .mcp_server.pid (2 tokens, 1 line)
+├── 📄 Cargo.toml (206 tokens, 29 lines)
+├── 📄 claude-fable-5-rust-rewrite-goal.md (3,394 tokens, 434 lines)
+├── 📁 coverage (1 folder, 1 file)
+│   └── 📄 lcov.info (17,359 tokens, 2,180 lines)
+├── 📁 crates (11 folders, 27 files)
+│   ├── 📁 tree_plus_cli (3 folders, 3 files)
+│   │   ├── 📄 Cargo.toml (86 tokens, 15 lines)
+│   │   ├── 📁 src (1 folder, 1 file)
+│   │   │   └── 📄 main.rs (1,332 tokens, 174 lines)
+│   │   └── 📁 tests (1 folder, 1 file)
+│   │       └── 📄 cli.rs (701 tokens, 92 lines)
+│   └── 📁 tree_plus_core (7 folders, 24 files)
+│       ├── 📁 benches (1 folder, 1 file)
+│       │   └── 📄 tree_plus_bench.rs (608 tokens, 78 lines)
+│       ├── 📄 Cargo.toml (228 tokens, 36 lines)
+│       ├── 📁 examples (1 folder, 2 files)
+│       │   ├── 📄 dump_ast.rs (516 tokens, 55 lines)
+│       │   └── 📄 extract.rs (129 tokens, 16 lines)
+│       ├── 📁 src (3 folders, 18 files)
+│       │   ├── 📄 config.rs (304 tokens, 39 lines)
+│       │   ├── 📄 count.rs (1,346 tokens, 203 lines)
+│       │   ├── 📁 extract (2 folders, 10 files)
+│       │   │   ├── 📄 data.rs (5,115 tokens, 582 lines)
+│       │   │   ├── 📄 markdown.rs (1,531 tokens, 180 lines)
+│       │   │   ├── 📄 markers.rs (438 tokens, 60 lines)
+│       │   │   ├── 📄 mod.rs (2,520 tokens, 277 lines)
+│       │   │   ├── 📄 simple.rs (1,629 tokens, 216 lines)
+│       │   │   └── 📁 treesitter (1 folder, 5 files)
+│       │   │       ├── 📄 c_cpp.rs (5,979 tokens, 591 lines)
+│       │   │       ├── 📄 mod.rs (488 tokens, 66 lines)
+│       │   │       ├── 📄 python.rs (3,487 tokens, 346 lines)
+│       │   │       ├── 📄 rust.rs (2,785 tokens, 312 lines)
+│       │   │       └── 📄 typescript.rs (3,897 tokens, 420 lines)
+│       │   ├── 📄 ignore.rs (2,144 tokens, 307 lines)
+│       │   ├── 📄 lib.rs (222 tokens, 29 lines)
+│       │   ├── 📄 model.rs (928 tokens, 125 lines)
+│       │   ├── 📄 render.rs (2,741 tokens, 347 lines)
+│       │   ├── 📄 sort.rs (1,693 tokens, 214 lines)
+│       │   └── 📄 walk.rs (2,245 tokens, 260 lines)
+│       └── 📁 tests (1 folder, 2 files)
+│           ├── 📄 golden_parity.rs (1,809 tokens, 243 lines)
+│           └── 📄 robustness.rs (743 tokens, 86 lines)
+├── 📁 docs (1 folder, 4 files)
+│   ├── 📄 architecture.md (1,392 tokens, 113 lines)
+│   ├── 📄 language-roadmap.md (1,262 tokens, 64 lines)
+│   ├── 📄 performance.md (690 tokens, 64 lines)
+│   └── 📄 rust-port-differences.md (922 tokens, 67 lines)
+├── 📄 LICENSE (2,744 tokens, 81 lines)
+├── 📄 Makefile (801 tokens, 121 lines)
+├── 📄 nodemon.json (112 tokens, 24 lines)
+├── 📄 pyproject.toml (366 tokens, 51 lines)
+├── 📄 pytest.ini (20 tokens, 4 lines)
+├── 📄 README.md (99,851 tokens, 3,708 lines)
+├── 📁 tests (25 folders, 378 files)
+│   ├── 📄 .env.test (4 tokens, 0 lines)
+│   ├── 📄 build_absurdly_huge_jsonl.py (506 tokens, 65 lines)
+│   ├── 📁 dot_dot (2 folders, 4 files)
+│   │   ├── 📄 my_test_file.py (7 tokens, 2 lines)
+│   │   └── 📁 nested_dir (1 folder, 3 files)
+│   │       ├── 📄 .env.test (4 tokens, 0 lines)
+│   │       ├── 📄 pytest.ini (20 tokens, 4 lines)
+│   │       └── 📄 test_tp_dotdot.py (362 tokens, 52 lines)
+│   ├── 📁 empty_folder (2 folders, 0 files)
+│   │   └── 📁 is_empty (1 folder, 0 files)
+│   ├── 📁 folder_with_evil_logging (1 folder, 1 file)
+│   │   └── 📄 logging.py (11 tokens, 0 lines)
+│   ├── 📁 golden (6 folders, 246 files)
+│   │   ├── 📄 diff_components.py (417 tokens, 59 lines)
+│   │   ├── 📄 generate_legacy_goldens.py (1,499 tokens, 156 lines)
+│   │   └── 📁 legacy (5 folders, 244 files)
+│   │       ├── 📁 components (1 folder, 109 files)
+│   │       │   ├── 📄 tests__dot_dot__my_test_file.py.json (21 tokens, 5 lines)
+│   │       │   ├── 📄 tests__dot_dot__nested_dir__.env.test.json (22 tokens, 5
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__dot_dot__nested_dir__pytest.ini.json (17 tokens, 3
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__dot_dot__nested_dir__test_tp_dotdot.py.json (40
+│   │       │   │   tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__addamt.cobol.json (110
+│   │       │   │   tokens, 19 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
+│   │       │   │   (251 tokens, 39 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__JavaTest.java.json (236
+│   │       │   │   tokens, 28 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__JuliaTest.jl.json (189
+│   │       │   │   tokens, 16 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__KotlinTest.kt.json (533
+│   │       │   │   tokens, 51 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__lesson.cbl.json (259
+│   │       │   │   tokens, 44 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__LuaTest.lua.json (39
+│   │       │   │   tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__ObjectiveCTest.m.json (65
+│   │       │   │   tokens, 9 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__OcamlTest.ml.json (40
+│   │       │   │   tokens, 8 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__test.js.json (355 tokens,
+│   │       │   │   39 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__test.ts.json (541 tokens,
+│   │       │   │   40 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__apl_test.apl.json (49
+│   │       │   │   tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__c_test.c.json (290 tokens,
+│   │       │   │   38 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__go_test.go.json (111
+│   │       │   │   tokens, 11 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__PerlTest.pl.json (50
+│   │       │   │   tokens, 8 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__PhpTest.php.json (54
+│   │       │   │   tokens, 9 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__PowershellTest.ps1.json
+│   │       │   │   (315 tokens, 26 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__ScalaTest.scala.json (134
+│   │       │   │   tokens, 15 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__test.csv.json (31 tokens,
+│   │       │   │   9 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__bash_test.sh.json (55
+│   │       │   │   tokens, 10 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__cpp_test.cpp.json (591
+│   │       │   │   tokens, 73 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__csharp_test.cs.json (592
+│   │       │   │   tokens, 47 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__hallucination.tex.json
+│   │       │   │   (299 tokens, 29 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__ruby_test.rb.json (96
+│   │       │   │   tokens, 13 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__swift_test.swift.json (200
+│   │       │   │   tokens, 22 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.lean.json (151
+│   │       │   │   tokens, 14 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.capnp.json (124
+│   │       │   │   tokens, 22 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.graphql.json (102
+│   │       │   │   tokens, 18 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.proto.json (159
+│   │       │   │   tokens, 22 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.sqlite.json (73
+│   │       │   │   tokens, 12 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_Cargo.toml.json (69
+│   │       │   │   tokens, 11 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group3__test_json_rpc_2_0.json.json
+│   │       │   │   (48 tokens, 10 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_openapi.yaml.json
+│   │       │   │   (120 tokens, 15 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_openrpc.json.json (78
+│   │       │   │   tokens, 13 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_pyproject.toml.json
+│   │       │   │   (99 tokens, 15 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__erl_test.erl.json (212
+│   │       │   │   tokens, 19 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__haskell_test.hs.json (121
+│   │       │   │   tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__mathematica_test.nb.json
+│   │       │   │   (35 tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__matlab_test.m.json (35
+│   │       │   │   tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__RTest.R.json (50 tokens, 8
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group4__rust_test.rs.json (775
+│   │       │   │   tokens, 49 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__test.zig.json (139 tokens,
+│   │       │   │   10 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__test_fsharp.fs.json (80
+│   │       │   │   tokens, 10 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__test_tcl_tk.tcl.json (41
+│   │       │   │   tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__tf_test.tf.json (67
+│   │       │   │   tokens, 11 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__ansible_test.yml.json (34
+│   │       │   │   tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app-routing.module.ts.json
+│   │       │   │   (157 tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app.component.ts.json (104
+│   │       │   │   tokens, 10 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app.component.spec.ts.json
+│   │       │   │   (67 tokens, 9 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app.module.ts.json (87
+│   │       │   │   tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__checkbox_test.md.json (146
+│   │       │   │   tokens, 22 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__checkbox_test.txt.json
+│   │       │   │   (104 tokens, 12 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__environment.test.ts.json
+│   │       │   │   (46 tokens, 9 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__hello_world.pyi.json (44
+│   │       │   │   tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__k8s_test.yaml.json (42
+│   │       │   │   tokens, 7 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__Makefile.json (62 tokens,
+│   │       │   │   13 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__requirements_test.txt.json
+│   │       │   │   (59 tokens, 13 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__rust_todo_test.rs.json (75
+│   │       │   │   tokens, 12 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__sql_test.sql.json (190
+│   │       │   │   tokens, 26 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group5__standard-app-routing.module.t
+│   │       │   │   s.json (106 tokens, 5 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__test.env.json (98 tokens,
+│   │       │   │   24 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__testJsonSchema.json.json
+│   │       │   │   (59 tokens, 8 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__testPackage.json.json (74
+│   │       │   │   tokens, 12 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__tickets.component.ts.json
+│   │       │   │   (647 tokens, 53 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__catastrophic.c.json (1,677
+│   │       │   │   tokens, 195 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__cpp_examples_impl.cc.json
+│   │       │   │   (50 tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__cpp_examples_impl.cu.json
+│   │       │   │   (43 tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__cpp_examples_impl.h.json
+│   │       │   │   (41 tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__edge_case.hpp.json (18
+│   │       │   │   tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__fractal.thy.json (816
+│   │       │   │   tokens, 29 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group6__Microsoft.PowerShell_profile.
+│   │       │   │   ps1.json (354 tokens, 34 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group6__python_complex_class.py.json
+│   │       │   │   (30 tokens, 5 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__ramda__cloneRegExp.js.json
+│   │       │   │   (33 tokens, 5 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__ramda_prop.js.json (317
+│   │       │   │   tokens, 8 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__tensorflow_flags.h.json
+│   │       │   │   (2,584 tokens, 101 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__test.f.json (112 tokens,
+│   │       │   │   10 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__torch.rst.json (34 tokens,
+│   │       │   │   6 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__yc.html.json (17 tokens, 3
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group7__absurdly_huge.jsonl.json
+│   │       │   │   (68 tokens, 13 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__angular_crud.ts.json (217
+│   │       │   │   tokens, 17 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__structure.py.json (338
+│   │       │   │   tokens, 34 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__test.wgsl.json (288
+│   │       │   │   tokens, 19 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__test.metal.json (135
+│   │       │   │   tokens, 10 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__clojure_test.clj.json
+│   │       │   │   (63 tokens, 12 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__LispTest.lisp.json (29
+│   │       │   │   tokens, 6 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__racket_struct.rkt.json
+│   │       │   │   (25 tokens, 5 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__test_scheme.scm.json
+│   │       │   │   (54 tokens, 10 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__AAPLShaders.metal.json
+│   │       │   │   (1,202 tokens, 29 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__crystal_test.cr.json
+│   │       │   │   (20 tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__dart_test.dart.json
+│   │       │   │   (20 tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__elixir_test.exs.json
+│   │       │   │   (20 tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__forward.frag.json (19
+│   │       │   │   tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__forward.vert.json (19
+│   │       │   │   tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__nodemon.json.json (19
+│   │       │   │   tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__sas_test.sas.json (19
+│   │       │   │   tokens, 3 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group_todo__test_setup_py.test.json
+│   │       │   │   (21 tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__testTypings.d.ts.json
+│   │       │   │   (20 tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__vba_test.bas.json (19
+│   │       │   │   tokens, 3 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__wgsl_test.wgsl.json
+│   │       │   │   (70 tokens, 7 lines)
+│   │       │   ├── 📄 tests__path_to_test__class_method_type.py.json (340
+│   │       │   │   tokens, 32 lines)
+│   │       │   ├── 📄 tests__path_to_test__empty.py.json (15 tokens, 3 lines)
+│   │       │   ├── 📄 tests__path_to_test__file.md.json (20 tokens, 5 lines)
+│   │       │   ├── 📄 tests__path_to_test__file.py.json (21 tokens, 5 lines)
+│   │       │   ├── 📄 tests__path_to_test__file.txt.json (15 tokens, 3 lines)
+│   │       │   └── 📄 tests__path_to_test__version.py.json (23 tokens, 5 lines)
+│   │       ├── 📁 counts (1 folder, 109 files)
+│   │       │   ├── 📄 tests__dot_dot__my_test_file.py.json (20 tokens, 0 lines)
+│   │       │   ├── 📄 tests__dot_dot__nested_dir__.env.test.json (21 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__dot_dot__nested_dir__pytest.ini.json (22 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__dot_dot__nested_dir__test_tp_dotdot.py.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__addamt.cobol.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__CUSTOMER-INVOICE.CBL.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__JavaTest.java.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__JuliaTest.jl.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__KotlinTest.kt.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__lesson.cbl.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__LuaTest.lua.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__ObjectiveCTest.m.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__OcamlTest.ml.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group1__test.js.json (22 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group1__test.ts.json (22 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group2__apl_test.apl.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__c_test.c.json (23 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__go_test.go.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__PerlTest.pl.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__PhpTest.php.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__PowershellTest.ps1.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__ScalaTest.scala.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group2__test.csv.json (15 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__bash_test.sh.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__cpp_test.cpp.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__csharp_test.cs.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__hallucination.tex.json (25
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__ruby_test.rb.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__swift_test.swift.json (25
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.lean.json (23 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.capnp.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.graphql.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.proto.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test.sqlite.json (16
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_Cargo.toml.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group3__test_json_rpc_2_0.json.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_openapi.yaml.json (25
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_openrpc.json.json (25
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group3__test_pyproject.toml.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__erl_test.erl.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__haskell_test.hs.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__mathematica_test.nb.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__matlab_test.m.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__RTest.R.json (22 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group4__rust_test.rs.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__test.zig.json (22 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__test_fsharp.fs.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__test_tcl_tk.tcl.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group4__tf_test.tf.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__ansible_test.yml.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app-routing.module.ts.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app.component.ts.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app.component.spec.ts.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__app.module.ts.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__checkbox_test.md.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__checkbox_test.txt.json (25
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__environment.test.ts.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__hello_world.pyi.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__k8s_test.yaml.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__Makefile.json (22 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__requirements_test.txt.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__rust_todo_test.rs.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__sql_test.sql.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group5__standard-app-routing.module.t
+│   │       │   │   s.json (28 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__test.env.json (22 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__testJsonSchema.json.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__testPackage.json.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group5__tickets.component.ts.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__catastrophic.c.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__cpp_examples_impl.cc.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__cpp_examples_impl.cu.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__cpp_examples_impl.h.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__edge_case.hpp.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__fractal.thy.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group6__Microsoft.PowerShell_profile.
+│   │       │   │   ps1.json (29 tokens, 0 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group6__python_complex_class.py.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__ramda__cloneRegExp.js.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__ramda_prop.js.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__tensorflow_flags.h.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__test.f.json (22 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group6__torch.rst.json (22 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group6__yc.html.json (23 tokens, 0
+│   │       │   │   lines)
+│   │       │   ├── 📄 tests__more_languages__group7__absurdly_huge.jsonl.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__angular_crud.ts.json (25
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__structure.py.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__test.wgsl.json (23 tokens,
+│   │       │   │   0 lines)
+│   │       │   ├── 📄 tests__more_languages__group7__test.metal.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__clojure_test.clj.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__LispTest.lisp.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__racket_struct.rkt.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_lisp__test_scheme.scm.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__AAPLShaders.metal.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__crystal_test.cr.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__dart_test.dart.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__elixir_test.exs.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__forward.frag.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__forward.vert.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__nodemon.json.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__sas_test.sas.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄
+│   │       │   │   tests__more_languages__group_todo__test_setup_py.test.json
+│   │       │   │   (26 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__testTypings.d.ts.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__vba_test.bas.json (24
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__more_languages__group_todo__wgsl_test.wgsl.json
+│   │       │   │   (25 tokens, 0 lines)
+│   │       │   ├── 📄 tests__path_to_test__class_method_type.py.json (23
+│   │       │   │   tokens, 0 lines)
+│   │       │   ├── 📄 tests__path_to_test__empty.py.json (19 tokens, 0 lines)
+│   │       │   ├── 📄 tests__path_to_test__file.md.json (19 tokens, 0 lines)
+│   │       │   ├── 📄 tests__path_to_test__file.py.json (19 tokens, 0 lines)
+│   │       │   ├── 📄 tests__path_to_test__file.txt.json (20 tokens, 0 lines)
+│   │       │   └── 📄 tests__path_to_test__version.py.json (20 tokens, 0 lines)
+│   │       ├── 📁 trees (1 folder, 13 files)
+│   │       │   ├── 📄 dot_dot.txt (99 tokens, 10 lines)
+│   │       │   ├── 📄 more_languages.txt (22,977 tokens, 2,225 lines)
+│   │       │   ├── 📄 more_languages_group1.txt (2,909 tokens, 374 lines)
+│   │       │   ├── 📄 more_languages_group2.txt (1,111 tokens, 135 lines)
+│   │       │   ├── 📄 more_languages_group3.txt (3,035 tokens, 346 lines)
+│   │       │   ├── 📄 more_languages_group4.txt (1,737 tokens, 214 lines)
+│   │       │   ├── 📄 more_languages_group5.txt (2,288 tokens, 257 lines)
+│   │       │   ├── 📄 more_languages_group6.txt (6,838 tokens, 615 lines)
+│   │       │   ├── 📄 more_languages_group7.txt (1,148 tokens, 126 lines)
+│   │       │   ├── 📄 more_languages_group_lisp.txt (155 tokens, 22 lines)
+│   │       │   ├── 📄 more_languages_group_todo.txt (1,491 tokens, 111 lines)
+│   │       │   ├── 📄 multi_seed.txt (3,789 tokens, 427 lines)
+│   │       │   └── 📄 path_to_test.txt (445 tokens, 51 lines)
+│   │       └── 📁 trees_v1 (1 folder, 13 files)
+│   │           ├── 📄 dot_dot.txt (99 tokens, 10 lines)
+│   │           ├── 📄 more_languages.txt (10,782 tokens, 1,122 lines)
+│   │           ├── 📄 more_languages_group1.txt (1,130 tokens, 146 lines)
+│   │           ├── 📄 more_languages_group2.txt (415 tokens, 54 lines)
+│   │           ├── 📄 more_languages_group3.txt (1,240 tokens, 149 lines)
+│   │           ├── 📄 more_languages_group4.txt (1,020 tokens, 123 lines)
+│   │           ├── 📄 more_languages_group5.txt (2,099 tokens, 235 lines)
+│   │           ├── 📄 more_languages_group6.txt (2,821 tokens, 300 lines)
+│   │           ├── 📄 more_languages_group7.txt (705 tokens, 83 lines)
+│   │           ├── 📄 more_languages_group_lisp.txt (52 tokens, 5 lines)
+│   │           ├── 📄 more_languages_group_todo.txt (140 tokens, 13 lines)
+│   │           ├── 📄 multi_seed.txt (1,783 tokens, 199 lines)
+│   │           └── 📄 path_to_test.txt (445 tokens, 51 lines)
+│   ├── 📁 more_languages (10 folders, 99 files)
+│   │   ├── 📁 group1 (1 folder, 11 files)
+│   │   │   ├── 📄 addamt.cobol (441 tokens, 40 lines)
+│   │   │   ├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+│   │   │   ├── 📄 JavaTest.java (578 tokens, 86 lines)
+│   │   │   ├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+│   │   │   ├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+│   │   │   ├── 📄 lesson.cbl (635 tokens, 78 lines)
+│   │   │   ├── 📄 LuaTest.lua (83 tokens, 16 lines)
+│   │   │   ├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+│   │   │   ├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+│   │   │   ├── 📄 test.js (757 tokens, 154 lines)
+│   │   │   └── 📄 test.ts (832 tokens, 165 lines)
+│   │   ├── 📁 group2 (1 folder, 8 files)
+│   │   │   ├── 📄 apl_test.apl (28 tokens, 5 lines)
+│   │   │   ├── 📄 c_test.c (837 tokens, 142 lines)
+│   │   │   ├── 📄 go_test.go (179 tokens, 46 lines)
+│   │   │   ├── 📄 PerlTest.pl (63 tokens, 20 lines)
+│   │   │   ├── 📄 PhpTest.php (70 tokens, 19 lines)
+│   │   │   ├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)
+│   │   │   ├── 📄 ScalaTest.scala (171 tokens, 40 lines)
+│   │   │   └── 📄 test.csv (0 tokens, 0 lines)
+│   │   ├── 📁 group3 (1 folder, 16 files)
+│   │   │   ├── 📄 bash_test.sh (127 tokens, 22 lines)
+│   │   │   ├── 📄 cpp_test.cpp (1,670 tokens, 259 lines)
+│   │   │   ├── 📄 csharp_test.cs (957 tokens, 146 lines)
+│   │   │   ├── 📄 hallucination.tex (1,633 tokens, 126 lines)
+│   │   │   ├── 📄 ruby_test.rb (138 tokens, 37 lines)
+│   │   │   ├── 📄 swift_test.swift (469 tokens, 110 lines)
+│   │   │   ├── 📄 test.lean (289 tokens, 42 lines)
+│   │   │   ├── 📄 test.capnp (117 tokens, 30 lines)
+│   │   │   ├── 📄 test.graphql (66 tokens, 21 lines)
+│   │   │   ├── 📄 test.proto (142 tokens, 34 lines)
+│   │   │   ├── 📄 test.sqlite (0 tokens, 0 lines)
+│   │   │   ├── 📄 test_Cargo.toml (119 tokens, 18 lines)
+│   │   │   ├── 📄 test_json_rpc_2_0.json (26 tokens, 6 lines)
+│   │   │   ├── 📄 test_openapi.yaml (753 tokens, 92 lines)
+│   │   │   ├── 📄 test_openrpc.json (225 tokens, 44 lines)
+│   │   │   └── 📄 test_pyproject.toml (304 tokens, 39 lines)
+│   │   ├── 📁 group4 (1 folder, 10 files)
+│   │   │   ├── 📄 erl_test.erl (480 tokens, 68 lines)
+│   │   │   ├── 📄 haskell_test.hs (414 tokens, 41 lines)
+│   │   │   ├── 📄 mathematica_test.nb (133 tokens, 21 lines)
+│   │   │   ├── 📄 matlab_test.m (48 tokens, 12 lines)
+│   │   │   ├── 📄 RTest.R (367 tokens, 46 lines)
+│   │   │   ├── 📄 rust_test.rs (1,368 tokens, 259 lines)
+│   │   │   ├── 📄 test.zig (397 tokens, 60 lines)
+│   │   │   ├── 📄 test_fsharp.fs (92 tokens, 27 lines)
+│   │   │   ├── 📄 test_tcl_tk.tcl (54 tokens, 16 lines)
+│   │   │   └── 📄 tf_test.tf (202 tokens, 38 lines)
+│   │   ├── 📁 group5 (1 folder, 19 files)
+│   │   │   ├── 📄 ansible_test.yml (55 tokens, 14 lines)
+│   │   │   ├── 📄 app-routing.module.ts (287 tokens, 28 lines)
+│   │   │   ├── 📄 app.component.spec.ts (410 tokens, 47 lines)
+│   │   │   ├── 📄 app.component.ts (271 tokens, 45 lines)
+│   │   │   ├── 📄 app.module.ts (374 tokens, 43 lines)
+│   │   │   ├── 📄 checkbox_test.md (191 tokens, 29 lines)
+│   │   │   ├── 📄 checkbox_test.txt (257 tokens, 33 lines)
+│   │   │   ├── 📄 environment.test.ts (197 tokens, 19 lines)
+│   │   │   ├── 📄 hello_world.pyi (22 tokens, 3 lines)
+│   │   │   ├── 📄 k8s_test.yaml (140 tokens, 37 lines)
+│   │   │   ├── 📄 Makefile (714 tokens, 84 lines)
+│   │   │   ├── 📄 requirements_test.txt (29 tokens, 10 lines)
+│   │   │   ├── 📄 rust_todo_test.rs (92 tokens, 26 lines)
+│   │   │   ├── 📄 sql_test.sql (270 tokens, 51 lines)
+│   │   │   ├── 📄 standard-app-routing.module.ts (100 tokens, 16 lines)
+│   │   │   ├── 📄 test.env (190 tokens, 25 lines)
+│   │   │   ├── 📄 testJsonSchema.json (421 tokens, 48 lines)
+│   │   │   ├── 📄 testPackage.json (349 tokens, 43 lines)
+│   │   │   └── 📄 tickets.component.ts (7,160 tokens, 903 lines)
+│   │   ├── 📁 group6 (1 folder, 14 files)
+│   │   │   ├── 📄 catastrophic.c (5,339 tokens, 754 lines)
+│   │   │   ├── 📄 cpp_examples_impl.cc (60 tokens, 10 lines)
+│   │   │   ├── 📄 cpp_examples_impl.cu (37 tokens, 10 lines)
+│   │   │   ├── 📄 cpp_examples_impl.h (22 tokens, 6 lines)
+│   │   │   ├── 📄 edge_case.hpp (426 tokens, 28 lines)
+│   │   │   ├── 📄 fractal.thy (1,712 tokens, 147 lines)
+│   │   │   ├── 📄 Microsoft.PowerShell_profile.ps1 (3,346 tokens, 497 lines)
+│   │   │   ├── 📄 python_complex_class.py (10 tokens, 2 lines)
+│   │   │   ├── 📄 ramda__cloneRegExp.js (173 tokens, 9 lines)
+│   │   │   ├── 📄 ramda_prop.js (646 tokens, 85 lines)
+│   │   │   ├── 📄 tensorflow_flags.h (7,628 tokens, 668 lines)
+│   │   │   ├── 📄 test.f (181 tokens, 30 lines)
+│   │   │   ├── 📄 torch.rst (60 tokens, 8 lines)
+│   │   │   └── 📄 yc.html (9,063 tokens, 169 lines)
+│   │   ├── 📁 group7 (1 folder, 5 files)
+│   │   │   ├── 📄 absurdly_huge.jsonl (8,347 tokens, 126 lines)
+│   │   │   ├── 📄 angular_crud.ts (1,192 tokens, 148 lines)
+│   │   │   ├── 📄 structure.py (400 tokens, 92 lines)
+│   │   │   ├── 📄 test.wgsl (528 tokens, 87 lines)
+│   │   │   └── 📄 test.metal (272 tokens, 34 lines)
+│   │   ├── 📁 group_lisp (1 folder, 4 files)
+│   │   │   ├── 📄 clojure_test.clj (682 tokens, 85 lines)
+│   │   │   ├── 📄 LispTest.lisp (25 tokens, 6 lines)
+│   │   │   ├── 📄 racket_struct.rkt (14 tokens, 1 line)
+│   │   │   └── 📄 test_scheme.scm (360 tokens, 44 lines)
+│   │   └── 📁 group_todo (1 folder, 12 files)
+│   │       ├── 📄 AAPLShaders.metal (5,780 tokens, 566 lines)
+│   │       ├── 📄 crystal_test.cr (48 tokens, 15 lines)
+│   │       ├── 📄 dart_test.dart (108 tokens, 24 lines)
+│   │       ├── 📄 elixir_test.exs (39 tokens, 10 lines)
+│   │       ├── 📄 forward.frag (739 tokens, 87 lines)
+│   │       ├── 📄 forward.vert (359 tokens, 48 lines)
+│   │       ├── 📄 nodemon.json (118 tokens, 20 lines)
+│   │       ├── 📄 sas_test.sas (97 tokens, 22 lines)
+│   │       ├── 📄 test_setup_py.test (133 tokens, 24 lines)
+│   │       ├── 📄 testTypings.d.ts (158 tokens, 23 lines)
+│   │       ├── 📄 vba_test.bas (67 tokens, 16 lines)
+│   │       └── 📄 wgsl_test.wgsl (94 tokens, 17 lines)
+│   ├── 📁 path_to_test (1 folder, 6 files)
+│   │   ├── 📄 class_method_type.py (525 tokens, 101 lines)
+│   │   ├── 📄 empty.py (0 tokens, 0 lines)
+│   │   ├── 📄 file.md (11 tokens, 2 lines)
+│   │   ├── 📄 file.py (18 tokens, 3 lines)
+│   │   ├── 📄 file.txt (10 tokens, 2 lines)
+│   │   └── 📄 version.py (13 tokens, 2 lines)
+│   ├── 📄 pytest.ini (20 tokens, 4 lines)
+│   ├── 📁 readme_updates (1 folder, 6 files)
+│   │   ├── 📄 .gitkeep (0 tokens, 0 lines)
+│   │   ├── 📄 medium_README_sink.md (37,636 tokens, 3,733 lines)
+│   │   ├── 📄 medium_README_source.md (149 tokens, 48 lines)
+│   │   ├── 📄 mini_README_sink.md (7,833 tokens, 785 lines)
+│   │   ├── 📄 mini_README_source.md (27 tokens, 8 lines)
+│   │   └── 📄 renamed_dry_run_README.md (38,153 tokens, 3,705 lines)
+│   ├── 📄 tensorflow_expectation.py (2,554 tokens, 97 lines)
+│   ├── 📄 test_cli.py (2,218 tokens, 252 lines)
+│   ├── 📄 test_deploy.py (2,045 tokens, 225 lines)
+│   ├── 📄 test_dotenv.py (45 tokens, 10 lines)
+│   ├── 📄 test_e2e.py (3,630 tokens, 429 lines)
+│   ├── 📄 test_engine.py (3,869 tokens, 424 lines)
+│   ├── 📄 test_more_language_units.py (24,414 tokens, 2,559 lines)
+│   ├── 📄 test_programs.py (509 tokens, 88 lines)
+│   ├── 📄 test_units.py (1,914 tokens, 261 lines)
+│   ├── 📄 test_web.py (42 tokens, 6 lines)
+│   └── 📁 version_increments (1 folder, 3 files)
+│       ├── 📄 renamed_dry_run_version.py (12 tokens, 1 line)
+│       ├── 📄 test_sink_version.py (12 tokens, 1 line)
+│       └── 📄 test_source_version.py (12 tokens, 2 lines)
+├── 📄 tree_plus_cli.py (2,283 tokens, 331 lines)
+├── 📁 tree_plus_programs (1 folder, 4 files)
+│   ├── 📄 hello_tree_plus.py (545 tokens, 80 lines)
+│   ├── 📄 rewrite.py (4,017 tokens, 471 lines)
+│   ├── 📄 stub_tests.py (1,348 tokens, 180 lines)
+│   └── 📄 test_stub_tests.py (79 tokens, 20 lines)
+└── 📁 tree_plus_src (2 folders, 10 files)
+    ├── 📄 count_tokens_lines.py (1,323 tokens, 209 lines)
+    ├── 📄 debug.py (186 tokens, 39 lines)
+    ├── 📄 deploy.py (2,058 tokens, 230 lines)
+    ├── 📄 engine.py (12,042 tokens, 1,438 lines)
+    ├── 📄 ignore.py (2,342 tokens, 332 lines)
+    ├── 📄 isabelle_symbols.py (2,146 tokens, 462 lines)
+    ├── 📄 parse_file.py (27,040 tokens, 2,959 lines)
+    ├── 📁 scripts (1 folder, 1 file)
+    │   └── 📄 alias_tree_plus.sh (241 tokens, 30 lines)
+    ├── 📄 version.py (12 tokens, 1 line)
+    └── 📄 web.py (2,409 tokens, 321 lines)
diff --git a/tests/golden/legacy/trees_v1/dot_dot.txt b/tests/golden/legacy/trees_v1/dot_dot.txt
new file mode 100644
index 0000000..d6db06d
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/dot_dot.txt
@@ -0,0 +1,10 @@
+📁 dot_dot (2 folders, 4 files)
+├── 📄 my_test_file.py (7 tokens, 2 lines)
+│   └── def dot_dot_dot()
+└── 📁 nested_dir (1 folder, 3 files)
+    ├── 📄 .env.test (4 tokens, 0 lines)
+    │   └── DEBUG_TREE_PLUS
+    ├── 📄 pytest.ini (20 tokens, 4 lines)
+    └── 📄 test_tp_dotdot.py (362 tokens, 52 lines)
+        ├── def ignore_tokens_lines_test(text: str) -> str
+        └── def test_tree_plus_dotdot()
diff --git a/tests/golden/legacy/trees_v1/more_languages.txt b/tests/golden/legacy/trees_v1/more_languages.txt
new file mode 100644
index 0000000..a1de0a9
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages.txt
@@ -0,0 +1,1122 @@
+📁 more_languages (10 folders, 99 files)
+├── 📁 group1 (1 folder, 11 files)
+│   ├── 📄 addamt.cobol (441 tokens, 40 lines)
+│   ├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+│   ├── 📄 JavaTest.java (578 tokens, 86 lines)
+│   ├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+│   ├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+│   ├── 📄 lesson.cbl (635 tokens, 78 lines)
+│   ├── 📄 LuaTest.lua (83 tokens, 16 lines)
+│   ├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+│   ├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+│   ├── 📄 test.js (757 tokens, 154 lines)
+│   │   ├── class MyClass
+│   │   ├──   myMethod()
+│   │   ├──   async asyncMethod(a, b)
+│   │   ├──   methodWithDefaultParameters(a = 5, b = 10)
+│   │   ├──   multilineMethod(
+│   │   │       c,
+│   │   │       d
+│   │   │     )
+│   │   ├──   multilineMethodWithDefaults(
+│   │   │       t = "tree",
+│   │   │       p = "plus"
+│   │   │     )
+│   │   ├── function myFunction(param1, param2)
+│   │   ├── function multilineFunction(
+│   │   │     param1,
+│   │   │     param2
+│   │   │   )
+│   │   ├── const arrowFunction = () =>
+│   │   ├── const parametricArrow = (a, b) =>
+│   │   ├── function ()
+│   │   ├── function outerFunction(outerParam)
+│   │   ├──   function innerFunction(innerParam)
+│   │   ├──   innerFunction("inner")
+│   │   ├── const myObject = {
+│   │   ├──   myMethod: function (stuff)
+│   │   ├── let myArrowObject = {
+│   │   ├──   myArrow: ({
+│   │   │       a,
+│   │   │       b,
+│   │   │       c,
+│   │   │     }) =>
+│   │   ├── const myAsyncArrowFunction = async () =>
+│   │   ├── function functionWithRestParameters(...args)
+│   │   ├── const namedFunctionExpression = function myNamedFunction()
+│   │   ├── const multilineArrowFunction = (
+│   │   │     a,
+│   │   │     b
+│   │   │   ) =>
+│   │   ├── function functionReturningFunction()
+│   │   ├──   return function ()
+│   │   ├── function destructuringOnMultipleLines({
+│   │   │     a,
+│   │   │     b,
+│   │   │   })
+│   │   ├── const arrowFunctionWithDestructuring = ({ a, b }) =>
+│   │   ├── const multilineDestructuringArrow = ({
+│   │   │     a,
+│   │   │     b,
+│   │   │   }) =>
+│   │   ├── async function asyncFunctionWithErrorHandling()
+│   │   ├── class Car
+│   │   ├──   constructor(brand)
+│   │   ├──   present()
+│   │   ├── class Model extends Car
+│   │   ├──   constructor(brand, mod)
+│   │   ├──     super(brand)
+│   │   └──   show()
+│   └── 📄 test.ts (832 tokens, 165 lines)
+│       ├── type MyType
+│       ├── interface MyInterface
+│       ├── class TsClass
+│       ├──   myMethod()
+│       ├──   myMethodWithArgs(param1: string, param2: number): void
+│       ├──   static myStaticMethod<T>(param: T): T
+│       ├──   multilineMethod(
+│       │       c: number,
+│       │       d: number
+│       │     ): number
+│       ├──   multilineMethodWithDefaults(
+│       │       t: string = "tree",
+│       │       p: string = "plus"
+│       │     ): string
+│       ├── export class AdvancedComponent implements MyInterface
+│       ├──   async myAsyncMethod(
+│       │       a: string,
+│       │       b: number,
+│       │       c: string
+│       │     ): Promise<void>
+│       ├──   genericMethod<T, U>(
+│       │       arg1: T,
+│       │       arg2: U
+│       │     ): [T, U]
+│       ├── export class TicketsComponent implements MyInterface
+│       ├──   async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String
+│       │   })
+│       ├── function tsFunction()
+│       ├── function tsFunctionSigned(
+│       │     param1: number,
+│       │     param2: number
+│       │   ): void
+│       ├── export default async function tsFunctionComplicated<A, B, C>({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string>
+│       ├── const tsArrowFunctionSigned = ({
+│       │     a,
+│       │     b,
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │   }) =>
+│       ├── export const tsComplicatedArrow = async ({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string> =>
+│       ├── const arrowFunction = () =>
+│       ├── const arrow = (a: String, b: Number) =>
+│       ├── const asyncArrowFunction = async () =>
+│       ├── const asyncArrow = async (a: String, b: Number) =>
+│       ├── let weirdArrow = () =>
+│       ├── const asyncPromiseArrow = async (): Promise<void> =>
+│       ├── let myWeirdArrowSigned = (x: number): number =>
+│       ├── class Person
+│       ├──   constructor(private firstName: string, private lastName: string)
+│       ├──   getFullName(): string
+│       ├──   describe(): string
+│       ├── class Employee extends Person
+│       ├──   constructor(
+│       │       firstName: string,
+│       │       lastName: string,
+│       │       private jobTitle: string
+│       │     )
+│       ├──     super(firstName, lastName)
+│       ├──   describe(): string
+│       ├── interface Shape
+│       └── interface Square extends Shape
+├── 📁 group2 (1 folder, 8 files)
+│   ├── 📄 apl_test.apl (28 tokens, 5 lines)
+│   ├── 📄 c_test.c (837 tokens, 142 lines)
+│   │   ├── struct Point
+│   │   ├──     int x;
+│   │   ├──     int y;
+│   │   ├── struct Point getOrigin()
+│   │   ├── float mul_two_floats(float x1, float x2)
+│   │   ├── enum days
+│   │   ├──     SUN,
+│   │   ├──     MON,
+│   │   ├──     TUE,
+│   │   ├──     WED,
+│   │   ├──     THU,
+│   │   ├──     FRI,
+│   │   ├──     SAT
+│   │   ├── long add_two_longs(long x1, long x2)
+│   │   ├── double multiplyByTwo(double num)
+│   │   ├── char getFirstCharacter(char *str)
+│   │   ├── void greet(Person p)
+│   │   ├── typedef struct
+│   │   ├──     char name[50];
+│   │   ├── } Person;
+│   │   ├── int main()
+│   │   ├── int* getArrayStart(int arr[], int size)
+│   │   ├── long complexFunctionWithMultipleArguments(
+│   │   │       int param1,
+│   │   │       double param2,
+│   │   │       char *param3,
+│   │   │       struct Point point
+│   │   │   )
+│   │   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   │   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   │   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   │   *channel, int channellen, int is_pattern)
+│   │   ├──     while((ln = listNext(&li)))
+│   │   ├── static struct config
+│   │   ├──     aeEventLoop *el;
+│   │   ├──     cliConnInfo conn_info;
+│   │   ├──     const char *hostsocket;
+│   │   ├──     int tls;
+│   │   ├──     struct cliSSLconfig sslconfig;
+│   │   └── } config;
+│   ├── 📄 go_test.go (179 tokens, 46 lines)
+│   ├── 📄 PerlTest.pl (63 tokens, 20 lines)
+│   ├── 📄 PhpTest.php (70 tokens, 19 lines)
+│   ├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)
+│   ├── 📄 ScalaTest.scala (171 tokens, 40 lines)
+│   └── 📄 test.csv (0 tokens, 0 lines)
+│       ├── Name
+│       ├── Age
+│       ├── Country
+│       ├── City
+│       └── Email
+├── 📁 group3 (1 folder, 16 files)
+│   ├── 📄 bash_test.sh (127 tokens, 22 lines)
+│   ├── 📄 cpp_test.cpp (1,670 tokens, 259 lines)
+│   │   ├── class Person
+│   │   ├──     std::string name;
+│   │   ├── public:
+│   │   ├──     Person(std::string n) : name(n)
+│   │   ├──     void greet()
+│   │   ├── void globalGreet()
+│   │   ├── int main()
+│   │   ├── void printMessage(const std::string &message)
+│   │   ├── template<typename T>
+│   │   │   void printVector(const std::vector<T>& vec)
+│   │   ├── struct Point
+│   │   ├──     int x, y;
+│   │   ├──     Point(int x, int y) : x(x), y(y)
+│   │   ├── class Animal
+│   │   ├── public:
+│   │   ├──     Animal(const std::string &name) : name(name)
+│   │   ├──     virtual void speak() const
+│   │   ├──     virtual ~Animal()
+│   │   ├── protected:
+│   │   ├──     std::string name;
+│   │   ├── class Dog : public Animal
+│   │   ├── public:
+│   │   ├──     Dog(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── class Cat : public Animal
+│   │   ├── public:
+│   │   ├──     Cat(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   │   num_layers,
+│   │   │                                int batch_size, int max_seq_length,
+│   │   │   float dropout,
+│   │   │                                bool bidirectional, bool
+│   │   │   cudnn_allow_tf32,
+│   │   │                                int workspace_size, int
+│   │   │   reserve_space_size)
+│   │   ├── int main()
+│   │   ├── enum ECarTypes
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV,
+│   │   ├──   Wagon
+│   │   ├── ECarTypes GetPreferredCarType()
+│   │   ├── enum ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── enum class ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── void myFunction(string fname, int age)
+│   │   ├── template <typename T> T cos(T)
+│   │   ├── template <typename T> T sin(T)
+│   │   ├── template <typename T> T sqrt(T)
+│   │   ├── template<typename T> struct VLEN
+│   │   ├── template<typename T> class arr
+│   │   ├──   private:
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──   public:
+│   │   ├──     arr() : p(0), sz(0)
+│   │   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   │   ├──     arr(arr &&other)
+│   │   │         : p(other.p), sz(other.sz)
+│   │   ├──     ~arr()
+│   │   ├──     void resize(size_t n)
+│   │   ├──     T &operator[](size_t idx)
+│   │   ├──     T *data()
+│   │   ├──     size_t size() const
+│   │   ├── class Buffer
+│   │   ├──  private:
+│   │   ├──   void* ptr_;
+│   │   └── std::tuple<array, array, array> quantize(
+│   │           const array& w,
+│   │           int group_size,
+│   │           int bits,
+│   │           StreamOrDevice s)
+│   ├── 📄 csharp_test.cs (957 tokens, 146 lines)
+│   ├── 📄 hallucination.tex (1,633 tokens, 126 lines)
+│   ├── 📄 ruby_test.rb (138 tokens, 37 lines)
+│   ├── 📄 swift_test.swift (469 tokens, 110 lines)
+│   ├── 📄 test.lean (289 tokens, 42 lines)
+│   ├── 📄 test.capnp (117 tokens, 30 lines)
+│   ├── 📄 test.graphql (66 tokens, 21 lines)
+│   ├── 📄 test.proto (142 tokens, 34 lines)
+│   ├── 📄 test.sqlite (0 tokens, 0 lines)
+│   │   ├── students table:
+│   │   ├──    id integer primary key
+│   │   ├──    name text not null
+│   │   ├──    age integer not null
+│   │   ├── courses table:
+│   │   ├──    id integer primary key
+│   │   ├──    title text not null
+│   │   └──    credits integer not null
+│   ├── 📄 test_Cargo.toml (119 tokens, 18 lines)
+│   │   ├── name: test_cargo
+│   │   ├── version: 0.1.0
+│   │   ├── description: A test Cargo.toml
+│   │   ├── license: MIT OR Apache-2.0
+│   │   ├── dependencies:
+│   │   ├──   clap 4.4
+│   │   └──   sqlx 0.7 (features: runtime-tokio, tls-rustls)
+│   ├── 📄 test_json_rpc_2_0.json (26 tokens, 6 lines)
+│   │   ├── jsonrpc: 2.0
+│   │   ├── method: subtract
+│   │   ├── params:
+│   │   ├──     minuend: 42
+│   │   ├──     subtrahend: 23
+│   │   └── id: 1
+│   ├── 📄 test_openapi.yaml (753 tokens, 92 lines)
+│   │   ├── openapi: 3.0.1
+│   │   ├──     title: TODO Plugin
+│   │   ├──     description: A plugin to create and manage TODO lists using
+│   │   │   ChatGPT.
+│   │   ├──     version: v1
+│   │   ├── servers:
+│   │   ├──     - url: PLUGIN_HOSTNAME
+│   │   ├── paths:
+│   │   ├──     '/todos/{username}':
+│   │   ├──         GET (getTodos): Get the list of todos
+│   │   ├──         POST (addTodo): Add a todo to the list
+│   │   └──         DELETE (deleteTodo): Delete a todo from the list
+│   ├── 📄 test_openrpc.json (225 tokens, 44 lines)
+│   │   ├── openrpc: 1.2.1
+│   │   ├── info:
+│   │   ├──     title: Demo Petstore
+│   │   ├──     version: 1.0.0
+│   │   ├── methods:
+│   │   ├──     listPets: List all pets
+│   │   ├──         params:
+│   │   ├──             - limit: integer
+│   │   └──         result: pets = An array of pets
+│   └── 📄 test_pyproject.toml (304 tokens, 39 lines)
+│       ├── name: tree_plus
+│       ├── version: 1.0.8
+│       ├── description: A `tree` util enhanced with tokens, lines, and
+│       │   components.
+│       ├── License :: OSI Approved :: Apache Software License
+│       ├── License :: OSI Approved :: MIT License
+│       ├── dependencies:
+│       ├──     tiktoken
+│       ├──     PyYAML
+│       ├──     click
+│       ├──     rich
+│       └──     tomli
+├── 📁 group4 (1 folder, 10 files)
+│   ├── 📄 erl_test.erl (480 tokens, 68 lines)
+│   ├── 📄 haskell_test.hs (414 tokens, 41 lines)
+│   ├── 📄 mathematica_test.nb (133 tokens, 21 lines)
+│   ├── 📄 matlab_test.m (48 tokens, 12 lines)
+│   ├── 📄 RTest.R (367 tokens, 46 lines)
+│   ├── 📄 rust_test.rs (1,368 tokens, 259 lines)
+│   │   ├── fn at_beginning<'a>(&'a str)
+│   │   ├── pub enum Days<E: EdgeCase> {
+│   │   │       #\[default]
+│   │   │       Sun,
+│   │   │       Mon,
+│   │   │       #\[error("edge case {idx}, expected at least {} and at most {}",
+│   │   │   .limits.lo, .limits.hi)]
+│   │   │       Tue,
+│   │   │       Wed,
+│   │   │       Thu(i16, bool),
+│   │   │       Fri { day: u8 },
+│   │   │       Sat {
+│   │   │           urday: String,
+│   │   │           edge_case: E,
+│   │   │       },
+│   │   │   }
+│   │   ├── struct Point
+│   │   ├── impl Point
+│   │   ├──     fn get_origin() -> Point
+│   │   ├── struct Person
+│   │   ├── impl Person
+│   │   ├──     fn greet(&self)
+│   │   ├── fn add_two_longs(x1: i64, x2: i64) -> i64
+│   │   ├── fn add_two_longs_longer(
+│   │   │       x1: i64,
+│   │   │       x2: i64,
+│   │   │   ) -> i64
+│   │   ├── const fn multiply_by_two(num: f64) -> f64
+│   │   ├── fn get_first_character(s: &str) -> Option<char>
+│   │   ├── trait Drawable
+│   │   ├──     fn draw(&self)
+│   │   ├── impl Drawable for Point
+│   │   ├──     fn draw(&self)
+│   │   ├── fn with_generic<D: Drawable>(d: D)
+│   │   ├── fn with_generic<D>(d: D)
+│   │   │   where
+│   │   │       D: Drawable
+│   │   ├── fn main()
+│   │   ├── pub struct VisibleStruct
+│   │   ├── mod my_module
+│   │   ├──     pub struct AlsoVisibleStruct<T>(T, T)
+│   │   ├── macro_rules! say_hello
+│   │   ├── #[macro_export]
+│   │   │   macro_rules! hello_tree_plus
+│   │   ├── pub mod lib
+│   │   ├──     pub mod interfaces
+│   │   ├──     mod engine
+│   │   ├── pub fn flow<S1, S2, S3, S4, E, T, L>(
+│   │   │       source: S1,
+│   │   │       extractor: E,
+│   │   │       inbox: S2,
+│   │   │       transformer: T,
+│   │   │       outbox: S3,
+│   │   │       loader: L,
+│   │   │       sink: &mut S4,
+│   │   │   ) -> Result<(), Box<dyn Error>>
+│   │   │   where
+│   │   │       S1: Extractable,
+│   │   │       S2: Extractable + Loadable,
+│   │   │       S3: Extractable + Loadable,
+│   │   │       S4: Loadable,
+│   │   │       E: Extractor<S1, S2>,
+│   │   │       T: Transformer<S2, S3>,
+│   │   │       L: Loader<S3, S4>
+│   │   ├── trait Container
+│   │   ├──     fn items(&self) -> impl Iterator<Item = Widget>
+│   │   ├── trait HttpService
+│   │   ├──     async fn fetch(&self, url: Url) -> HtmlBody
+│   │   ├── struct Pair<T, U>
+│   │   ├── trait Transformer<T>
+│   │   ├──     fn transform(&self, input: T) -> T
+│   │   ├── impl<T: std::ops::Add<Output = T> + Copy> Transformer<T> for Pair<T,
+│   │   │   T>
+│   │   ├──     fn transform(&self, input: T) -> T
+│   │   ├── fn main()
+│   │   ├── async fn handle_get(State(pool): State<PgPool>) ->
+│   │   │   Result<Html<String>, (StatusCode, String)>
+│   │   │   where
+│   │   │       Bion: Cool
+│   │   ├── #[macro_export]
+│   │   │   macro_rules! unit
+│   │   ├──             fn insert(
+│   │   │                   &mut self,
+│   │   │                   key: (),
+│   │   │                   value: $unit_dtype,
+│   │   │               ) -> Result<Option<$unit_dtype>, ETLError>
+│   │   ├── pub async fn handle_get_axum_route(
+│   │   │       Session { maybe_claims }: Session,
+│   │   │       Path(RouteParams {
+│   │   │           alpha,
+│   │   │           bravo,
+│   │   │           charlie,
+│   │   │           edge_case
+│   │   │       }): Path<RouteParams>,
+│   │   │   ) -> ServerResult<Response>
+│   │   ├── fn encode_pipeline(cmds: &[Cmd], atomic: bool) -> Vec<u8>
+│   │   ├── pub async fn handle_post_yeet(
+│   │   │       State(auth_backend): State<AuthBackend>,
+│   │   │       Session { maybe_claims }: Session,
+│   │   │       Form(yeet_form): Form<YeetForm>,
+│   │   │   ) -> Result<Response, AuthError>
+│   │   └── pub async fn handle_get_thingy(
+│   │           session: Session,
+│   │           State(ApiBackend {
+│   │               page_cache,
+│   │               auth_backend,
+│   │               library_sql,
+│   │               some_data_cache,
+│   │               metadata_cache,
+│   │               thingy_client,
+│   │               ..
+│   │           }): State<ApiBackend>,
+│   │       ) -> ServerResult<Response>
+│   ├── 📄 test.zig (397 tokens, 60 lines)
+│   ├── 📄 test_fsharp.fs (92 tokens, 27 lines)
+│   ├── 📄 test_tcl_tk.tcl (54 tokens, 16 lines)
+│   └── 📄 tf_test.tf (202 tokens, 38 lines)
+├── 📁 group5 (1 folder, 19 files)
+│   ├── 📄 ansible_test.yml (55 tokens, 14 lines)
+│   │   ├── Install package
+│   │   ├── Start service
+│   │   └── Create user
+│   ├── 📄 app-routing.module.ts (287 tokens, 28 lines)
+│   │   ├── const routes: Routes = [
+│   │   │       { path: '', redirectTo: 'login', pathMatch: 'full' },
+│   │   │       { path: '*', redirectTo: 'login' },
+│   │   │       { path: 'home', component: HomeComponent },
+│   │   │       { path: 'login', component: LoginComponent },
+│   │   │       { path: 'register', component: RegisterComponent },
+│   │   │       { path: 'events', component: EventsComponent },
+│   │   │       { path: 'invites', component: InvitesComponent },
+│   │   │       { path: 'rewards', component: RewardsComponent },
+│   │   │       { path: 'profile', component: ProfileComponent },
+│   │   │   ];
+│   │   └── export class AppRoutingModule
+│   ├── 📄 app.component.spec.ts (410 tokens, 47 lines)
+│   │   ├── describe 'AppComponent'
+│   │   ├──     it should create the app
+│   │   ├──     it should welcome the user
+│   │   ├──     it should welcome 'Jimbo'
+│   │   └──     it should request login if not logged in
+│   ├── 📄 app.component.ts (271 tokens, 45 lines)
+│   │   ├── export class AppComponent
+│   │   ├──   constructor(
+│   │   │       private http: HttpClient,
+│   │   │       private loginService: LoginService,
+│   │   │       private stripeService: StripeService
+│   │   │     )
+│   │   ├──   constructor(private loginService: LoginService)
+│   │   ├──   checkSession()
+│   │   ├──   async goToEvent(event_id: string)
+│   │   └──   valInvitedBy(event: any, event_id: string)
+│   ├── 📄 app.module.ts (374 tokens, 43 lines)
+│   │   ├── @NgModule({
+│   │   │       declarations: [
+│   │   │           AppComponent,
+│   │   │           HomeComponent,
+│   │   │           LoginComponent,
+│   │   │           RegisterComponent,
+│   │   │           EventsComponent,
+│   │   │           InvitesComponent,
+│   │   │           RewardsComponent,
+│   │   │           ProfileComponent
+│   │   └── export class AppModule
+│   ├── 📄 checkbox_test.md (191 tokens, 29 lines)
+│   │   ├── # My Checkbox Test
+│   │   ├── ## My No Parens Test
+│   │   ├── ## My Empty href Test
+│   │   ├── ## My other url Test [Q&A]
+│   │   ├── ## My other other url Test [Q&A]
+│   │   ├── ## My 2nd other url Test [Q&A]
+│   │   ├── ## My 3rd other url Test [Q&A]
+│   │   ├── - [ ] Task 1
+│   │   ├──     - [ ] No Space Task 1.1
+│   │   ├──     - [ ] Two Spaces Task 1.2
+│   │   ├──         - [ ] Subtask 1.2.1
+│   │   ├── - [ ] Task 2
+│   │   ├── - [x] Task 3
+│   │   ├──     - [ ] Subtask 3.1
+│   │   ├── - [x] Task 6
+│   │   ├──     - [x] Subtask 6.1
+│   │   ├──         - [ ] Handle edge cases
+│   │   └── # My Codeblock Test
+│   ├── 📄 checkbox_test.txt (257 tokens, 33 lines)
+│   │   ├── - [ ] fix phone number format +1
+│   │   ├── - [ ] add forgot password
+│   │   ├── - [ ] ? add email verification
+│   │   ├── - [ ] store token the right way
+│   │   ├── - [ ] test nesting of checkboxes
+│   │   ├── - [ ] user can use option to buy ticket at 2-referred price
+│   │   ├── - [ ] CTA refer 2 people to get instant lower price
+│   │   └── - [ ] form to send referrals
+│   ├── 📄 environment.test.ts (197 tokens, 19 lines)
+│   │   ├── environment:
+│   │   ├──    production
+│   │   ├──    cognitoUserPoolId
+│   │   ├──    cognitoAppClientId
+│   │   └──    apiurl
+│   ├── 📄 hello_world.pyi (22 tokens, 3 lines)
+│   │   ├── @final
+│   │   │   class dtype(Generic[_DTypeScalar_co])
+│   │   └──     names: None | tuple[builtins.str, ...]
+│   ├── 📄 k8s_test.yaml (140 tokens, 37 lines)
+│   │   ├── apps/v1.Deployment -> my-app
+│   │   ├── v1.Service -> my-service
+│   │   └── v1.ConfigMap -> my-config
+│   ├── 📄 Makefile (714 tokens, 84 lines)
+│   │   ├── include dotenv/dev.env
+│   │   ├── .PHONY: dev
+│   │   ├── dev
+│   │   ├── services-down
+│   │   ├── services-stop: services-down
+│   │   ├── define CHECK_POSTGRES
+│   │   ├── damage-report
+│   │   ├── tail-logs
+│   │   └── cloud
+│   ├── 📄 requirements_test.txt (29 tokens, 10 lines)
+│   │   ├── psycopg2-binary
+│   │   ├── pytest
+│   │   ├── coverage
+│   │   ├── flask[async]
+│   │   ├── flask_cors
+│   │   ├── stripe
+│   │   ├── pyjwt[crypto]
+│   │   ├── cognitojwt[async]
+│   │   └── flask-lambda
+│   ├── 📄 rust_todo_test.rs (92 tokens, 26 lines)
+│   │   ├── TODO: This todo tests parse_todo
+│   │   ├── enum Color {
+│   │   │       Red,
+│   │   │       Blue,
+│   │   │       Green,
+│   │   │   }
+│   │   ├── struct Point
+│   │   ├── trait Drawable
+│   │   ├──     fn draw(&self)
+│   │   ├── impl Drawable for Point
+│   │   ├──     fn draw(&self)
+│   │   └── fn main()
+│   ├── 📄 sql_test.sql (270 tokens, 51 lines)
+│   ├── 📄 standard-app-routing.module.ts (100 tokens, 16 lines)
+│   │   └── const routes: Routes = [
+│   │         { path: '', component: HomeComponent },
+│   │         {
+│   │           path: 'heroes',
+│   │           component: HeroesListComponent,
+│   │           children: [
+│   │             { path: ':id', component: HeroDetailComponent },
+│   │             { path: 'new', component: HeroFormComponent },
+│   │           ],
+│   │         },
+│   │         { path: '**', component: PageNotFoundComponent },
+│   │       ];
+│   ├── 📄 test.env (190 tokens, 25 lines)
+│   │   ├── PROMO_PATH
+│   │   ├── PRODUCTION
+│   │   ├── SQL_SCHEMA_PATH
+│   │   ├── DB_LOGS
+│   │   ├── DB_LOG
+│   │   ├── PGPASSWORD
+│   │   ├── PGDATABASE
+│   │   ├── PGHOST
+│   │   ├── PGPORT
+│   │   ├── PGUSER
+│   │   ├── SERVER_LOG
+│   │   ├── SERVER_LOGS
+│   │   ├── API_URL
+│   │   ├── APP_LOGS
+│   │   ├── APP_LOG
+│   │   ├── APP_URL
+│   │   ├── COGNITO_USER_POOL_ID
+│   │   ├── COGNITO_APP_CLIENT_ID
+│   │   ├── AWS_REGION
+│   │   └── STRIPE_SECRET_KEY
+│   ├── 📄 testJsonSchema.json (421 tokens, 48 lines)
+│   │   ├── $schema: http://json-schema.org/draft-07/schema#
+│   │   ├── type: object
+│   │   ├── title: random_test
+│   │   └── description: A promoter's activites related to events
+│   ├── 📄 testPackage.json (349 tokens, 43 lines)
+│   │   ├── name: 'promo-app'
+│   │   ├── version: 0.0.0
+│   │   ├── scripts:
+│   │   ├──     ng: 'ng'
+│   │   ├──     start: 'ng serve'
+│   │   ├──     build: 'ng build'
+│   │   ├──     watch: 'ng build --watch --configuration development'
+│   │   └──     test: 'ng test'
+│   └── 📄 tickets.component.ts (7,160 tokens, 903 lines)
+│       ├── interface EnrichedTicket extends Ticket
+│       ├── interface SpinConfig
+│       ├── interface RotationState
+│       ├── interface SpeakInput
+│       ├── const formatSpeakInput = (input: SpeakInput): string =>
+│       ├── function hourToSpeech(hour: number, minute: number, period: string):
+│       │   string
+│       ├── export class TicketsComponent implements AfterViewInit
+│       ├──   speak(input: SpeakInput)
+│       ├──   speakEvent(ticket: EnrichedTicket): void
+│       ├──   formatEvent(ticket: EnrichedTicket): string
+│       ├──   speakVenue(ticket: EnrichedTicket): void
+│       ├──   formatDate(date: Date, oneLiner: boolean = false): string
+│       ├──   formatDateForSpeech(date: Date): string
+│       ├──   async spinQRCode(
+│       │       event: PointerEvent,
+│       │       config: SpinConfig = DEFAULT_SPIN_CONFIG
+│       │     )
+│       ├──   private animateRotation(
+│       │       imgElement: HTMLElement,
+│       │       targetRotation: number,
+│       │       config: SpinConfig,
+│       │       cleanup: () => void
+│       │     )
+│       ├──     const animate = (currentTime: number) =>
+│       ├──         requestAnimationFrame(animate)
+│       ├──         cleanup()
+│       ├──     requestAnimationFrame(animate)
+│       ├──   private getNext90Degree(currentRotation: number): number
+│       ├──   private getCurrentRotation(matrix: string): number
+│       ├──   ngAfterViewInit()
+│       ├──       const mouseEnterListener = () =>
+│       ├──       const mouseLeaveListener = () =>
+│       ├──   ngOnDestroy()
+│       ├──   toggleColumn(event: MatOptionSelectionChange, column: string)
+│       ├──   adjustColumns(event?: Event)
+│       ├──   onResize(event: Event)
+│       ├──   async ngOnInit()
+│       ├──   async loadTickets(): Promise<void>
+│       ├──   onDateRangeChange(
+│       │       type: "start" | "end",
+│       │       event: MatDatepickerInputEvent<Date>
+│       │     )
+│       ├──   applyFilter(column: string): void
+│       ├──   formatDateForComparison(date: Date): string
+│       ├──   constructor(private renderer: Renderer2)
+│       ├──   onFilterChange(event: Event, column: string)
+│       ├──   onLatitudeChange(event: Event)
+│       ├──   onLongitudeChange(event: Event)
+│       ├──   onRadiusChange(event: Event)
+│       ├──   sortData(sort: Sort): void
+│       ├──   onRowClick(event: Event, row: any)
+│       ├── function isDate(value: Date | undefined | null): value is Date
+│       ├── function isNonNullNumber(value: number | null): value is number
+│       ├── function hasLocation(
+│       │     ticket: any
+│       │   ): ticket is
+│       ├── const create_faker_ticket = async () =>
+│       ├── function compare(a: number | string, b: number | string, isAsc:
+│       │   boolean)
+│       ├── function compare_dates(a: Date, b: Date, isAsc: boolean)
+│       ├── async function mockMoreTickets(): Promise<Ticket[]>
+│       ├── const mockTickets = async () =>
+│       └── const renderQRCode = async (text: String): Promise<string> =>
+├── 📁 group6 (1 folder, 14 files)
+│   ├── 📄 catastrophic.c (5,339 tokens, 754 lines)
+│   │   ├── TODO: technically we should use a proper parser
+│   │   ├── struct Point
+│   │   ├──     int x;
+│   │   ├──     int y;
+│   │   ├── struct Point getOrigin()
+│   │   ├── float mul_two_floats(float x1, float x2)
+│   │   ├── enum days
+│   │   ├──     SUN,
+│   │   ├──     MON,
+│   │   ├──     TUE,
+│   │   ├──     WED,
+│   │   ├──     THU,
+│   │   ├──     FRI,
+│   │   ├──     SAT
+│   │   ├── enum worker_pool_flags
+│   │   ├──         POOL_BH                 = 1 << 0,
+│   │   ├──         POOL_MANAGER_ACTIVE     = 1 << 1,
+│   │   ├──         POOL_DISASSOCIATED      = 1 << 2,
+│   │   ├──         POOL_BH_DRAINING        = 1 << 3,
+│   │   ├── enum worker_flags
+│   │   ├──         WORKER_DIE              = 1 << 1,
+│   │   ├──         WORKER_IDLE             = 1 << 2,
+│   │   ├──         WORKER_PREP             = 1 << 3,
+│   │   ├──         WORKER_CPU_INTENSIVE    = 1 << 6,
+│   │   ├──         WORKER_UNBOUND          = 1 << 7,
+│   │   ├──         WORKER_REBOUND          = 1 << 8,
+│   │   ├──         WORKER_NOT_RUNNING      = WORKER_PREP | WORKER_CPU_INTENSIVE
+│   │   │   |
+│   │   │                                     WORKER_UNBOUND | WORKER_REBOUND,
+│   │   ├── struct worker_pool
+│   │   ├──   raw_spinlock_t    lock;
+│   │   ├──   int      cpu;
+│   │   ├──   int      node;
+│   │   ├──   int      id;
+│   │   ├──   unsigned int    flags;
+│   │   ├──   unsigned long    watchdog_ts;
+│   │   ├──   bool      cpu_stall;
+│   │   ├──   int      nr_running;
+│   │   ├──   struct list_head  worklist;
+│   │   ├──   int      nr_workers;
+│   │   ├──   int      nr_idle;
+│   │   ├──   struct list_head  idle_list;
+│   │   ├──   struct timer_list  idle_timer;
+│   │   ├──   struct work_struct      idle_cull_work;
+│   │   ├──   struct timer_list  mayday_timer;
+│   │   ├──   struct worker    *manager;
+│   │   ├──   struct list_head  workers;
+│   │   ├──   struct ida    worker_ida;
+│   │   ├──   struct workqueue_attrs  *attrs;
+│   │   ├──   struct hlist_node  hash_node;
+│   │   ├──   int      refcnt;
+│   │   ├──   struct rcu_head    rcu;
+│   │   ├── long add_two_longs(long x1, long x2)
+│   │   ├── double multiplyByTwo(double num)
+│   │   ├── char getFirstCharacter(char *str)
+│   │   ├── void greet(Person p)
+│   │   ├── typedef struct
+│   │   ├──     char name[50];
+│   │   ├── } Person;
+│   │   ├── typedef struct PersonA
+│   │   ├──     char name[50];
+│   │   ├── } PersonB;
+│   │   ├── int main()
+│   │   ├── int* getArrayStart(int arr[], int size)
+│   │   ├── long complexFunctionWithMultipleArguments(
+│   │   │       int param1,
+│   │   │       double param2,
+│   │   │       char *param3,
+│   │   │       struct Point point
+│   │   │   )
+│   │   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   │   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   │   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   │   *channel, int channellen, int is_pattern)
+│   │   ├──     while((ln = listNext(&li)))
+│   │   ├── static struct config
+│   │   ├──     aeEventLoop *el;
+│   │   ├──     cliConnInfo conn_info;
+│   │   ├──     const char *hostsocket;
+│   │   ├──     int tls;
+│   │   ├──     struct cliSSLconfig sslconfig;
+│   │   ├── } config;
+│   │   ├── class Person
+│   │   ├──     std::string name;
+│   │   ├── public:
+│   │   ├──     Person(std::string n) : name(n)
+│   │   ├──     void greet()
+│   │   ├── void globalGreet()
+│   │   ├── int main()
+│   │   ├── void printMessage(const std::string &message)
+│   │   ├── template<typename T>
+│   │   │   void printVector(const std::vector<T>& vec)
+│   │   ├── struct foo
+│   │   ├──     char x;
+│   │   ├──     struct foo_in
+│   │   ├──           char* y;
+│   │   ├──           short z;
+│   │   ├──     } inner;
+│   │   ├── struct Point
+│   │   ├──     int x, y;
+│   │   ├──     Point(int x, int y) : x(x), y(y)
+│   │   ├── class Animal
+│   │   ├──   public:
+│   │   ├──     Animal(const std::string &name) : name(name)
+│   │   ├──     virtual void speak() const
+│   │   ├──     virtual ~Animal()
+│   │   ├── protected:
+│   │   ├──     std::string name;
+│   │   ├── class Dog : public Animal
+│   │   ├──   public:
+│   │   ├──     Dog(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── class Cat : public Animal
+│   │   ├──   public:
+│   │   ├──     Cat(const std::string &name) : Animal(name)
+│   │   ├──     void speak() const override
+│   │   ├── class CatDog: public Animal, public Cat, public Dog
+│   │   ├──   public:
+│   │   ├──       CatDog(const std::string &name) : Animal(name)
+│   │   ├──       int meow_bark()
+│   │   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   │   num_layers,
+│   │   │                                int batch_size, int max_seq_length,
+│   │   │   float dropout,
+│   │   │                                bool bidirectional, bool
+│   │   │   cudnn_allow_tf32,
+│   │   │                                int workspace_size, int
+│   │   │   reserve_space_size)
+│   │   ├── int main()
+│   │   ├── enum ECarTypes
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV,
+│   │   ├──   Wagon
+│   │   ├── ECarTypes GetPreferredCarType()
+│   │   ├── enum ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── enum class ECarTypes : uint8_t
+│   │   ├──   Sedan,
+│   │   ├──   Hatchback,
+│   │   ├──   SUV = 254,
+│   │   ├──   Hybrid
+│   │   ├── void myFunction(string fname, int age)
+│   │   ├── template <typename T> T cos(T)
+│   │   ├── template <typename T> T sin(T)
+│   │   ├── template <typename T> T sqrt(T)
+│   │   ├── template<typename T> struct VLEN
+│   │   ├── template<typename T> class arr
+│   │   ├──   private:
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──     static T *ralloc(size_t num)
+│   │   ├──     static void dealloc(T *ptr)
+│   │   ├──   public:
+│   │   ├──     arr() : p(0), sz(0)
+│   │   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   │   ├──     arr(arr &&other)
+│   │   │         : p(other.p), sz(other.sz)
+│   │   ├──     ~arr()
+│   │   ├──     void resize(size_t n)
+│   │   ├──     T &operator[](size_t idx)
+│   │   ├──     T *data()
+│   │   ├──     size_t size() const
+│   │   ├── class Buffer
+│   │   ├──  private:
+│   │   ├──   void* ptr_;
+│   │   ├── std::tuple<array, array, array> quantize(
+│   │   │       const array& w,
+│   │   │       int group_size,
+│   │   │       int bits,
+│   │   │       StreamOrDevice s)
+│   │   ├── #define PY_SSIZE_T_CLEAN
+│   │   ├── #define PLATFORM_IS_X86
+│   │   ├── #define PLATFORM_WINDOWS
+│   │   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   │   ├── static int GetXCR0EAX()
+│   │   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   │   ├── static int GetXCR0EAX()
+│   │   ├──   asm("XGETBV" : "=a"(eax), "=d"(edx) : "c"(0))
+│   │   ├── static void ReportMissingCpuFeature(const char* name)
+│   │   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   │   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   │   ├── static PyMethodDef cpu_feature_guard_methods[]
+│   │   ├── static struct PyModuleDef cpu_feature_guard_module
+│   │   ├── #define EXPORT_SYMBOL __declspec(dllexport)
+│   │   ├── #define EXPORT_SYMBOL __attribute__ ((visibility("default")))
+│   │   ├── EXPORT_SYMBOL PyMODINIT_FUNC PyInit_cpu_feature_guard(void)
+│   │   ├── typedef struct
+│   │   ├──     GPT2Config config;
+│   │   ├──     ParameterTensors params;
+│   │   ├──     size_t param_sizes[NUM_PARAMETER_TENSORS];
+│   │   ├──     float* params_memory;
+│   │   ├──     size_t num_parameters;
+│   │   ├──     ParameterTensors grads;
+│   │   ├──     float* grads_memory;
+│   │   ├──     float* m_memory;
+│   │   ├──     float* v_memory;
+│   │   ├──     ActivationTensors acts;
+│   │   ├──     size_t act_sizes[NUM_ACTIVATION_TENSORS];
+│   │   ├──     float* acts_memory;
+│   │   ├──     size_t num_activations;
+│   │   ├──     ActivationTensors grads_acts;
+│   │   ├──     float* grads_acts_memory;
+│   │   ├──     int batch_size;
+│   │   ├──     int seq_len;
+│   │   ├──     int* inputs;
+│   │   ├──     int* targets;
+│   │   ├──     float mean_loss;
+│   │   └── } GPT2;
+│   ├── 📄 cpp_examples_impl.cc (60 tokens, 10 lines)
+│   │   ├── PYBIND11_MODULE(cpp_examples, m)
+│   │   └──     m.def("add", &add<int>, "An example function to add two
+│   │       numbers.")
+│   ├── 📄 cpp_examples_impl.cu (37 tokens, 10 lines)
+│   │   ├── template <typename T>
+│   │   │   T add(T a, T b)
+│   │   └── template <>
+│   │       int add<int>(int a, int b)
+│   ├── 📄 cpp_examples_impl.h (22 tokens, 6 lines)
+│   │   ├── template <typename T>
+│   │   │   T add(T a, T b)
+│   │   └── template <>
+│   │       int add<int>(int, int)
+│   ├── 📄 edge_case.hpp (426 tokens, 28 lines)
+│   ├── 📄 fractal.thy (1,712 tokens, 147 lines)
+│   ├── 📄 Microsoft.PowerShell_profile.ps1 (3,346 tokens, 497 lines)
+│   ├── 📄 python_complex_class.py (10 tokens, 2 lines)
+│   │   └── class Box(Space[NDArray[Any]])
+│   ├── 📄 ramda__cloneRegExp.js (173 tokens, 9 lines)
+│   │   └── export default function _cloneRegExp(pattern)
+│   ├── 📄 ramda_prop.js (646 tokens, 85 lines)
+│   │   ├── /**
+│   │   │    * Returns a function that when supplied an object returns the
+│   │   │   indicated
+│   │   │    * property of that object, if it exists.
+│   │   │    * @category Object
+│   │   │    * @typedefn Idx = String | Int | Symbol
+│   │   │    * @sig Idx -> {s: a} -> a | Undefined
+│   │   │    * @param {String|Number} p The property name or array index
+│   │   │    * @param {Object} obj The object to query
+│   │   │    * @return {*} The value at `obj.p`.
+│   │   │    */
+│   │   │   var prop = _curry2(function prop(p, obj)
+│   │   ├── /**
+│   │   │    * Solves equations of the form a * x = b
+│   │   │    * @param {{
+│   │   │    *  z: number
+│   │   │    * }} x
+│   │   │    */
+│   │   │   function foo(x)
+│   │   ├── /**
+│   │   │    * Deconstructs an array field from the input documents to output a
+│   │   │   document for each element.
+│   │   │    * Each output document is the input document with the value of the
+│   │   │   array field replaced by the element.
+│   │   │    * @category Object
+│   │   │    * @sig String -> {k: [v]} -> [{k: v}]
+│   │   │    * @param {String} key The key to determine which property of the
+│   │   │   object should be unwound.
+│   │   │    * @param {Object} object The object containing the list to unwind
+│   │   │   at the property named by the key.
+│   │   │    * @return {List} A list of new objects, each having the given key
+│   │   │   associated to an item from the unwound list.
+│   │   │    */
+│   │   │   var unwind = _curry2(function(key, object)
+│   │   └──   return _map(function(item)
+│   ├── 📄 tensorflow_flags.h (7,628 tokens, 668 lines)
+│   │   ├── #define TENSORFLOW_CORE_CONFIG_FLAG_DEFS_H_
+│   │   ├── class Flags
+│   │   ├──  public:
+│   │   ├── bool SetterForXlaAutoJitFlag(const string& value)
+│   │   ├── bool SetterForXlaCallModuleDisabledChecks(const string& value)
+│   │   ├── void AppendMarkForCompilationPassFlagsInternal(std::vector<Flag>*
+│   │   │   flag_list)
+│   │   ├── void AllocateAndParseJitRtFlags()
+│   │   ├── void AllocateAndParseFlags()
+│   │   ├── void ResetFlags()
+│   │   ├── bool SetXlaAutoJitFlagFromFlagString(const string& value)
+│   │   ├── BuildXlaOpsPassFlags* GetBuildXlaOpsPassFlags()
+│   │   ├── MarkForCompilationPassFlags* GetMarkForCompilationPassFlags()
+│   │   ├── XlaSparseCoreFlags* GetXlaSparseCoreFlags()
+│   │   ├── XlaDeviceFlags* GetXlaDeviceFlags()
+│   │   ├── XlaOpsCommonFlags* GetXlaOpsCommonFlags()
+│   │   ├── XlaCallModuleFlags* GetXlaCallModuleFlags()
+│   │   ├── MlirCommonFlags* GetMlirCommonFlags()
+│   │   ├── void ResetJitCompilerFlags()
+│   │   ├── const JitRtFlags& GetJitRtFlags()
+│   │   ├── ConfigProto::Experimental::MlirBridgeRollout
+│   │   │   GetMlirBridgeRolloutState(
+│   │   │       std::optional<const ConfigProto> config_proto)
+│   │   ├── void AppendMarkForCompilationPassFlags(std::vector<Flag>* flag_list)
+│   │   ├── void DisableXlaCompilation()
+│   │   ├── void EnableXlaCompilation()
+│   │   ├── bool FailOnXlaCompilation()
+│   │   ├── #define TF_PY_DECLARE_FLAG(flag_name)
+│   │   └── PYBIND11_MODULE(flags_pybind, m)
+│   ├── 📄 test.f (181 tokens, 30 lines)
+│   ├── 📄 torch.rst (60 tokens, 8 lines)
+│   │   ├── # libtorch (C++-only)
+│   │   └── - Building libtorch using Python
+│   └── 📄 yc.html (9,063 tokens, 169 lines)
+├── 📁 group7 (1 folder, 5 files)
+│   ├── 📄 absurdly_huge.jsonl (8,347 tokens, 126 lines)
+│   │   ├── SMILES: str
+│   │   ├── Yield: float
+│   │   ├── Temperature: int
+│   │   ├── Pressure: float
+│   │   ├── Solvent: str
+│   │   ├── Success: bool
+│   │   ├── Reaction_Conditions: dict
+│   │   ├── Products: list
+│   │   └── EdgeCasesMissed: None
+│   ├── 📄 angular_crud.ts (1,192 tokens, 148 lines)
+│   │   ├── interface DBCommand<T = any>
+│   │   ├── export class IndexedDbService
+│   │   ├──     constructor()
+│   │   ├──     async create_connection({ db_name = 'client_db', table_name }:
+│   │   │   DBCommand)
+│   │   ├──                 upgrade(db)
+│   │   ├──     async create_model({ db_name, table_name, model }: DBCommand)
+│   │   ├──         verify_matching({ table_name, model })
+│   │   ├──     async read_key({ db_name, table_name, key }: DBCommand)
+│   │   ├──     async update_model({ db_name, table_name, model }: DBCommand)
+│   │   ├──         verify_matching({ table_name, model })
+│   │   ├──     async delete_key({ db_name, table_name, key }: DBCommand)
+│   │   ├──     async list_table({
+│   │   │           db_name,
+│   │   │           table_name,
+│   │   │           where,
+│   │   │       }: DBCommand & { where?: { [key: string]: string | number } })
+│   │   └──     async search_table(criteria: SearchCriteria)
+│   ├── 📄 structure.py (400 tokens, 92 lines)
+│   │   ├── @runtime_checkable
+│   │   │   class DataClass(Protocol)
+│   │   ├──     __dataclass_fields__: dict
+│   │   ├── class MyInteger(Enum)
+│   │   ├──     ONE = 1
+│   │   ├──     TWO = 2
+│   │   ├──     THREE = 42
+│   │   ├── class MyString(Enum)
+│   │   ├──     AAA1 = "aaa"
+│   │   ├──     BB_B = """edge
+│   │   │   case"""
+│   │   ├── @dataclass(frozen=True, slots=True, kw_only=True)
+│   │   │   class Tool
+│   │   ├──     name: str
+│   │   ├──     description: str
+│   │   ├──     input_model: DataClass
+│   │   ├──     output_model: DataClass
+│   │   ├──     def execute(self, *args, **kwargs)
+│   │   ├──     @property
+│   │   │       def edge_case(self) -> str
+│   │   ├──     def should_still_see_me(self, x: bool = True) -> "Tool"
+│   │   ├── @dataclass
+│   │   │   class MyInput[T]
+│   │   ├──     name: str
+│   │   ├──     rank: MyInteger
+│   │   ├──     serial_n: int
+│   │   ├── @dataclass
+│   │   │   class Thingy
+│   │   ├──     is_edge_case: bool
+│   │   ├── @dataclass
+│   │   │   class MyOutput
+│   │   ├──     orders: str
+│   │   ├── class MyTools(Enum)
+│   │   ├──     TOOL_A = Tool(
+│   │   │           name="complicated",
+│   │   │           description="edge case!",
+│   │   │           input_model=MyInput[Thingy],
+│   │   │           output_model=MyOutput,
+│   │   │       )
+│   │   ├──     TOOL_B = Tool(
+│   │   │           name="""super
+│   │   │   complicated
+│   │   │   """,
+│   │   │           description="edge case!",
+│   │   │           input_model=MyInput,
+│   │   │           output_model=MyOutput,
+│   │   │       )
+│   │   ├── @final
+│   │   │   class dtype(Generic[_DTypeScalar_co])
+│   │   └──     names: None | tuple[builtins.str, ...]
+│   ├── 📄 test.wgsl (528 tokens, 87 lines)
+│   └── 📄 test.metal (272 tokens, 34 lines)
+├── 📁 group_lisp (1 folder, 4 files)
+│   ├── 📄 clojure_test.clj (682 tokens, 85 lines)
+│   ├── 📄 LispTest.lisp (25 tokens, 6 lines)
+│   ├── 📄 racket_struct.rkt (14 tokens, 1 line)
+│   └── 📄 test_scheme.scm (360 tokens, 44 lines)
+└── 📁 group_todo (1 folder, 12 files)
+    ├── 📄 AAPLShaders.metal (5,780 tokens, 566 lines)
+    ├── 📄 crystal_test.cr (48 tokens, 15 lines)
+    ├── 📄 dart_test.dart (108 tokens, 24 lines)
+    ├── 📄 elixir_test.exs (39 tokens, 10 lines)
+    ├── 📄 forward.frag (739 tokens, 87 lines)
+    ├── 📄 forward.vert (359 tokens, 48 lines)
+    ├── 📄 nodemon.json (118 tokens, 20 lines)
+    ├── 📄 sas_test.sas (97 tokens, 22 lines)
+    ├── 📄 test_setup_py.test (133 tokens, 24 lines)
+    ├── 📄 testTypings.d.ts (158 tokens, 23 lines)
+    ├── 📄 vba_test.bas (67 tokens, 16 lines)
+    └── 📄 wgsl_test.wgsl (94 tokens, 17 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages_group1.txt b/tests/golden/legacy/trees_v1/more_languages_group1.txt
new file mode 100644
index 0000000..1d32827
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group1.txt
@@ -0,0 +1,146 @@
+📁 group1 (1 folder, 11 files)
+├── 📄 addamt.cobol (441 tokens, 40 lines)
+├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+├── 📄 JavaTest.java (578 tokens, 86 lines)
+├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+├── 📄 lesson.cbl (635 tokens, 78 lines)
+├── 📄 LuaTest.lua (83 tokens, 16 lines)
+├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+├── 📄 test.js (757 tokens, 154 lines)
+│   ├── class MyClass
+│   ├──   myMethod()
+│   ├──   async asyncMethod(a, b)
+│   ├──   methodWithDefaultParameters(a = 5, b = 10)
+│   ├──   multilineMethod(
+│   │       c,
+│   │       d
+│   │     )
+│   ├──   multilineMethodWithDefaults(
+│   │       t = "tree",
+│   │       p = "plus"
+│   │     )
+│   ├── function myFunction(param1, param2)
+│   ├── function multilineFunction(
+│   │     param1,
+│   │     param2
+│   │   )
+│   ├── const arrowFunction = () =>
+│   ├── const parametricArrow = (a, b) =>
+│   ├── function ()
+│   ├── function outerFunction(outerParam)
+│   ├──   function innerFunction(innerParam)
+│   ├──   innerFunction("inner")
+│   ├── const myObject = {
+│   ├──   myMethod: function (stuff)
+│   ├── let myArrowObject = {
+│   ├──   myArrow: ({
+│   │       a,
+│   │       b,
+│   │       c,
+│   │     }) =>
+│   ├── const myAsyncArrowFunction = async () =>
+│   ├── function functionWithRestParameters(...args)
+│   ├── const namedFunctionExpression = function myNamedFunction()
+│   ├── const multilineArrowFunction = (
+│   │     a,
+│   │     b
+│   │   ) =>
+│   ├── function functionReturningFunction()
+│   ├──   return function ()
+│   ├── function destructuringOnMultipleLines({
+│   │     a,
+│   │     b,
+│   │   })
+│   ├── const arrowFunctionWithDestructuring = ({ a, b }) =>
+│   ├── const multilineDestructuringArrow = ({
+│   │     a,
+│   │     b,
+│   │   }) =>
+│   ├── async function asyncFunctionWithErrorHandling()
+│   ├── class Car
+│   ├──   constructor(brand)
+│   ├──   present()
+│   ├── class Model extends Car
+│   ├──   constructor(brand, mod)
+│   ├──     super(brand)
+│   └──   show()
+└── 📄 test.ts (832 tokens, 165 lines)
+    ├── type MyType
+    ├── interface MyInterface
+    ├── class TsClass
+    ├──   myMethod()
+    ├──   myMethodWithArgs(param1: string, param2: number): void
+    ├──   static myStaticMethod<T>(param: T): T
+    ├──   multilineMethod(
+    │       c: number,
+    │       d: number
+    │     ): number
+    ├──   multilineMethodWithDefaults(
+    │       t: string = "tree",
+    │       p: string = "plus"
+    │     ): string
+    ├── export class AdvancedComponent implements MyInterface
+    ├──   async myAsyncMethod(
+    │       a: string,
+    │       b: number,
+    │       c: string
+    │     ): Promise<void>
+    ├──   genericMethod<T, U>(
+    │       arg1: T,
+    │       arg2: U
+    │     ): [T, U]
+    ├── export class TicketsComponent implements MyInterface
+    ├──   async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String })
+    ├── function tsFunction()
+    ├── function tsFunctionSigned(
+    │     param1: number,
+    │     param2: number
+    │   ): void
+    ├── export default async function tsFunctionComplicated<A, B, C>({
+    │     a = 1 | 2,
+    │     b = "bob",
+    │     c = async () => "charlie",
+    │   }: {
+    │     a: number;
+    │     b: string;
+    │     c: () => Promise<string>;
+    │   }): Promise<string>
+    ├── const tsArrowFunctionSigned = ({
+    │     a,
+    │     b,
+    │   }: {
+    │     a: number;
+    │     b: string;
+    │   }) =>
+    ├── export const tsComplicatedArrow = async ({
+    │     a = 1 | 2,
+    │     b = "bob",
+    │     c = async () => "charlie",
+    │   }: {
+    │     a: number;
+    │     b: string;
+    │     c: () => Promise<string>;
+    │   }): Promise<string> =>
+    ├── const arrowFunction = () =>
+    ├── const arrow = (a: String, b: Number) =>
+    ├── const asyncArrowFunction = async () =>
+    ├── const asyncArrow = async (a: String, b: Number) =>
+    ├── let weirdArrow = () =>
+    ├── const asyncPromiseArrow = async (): Promise<void> =>
+    ├── let myWeirdArrowSigned = (x: number): number =>
+    ├── class Person
+    ├──   constructor(private firstName: string, private lastName: string)
+    ├──   getFullName(): string
+    ├──   describe(): string
+    ├── class Employee extends Person
+    ├──   constructor(
+    │       firstName: string,
+    │       lastName: string,
+    │       private jobTitle: string
+    │     )
+    ├──     super(firstName, lastName)
+    ├──   describe(): string
+    ├── interface Shape
+    └── interface Square extends Shape
diff --git a/tests/golden/legacy/trees_v1/more_languages_group2.txt b/tests/golden/legacy/trees_v1/more_languages_group2.txt
new file mode 100644
index 0000000..854ea5d
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group2.txt
@@ -0,0 +1,54 @@
+📁 group2 (1 folder, 8 files)
+├── 📄 apl_test.apl (28 tokens, 5 lines)
+├── 📄 c_test.c (837 tokens, 142 lines)
+│   ├── struct Point
+│   ├──     int x;
+│   ├──     int y;
+│   ├── struct Point getOrigin()
+│   ├── float mul_two_floats(float x1, float x2)
+│   ├── enum days
+│   ├──     SUN,
+│   ├──     MON,
+│   ├──     TUE,
+│   ├──     WED,
+│   ├──     THU,
+│   ├──     FRI,
+│   ├──     SAT
+│   ├── long add_two_longs(long x1, long x2)
+│   ├── double multiplyByTwo(double num)
+│   ├── char getFirstCharacter(char *str)
+│   ├── void greet(Person p)
+│   ├── typedef struct
+│   ├──     char name[50];
+│   ├── } Person;
+│   ├── int main()
+│   ├── int* getArrayStart(int arr[], int size)
+│   ├── long complexFunctionWithMultipleArguments(
+│   │       int param1,
+│   │       double param2,
+│   │       char *param3,
+│   │       struct Point point
+│   │   )
+│   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   *channel, int channellen, int is_pattern)
+│   ├──     while((ln = listNext(&li)))
+│   ├── static struct config
+│   ├──     aeEventLoop *el;
+│   ├──     cliConnInfo conn_info;
+│   ├──     const char *hostsocket;
+│   ├──     int tls;
+│   ├──     struct cliSSLconfig sslconfig;
+│   └── } config;
+├── 📄 go_test.go (179 tokens, 46 lines)
+├── 📄 PerlTest.pl (63 tokens, 20 lines)
+├── 📄 PhpTest.php (70 tokens, 19 lines)
+├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)
+├── 📄 ScalaTest.scala (171 tokens, 40 lines)
+└── 📄 test.csv (0 tokens, 0 lines)
+    ├── Name
+    ├── Age
+    ├── Country
+    ├── City
+    └── Email
diff --git a/tests/golden/legacy/trees_v1/more_languages_group3.txt b/tests/golden/legacy/trees_v1/more_languages_group3.txt
new file mode 100644
index 0000000..1b4769e
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group3.txt
@@ -0,0 +1,149 @@
+📁 group3 (1 folder, 16 files)
+├── 📄 bash_test.sh (127 tokens, 22 lines)
+├── 📄 cpp_test.cpp (1,670 tokens, 259 lines)
+│   ├── class Person
+│   ├──     std::string name;
+│   ├── public:
+│   ├──     Person(std::string n) : name(n)
+│   ├──     void greet()
+│   ├── void globalGreet()
+│   ├── int main()
+│   ├── void printMessage(const std::string &message)
+│   ├── template<typename T>
+│   │   void printVector(const std::vector<T>& vec)
+│   ├── struct Point
+│   ├──     int x, y;
+│   ├──     Point(int x, int y) : x(x), y(y)
+│   ├── class Animal
+│   ├── public:
+│   ├──     Animal(const std::string &name) : name(name)
+│   ├──     virtual void speak() const
+│   ├──     virtual ~Animal()
+│   ├── protected:
+│   ├──     std::string name;
+│   ├── class Dog : public Animal
+│   ├── public:
+│   ├──     Dog(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── class Cat : public Animal
+│   ├── public:
+│   ├──     Cat(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   num_layers,
+│   │                                int batch_size, int max_seq_length, float
+│   │   dropout,
+│   │                                bool bidirectional, bool cudnn_allow_tf32,
+│   │                                int workspace_size, int reserve_space_size)
+│   ├── int main()
+│   ├── enum ECarTypes
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV,
+│   ├──   Wagon
+│   ├── ECarTypes GetPreferredCarType()
+│   ├── enum ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── enum class ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── void myFunction(string fname, int age)
+│   ├── template <typename T> T cos(T)
+│   ├── template <typename T> T sin(T)
+│   ├── template <typename T> T sqrt(T)
+│   ├── template<typename T> struct VLEN
+│   ├── template<typename T> class arr
+│   ├──   private:
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──   public:
+│   ├──     arr() : p(0), sz(0)
+│   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   ├──     arr(arr &&other)
+│   │         : p(other.p), sz(other.sz)
+│   ├──     ~arr()
+│   ├──     void resize(size_t n)
+│   ├──     T &operator[](size_t idx)
+│   ├──     T *data()
+│   ├──     size_t size() const
+│   ├── class Buffer
+│   ├──  private:
+│   ├──   void* ptr_;
+│   └── std::tuple<array, array, array> quantize(
+│           const array& w,
+│           int group_size,
+│           int bits,
+│           StreamOrDevice s)
+├── 📄 csharp_test.cs (957 tokens, 146 lines)
+├── 📄 hallucination.tex (1,633 tokens, 126 lines)
+├── 📄 ruby_test.rb (138 tokens, 37 lines)
+├── 📄 swift_test.swift (469 tokens, 110 lines)
+├── 📄 test.lean (289 tokens, 42 lines)
+├── 📄 test.capnp (117 tokens, 30 lines)
+├── 📄 test.graphql (66 tokens, 21 lines)
+├── 📄 test.proto (142 tokens, 34 lines)
+├── 📄 test.sqlite (0 tokens, 0 lines)
+│   ├── students table:
+│   ├──    id integer primary key
+│   ├──    name text not null
+│   ├──    age integer not null
+│   ├── courses table:
+│   ├──    id integer primary key
+│   ├──    title text not null
+│   └──    credits integer not null
+├── 📄 test_Cargo.toml (119 tokens, 18 lines)
+│   ├── name: test_cargo
+│   ├── version: 0.1.0
+│   ├── description: A test Cargo.toml
+│   ├── license: MIT OR Apache-2.0
+│   ├── dependencies:
+│   ├──   clap 4.4
+│   └──   sqlx 0.7 (features: runtime-tokio, tls-rustls)
+├── 📄 test_json_rpc_2_0.json (26 tokens, 6 lines)
+│   ├── jsonrpc: 2.0
+│   ├── method: subtract
+│   ├── params:
+│   ├──     minuend: 42
+│   ├──     subtrahend: 23
+│   └── id: 1
+├── 📄 test_openapi.yaml (753 tokens, 92 lines)
+│   ├── openapi: 3.0.1
+│   ├──     title: TODO Plugin
+│   ├──     description: A plugin to create and manage TODO lists using ChatGPT.
+│   ├──     version: v1
+│   ├── servers:
+│   ├──     - url: PLUGIN_HOSTNAME
+│   ├── paths:
+│   ├──     '/todos/{username}':
+│   ├──         GET (getTodos): Get the list of todos
+│   ├──         POST (addTodo): Add a todo to the list
+│   └──         DELETE (deleteTodo): Delete a todo from the list
+├── 📄 test_openrpc.json (225 tokens, 44 lines)
+│   ├── openrpc: 1.2.1
+│   ├── info:
+│   ├──     title: Demo Petstore
+│   ├──     version: 1.0.0
+│   ├── methods:
+│   ├──     listPets: List all pets
+│   ├──         params:
+│   ├──             - limit: integer
+│   └──         result: pets = An array of pets
+└── 📄 test_pyproject.toml (304 tokens, 39 lines)
+    ├── name: tree_plus
+    ├── version: 1.0.8
+    ├── description: A `tree` util enhanced with tokens, lines, and components.
+    ├── License :: OSI Approved :: Apache Software License
+    ├── License :: OSI Approved :: MIT License
+    ├── dependencies:
+    ├──     tiktoken
+    ├──     PyYAML
+    ├──     click
+    ├──     rich
+    └──     tomli
diff --git a/tests/golden/legacy/trees_v1/more_languages_group4.txt b/tests/golden/legacy/trees_v1/more_languages_group4.txt
new file mode 100644
index 0000000..4a4641f
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group4.txt
@@ -0,0 +1,123 @@
+📁 group4 (1 folder, 10 files)
+├── 📄 erl_test.erl (480 tokens, 68 lines)
+├── 📄 haskell_test.hs (414 tokens, 41 lines)
+├── 📄 mathematica_test.nb (133 tokens, 21 lines)
+├── 📄 matlab_test.m (48 tokens, 12 lines)
+├── 📄 RTest.R (367 tokens, 46 lines)
+├── 📄 rust_test.rs (1,368 tokens, 259 lines)
+│   ├── fn at_beginning<'a>(&'a str)
+│   ├── pub enum Days<E: EdgeCase> {
+│   │       #\[default]
+│   │       Sun,
+│   │       Mon,
+│   │       #\[error("edge case {idx}, expected at least {} and at most {}",
+│   │   .limits.lo, .limits.hi)]
+│   │       Tue,
+│   │       Wed,
+│   │       Thu(i16, bool),
+│   │       Fri { day: u8 },
+│   │       Sat {
+│   │           urday: String,
+│   │           edge_case: E,
+│   │       },
+│   │   }
+│   ├── struct Point
+│   ├── impl Point
+│   ├──     fn get_origin() -> Point
+│   ├── struct Person
+│   ├── impl Person
+│   ├──     fn greet(&self)
+│   ├── fn add_two_longs(x1: i64, x2: i64) -> i64
+│   ├── fn add_two_longs_longer(
+│   │       x1: i64,
+│   │       x2: i64,
+│   │   ) -> i64
+│   ├── const fn multiply_by_two(num: f64) -> f64
+│   ├── fn get_first_character(s: &str) -> Option<char>
+│   ├── trait Drawable
+│   ├──     fn draw(&self)
+│   ├── impl Drawable for Point
+│   ├──     fn draw(&self)
+│   ├── fn with_generic<D: Drawable>(d: D)
+│   ├── fn with_generic<D>(d: D)
+│   │   where
+│   │       D: Drawable
+│   ├── fn main()
+│   ├── pub struct VisibleStruct
+│   ├── mod my_module
+│   ├──     pub struct AlsoVisibleStruct<T>(T, T)
+│   ├── macro_rules! say_hello
+│   ├── #[macro_export]
+│   │   macro_rules! hello_tree_plus
+│   ├── pub mod lib
+│   ├──     pub mod interfaces
+│   ├──     mod engine
+│   ├── pub fn flow<S1, S2, S3, S4, E, T, L>(
+│   │       source: S1,
+│   │       extractor: E,
+│   │       inbox: S2,
+│   │       transformer: T,
+│   │       outbox: S3,
+│   │       loader: L,
+│   │       sink: &mut S4,
+│   │   ) -> Result<(), Box<dyn Error>>
+│   │   where
+│   │       S1: Extractable,
+│   │       S2: Extractable + Loadable,
+│   │       S3: Extractable + Loadable,
+│   │       S4: Loadable,
+│   │       E: Extractor<S1, S2>,
+│   │       T: Transformer<S2, S3>,
+│   │       L: Loader<S3, S4>
+│   ├── trait Container
+│   ├──     fn items(&self) -> impl Iterator<Item = Widget>
+│   ├── trait HttpService
+│   ├──     async fn fetch(&self, url: Url) -> HtmlBody
+│   ├── struct Pair<T, U>
+│   ├── trait Transformer<T>
+│   ├──     fn transform(&self, input: T) -> T
+│   ├── impl<T: std::ops::Add<Output = T> + Copy> Transformer<T> for Pair<T, T>
+│   ├──     fn transform(&self, input: T) -> T
+│   ├── fn main()
+│   ├── async fn handle_get(State(pool): State<PgPool>) -> Result<Html<String>,
+│   │   (StatusCode, String)>
+│   │   where
+│   │       Bion: Cool
+│   ├── #[macro_export]
+│   │   macro_rules! unit
+│   ├──             fn insert(
+│   │                   &mut self,
+│   │                   key: (),
+│   │                   value: $unit_dtype,
+│   │               ) -> Result<Option<$unit_dtype>, ETLError>
+│   ├── pub async fn handle_get_axum_route(
+│   │       Session { maybe_claims }: Session,
+│   │       Path(RouteParams {
+│   │           alpha,
+│   │           bravo,
+│   │           charlie,
+│   │           edge_case
+│   │       }): Path<RouteParams>,
+│   │   ) -> ServerResult<Response>
+│   ├── fn encode_pipeline(cmds: &[Cmd], atomic: bool) -> Vec<u8>
+│   ├── pub async fn handle_post_yeet(
+│   │       State(auth_backend): State<AuthBackend>,
+│   │       Session { maybe_claims }: Session,
+│   │       Form(yeet_form): Form<YeetForm>,
+│   │   ) -> Result<Response, AuthError>
+│   └── pub async fn handle_get_thingy(
+│           session: Session,
+│           State(ApiBackend {
+│               page_cache,
+│               auth_backend,
+│               library_sql,
+│               some_data_cache,
+│               metadata_cache,
+│               thingy_client,
+│               ..
+│           }): State<ApiBackend>,
+│       ) -> ServerResult<Response>
+├── 📄 test.zig (397 tokens, 60 lines)
+├── 📄 test_fsharp.fs (92 tokens, 27 lines)
+├── 📄 test_tcl_tk.tcl (54 tokens, 16 lines)
+└── 📄 tf_test.tf (202 tokens, 38 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages_group5.txt b/tests/golden/legacy/trees_v1/more_languages_group5.txt
new file mode 100644
index 0000000..47b0a19
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group5.txt
@@ -0,0 +1,235 @@
+📁 group5 (1 folder, 19 files)
+├── 📄 ansible_test.yml (55 tokens, 14 lines)
+│   ├── Install package
+│   ├── Start service
+│   └── Create user
+├── 📄 app-routing.module.ts (287 tokens, 28 lines)
+│   ├── const routes: Routes = [
+│   │       { path: '', redirectTo: 'login', pathMatch: 'full' },
+│   │       { path: '*', redirectTo: 'login' },
+│   │       { path: 'home', component: HomeComponent },
+│   │       { path: 'login', component: LoginComponent },
+│   │       { path: 'register', component: RegisterComponent },
+│   │       { path: 'events', component: EventsComponent },
+│   │       { path: 'invites', component: InvitesComponent },
+│   │       { path: 'rewards', component: RewardsComponent },
+│   │       { path: 'profile', component: ProfileComponent },
+│   │   ];
+│   └── export class AppRoutingModule
+├── 📄 app.component.spec.ts (410 tokens, 47 lines)
+│   ├── describe 'AppComponent'
+│   ├──     it should create the app
+│   ├──     it should welcome the user
+│   ├──     it should welcome 'Jimbo'
+│   └──     it should request login if not logged in
+├── 📄 app.component.ts (271 tokens, 45 lines)
+│   ├── export class AppComponent
+│   ├──   constructor(
+│   │       private http: HttpClient,
+│   │       private loginService: LoginService,
+│   │       private stripeService: StripeService
+│   │     )
+│   ├──   constructor(private loginService: LoginService)
+│   ├──   checkSession()
+│   ├──   async goToEvent(event_id: string)
+│   └──   valInvitedBy(event: any, event_id: string)
+├── 📄 app.module.ts (374 tokens, 43 lines)
+│   ├── @NgModule({
+│   │       declarations: [
+│   │           AppComponent,
+│   │           HomeComponent,
+│   │           LoginComponent,
+│   │           RegisterComponent,
+│   │           EventsComponent,
+│   │           InvitesComponent,
+│   │           RewardsComponent,
+│   │           ProfileComponent
+│   └── export class AppModule
+├── 📄 checkbox_test.md (191 tokens, 29 lines)
+│   ├── # My Checkbox Test
+│   ├── ## My No Parens Test
+│   ├── ## My Empty href Test
+│   ├── ## My other url Test [Q&A]
+│   ├── ## My other other url Test [Q&A]
+│   ├── ## My 2nd other url Test [Q&A]
+│   ├── ## My 3rd other url Test [Q&A]
+│   ├── - [ ] Task 1
+│   ├──     - [ ] No Space Task 1.1
+│   ├──     - [ ] Two Spaces Task 1.2
+│   ├──         - [ ] Subtask 1.2.1
+│   ├── - [ ] Task 2
+│   ├── - [x] Task 3
+│   ├──     - [ ] Subtask 3.1
+│   ├── - [x] Task 6
+│   ├──     - [x] Subtask 6.1
+│   ├──         - [ ] Handle edge cases
+│   └── # My Codeblock Test
+├── 📄 checkbox_test.txt (257 tokens, 33 lines)
+│   ├── - [ ] fix phone number format +1
+│   ├── - [ ] add forgot password
+│   ├── - [ ] ? add email verification
+│   ├── - [ ] store token the right way
+│   ├── - [ ] test nesting of checkboxes
+│   ├── - [ ] user can use option to buy ticket at 2-referred price
+│   ├── - [ ] CTA refer 2 people to get instant lower price
+│   └── - [ ] form to send referrals
+├── 📄 environment.test.ts (197 tokens, 19 lines)
+│   ├── environment:
+│   ├──    production
+│   ├──    cognitoUserPoolId
+│   ├──    cognitoAppClientId
+│   └──    apiurl
+├── 📄 hello_world.pyi (22 tokens, 3 lines)
+│   ├── @final
+│   │   class dtype(Generic[_DTypeScalar_co])
+│   └──     names: None | tuple[builtins.str, ...]
+├── 📄 k8s_test.yaml (140 tokens, 37 lines)
+│   ├── apps/v1.Deployment -> my-app
+│   ├── v1.Service -> my-service
+│   └── v1.ConfigMap -> my-config
+├── 📄 Makefile (714 tokens, 84 lines)
+│   ├── include dotenv/dev.env
+│   ├── .PHONY: dev
+│   ├── dev
+│   ├── services-down
+│   ├── services-stop: services-down
+│   ├── define CHECK_POSTGRES
+│   ├── damage-report
+│   ├── tail-logs
+│   └── cloud
+├── 📄 requirements_test.txt (29 tokens, 10 lines)
+│   ├── psycopg2-binary
+│   ├── pytest
+│   ├── coverage
+│   ├── flask[async]
+│   ├── flask_cors
+│   ├── stripe
+│   ├── pyjwt[crypto]
+│   ├── cognitojwt[async]
+│   └── flask-lambda
+├── 📄 rust_todo_test.rs (92 tokens, 26 lines)
+│   ├── TODO: This todo tests parse_todo
+│   ├── enum Color {
+│   │       Red,
+│   │       Blue,
+│   │       Green,
+│   │   }
+│   ├── struct Point
+│   ├── trait Drawable
+│   ├──     fn draw(&self)
+│   ├── impl Drawable for Point
+│   ├──     fn draw(&self)
+│   └── fn main()
+├── 📄 sql_test.sql (270 tokens, 51 lines)
+├── 📄 standard-app-routing.module.ts (100 tokens, 16 lines)
+│   └── const routes: Routes = [
+│         { path: '', component: HomeComponent },
+│         {
+│           path: 'heroes',
+│           component: HeroesListComponent,
+│           children: [
+│             { path: ':id', component: HeroDetailComponent },
+│             { path: 'new', component: HeroFormComponent },
+│           ],
+│         },
+│         { path: '**', component: PageNotFoundComponent },
+│       ];
+├── 📄 test.env (190 tokens, 25 lines)
+│   ├── PROMO_PATH
+│   ├── PRODUCTION
+│   ├── SQL_SCHEMA_PATH
+│   ├── DB_LOGS
+│   ├── DB_LOG
+│   ├── PGPASSWORD
+│   ├── PGDATABASE
+│   ├── PGHOST
+│   ├── PGPORT
+│   ├── PGUSER
+│   ├── SERVER_LOG
+│   ├── SERVER_LOGS
+│   ├── API_URL
+│   ├── APP_LOGS
+│   ├── APP_LOG
+│   ├── APP_URL
+│   ├── COGNITO_USER_POOL_ID
+│   ├── COGNITO_APP_CLIENT_ID
+│   ├── AWS_REGION
+│   └── STRIPE_SECRET_KEY
+├── 📄 testJsonSchema.json (421 tokens, 48 lines)
+│   ├── $schema: http://json-schema.org/draft-07/schema#
+│   ├── type: object
+│   ├── title: random_test
+│   └── description: A promoter's activites related to events
+├── 📄 testPackage.json (349 tokens, 43 lines)
+│   ├── name: 'promo-app'
+│   ├── version: 0.0.0
+│   ├── scripts:
+│   ├──     ng: 'ng'
+│   ├──     start: 'ng serve'
+│   ├──     build: 'ng build'
+│   ├──     watch: 'ng build --watch --configuration development'
+│   └──     test: 'ng test'
+└── 📄 tickets.component.ts (7,160 tokens, 903 lines)
+    ├── interface EnrichedTicket extends Ticket
+    ├── interface SpinConfig
+    ├── interface RotationState
+    ├── interface SpeakInput
+    ├── const formatSpeakInput = (input: SpeakInput): string =>
+    ├── function hourToSpeech(hour: number, minute: number, period: string):
+    │   string
+    ├── export class TicketsComponent implements AfterViewInit
+    ├──   speak(input: SpeakInput)
+    ├──   speakEvent(ticket: EnrichedTicket): void
+    ├──   formatEvent(ticket: EnrichedTicket): string
+    ├──   speakVenue(ticket: EnrichedTicket): void
+    ├──   formatDate(date: Date, oneLiner: boolean = false): string
+    ├──   formatDateForSpeech(date: Date): string
+    ├──   async spinQRCode(
+    │       event: PointerEvent,
+    │       config: SpinConfig = DEFAULT_SPIN_CONFIG
+    │     )
+    ├──   private animateRotation(
+    │       imgElement: HTMLElement,
+    │       targetRotation: number,
+    │       config: SpinConfig,
+    │       cleanup: () => void
+    │     )
+    ├──     const animate = (currentTime: number) =>
+    ├──         requestAnimationFrame(animate)
+    ├──         cleanup()
+    ├──     requestAnimationFrame(animate)
+    ├──   private getNext90Degree(currentRotation: number): number
+    ├──   private getCurrentRotation(matrix: string): number
+    ├──   ngAfterViewInit()
+    ├──       const mouseEnterListener = () =>
+    ├──       const mouseLeaveListener = () =>
+    ├──   ngOnDestroy()
+    ├──   toggleColumn(event: MatOptionSelectionChange, column: string)
+    ├──   adjustColumns(event?: Event)
+    ├──   onResize(event: Event)
+    ├──   async ngOnInit()
+    ├──   async loadTickets(): Promise<void>
+    ├──   onDateRangeChange(
+    │       type: "start" | "end",
+    │       event: MatDatepickerInputEvent<Date>
+    │     )
+    ├──   applyFilter(column: string): void
+    ├──   formatDateForComparison(date: Date): string
+    ├──   constructor(private renderer: Renderer2)
+    ├──   onFilterChange(event: Event, column: string)
+    ├──   onLatitudeChange(event: Event)
+    ├──   onLongitudeChange(event: Event)
+    ├──   onRadiusChange(event: Event)
+    ├──   sortData(sort: Sort): void
+    ├──   onRowClick(event: Event, row: any)
+    ├── function isDate(value: Date | undefined | null): value is Date
+    ├── function isNonNullNumber(value: number | null): value is number
+    ├── function hasLocation(
+    │     ticket: any
+    │   ): ticket is
+    ├── const create_faker_ticket = async () =>
+    ├── function compare(a: number | string, b: number | string, isAsc: boolean)
+    ├── function compare_dates(a: Date, b: Date, isAsc: boolean)
+    ├── async function mockMoreTickets(): Promise<Ticket[]>
+    ├── const mockTickets = async () =>
+    └── const renderQRCode = async (text: String): Promise<string> =>
diff --git a/tests/golden/legacy/trees_v1/more_languages_group6.txt b/tests/golden/legacy/trees_v1/more_languages_group6.txt
new file mode 100644
index 0000000..6aa597b
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group6.txt
@@ -0,0 +1,300 @@
+📁 group6 (1 folder, 14 files)
+├── 📄 catastrophic.c (5,339 tokens, 754 lines)
+│   ├── TODO: technically we should use a proper parser
+│   ├── struct Point
+│   ├──     int x;
+│   ├──     int y;
+│   ├── struct Point getOrigin()
+│   ├── float mul_two_floats(float x1, float x2)
+│   ├── enum days
+│   ├──     SUN,
+│   ├──     MON,
+│   ├──     TUE,
+│   ├──     WED,
+│   ├──     THU,
+│   ├──     FRI,
+│   ├──     SAT
+│   ├── enum worker_pool_flags
+│   ├──         POOL_BH                 = 1 << 0,
+│   ├──         POOL_MANAGER_ACTIVE     = 1 << 1,
+│   ├──         POOL_DISASSOCIATED      = 1 << 2,
+│   ├──         POOL_BH_DRAINING        = 1 << 3,
+│   ├── enum worker_flags
+│   ├──         WORKER_DIE              = 1 << 1,
+│   ├──         WORKER_IDLE             = 1 << 2,
+│   ├──         WORKER_PREP             = 1 << 3,
+│   ├──         WORKER_CPU_INTENSIVE    = 1 << 6,
+│   ├──         WORKER_UNBOUND          = 1 << 7,
+│   ├──         WORKER_REBOUND          = 1 << 8,
+│   ├──         WORKER_NOT_RUNNING      = WORKER_PREP | WORKER_CPU_INTENSIVE |
+│   │                                     WORKER_UNBOUND | WORKER_REBOUND,
+│   ├── struct worker_pool
+│   ├──   raw_spinlock_t    lock;
+│   ├──   int      cpu;
+│   ├──   int      node;
+│   ├──   int      id;
+│   ├──   unsigned int    flags;
+│   ├──   unsigned long    watchdog_ts;
+│   ├──   bool      cpu_stall;
+│   ├──   int      nr_running;
+│   ├──   struct list_head  worklist;
+│   ├──   int      nr_workers;
+│   ├──   int      nr_idle;
+│   ├──   struct list_head  idle_list;
+│   ├──   struct timer_list  idle_timer;
+│   ├──   struct work_struct      idle_cull_work;
+│   ├──   struct timer_list  mayday_timer;
+│   ├──   struct worker    *manager;
+│   ├──   struct list_head  workers;
+│   ├──   struct ida    worker_ida;
+│   ├──   struct workqueue_attrs  *attrs;
+│   ├──   struct hlist_node  hash_node;
+│   ├──   int      refcnt;
+│   ├──   struct rcu_head    rcu;
+│   ├── long add_two_longs(long x1, long x2)
+│   ├── double multiplyByTwo(double num)
+│   ├── char getFirstCharacter(char *str)
+│   ├── void greet(Person p)
+│   ├── typedef struct
+│   ├──     char name[50];
+│   ├── } Person;
+│   ├── typedef struct PersonA
+│   ├──     char name[50];
+│   ├── } PersonB;
+│   ├── int main()
+│   ├── int* getArrayStart(int arr[], int size)
+│   ├── long complexFunctionWithMultipleArguments(
+│   │       int param1,
+│   │       double param2,
+│   │       char *param3,
+│   │       struct Point point
+│   │   )
+│   ├── keyPattern *ACLKeyPatternCreate(sds pattern, int flags)
+│   ├── sds sdsCatPatternString(sds base, keyPattern *pat)
+│   ├── static int ACLCheckChannelAgainstList(list *reference, const char
+│   │   *channel, int channellen, int is_pattern)
+│   ├──     while((ln = listNext(&li)))
+│   ├── static struct config
+│   ├──     aeEventLoop *el;
+│   ├──     cliConnInfo conn_info;
+│   ├──     const char *hostsocket;
+│   ├──     int tls;
+│   ├──     struct cliSSLconfig sslconfig;
+│   ├── } config;
+│   ├── class Person
+│   ├──     std::string name;
+│   ├── public:
+│   ├──     Person(std::string n) : name(n)
+│   ├──     void greet()
+│   ├── void globalGreet()
+│   ├── int main()
+│   ├── void printMessage(const std::string &message)
+│   ├── template<typename T>
+│   │   void printVector(const std::vector<T>& vec)
+│   ├── struct foo
+│   ├──     char x;
+│   ├──     struct foo_in
+│   ├──           char* y;
+│   ├──           short z;
+│   ├──     } inner;
+│   ├── struct Point
+│   ├──     int x, y;
+│   ├──     Point(int x, int y) : x(x), y(y)
+│   ├── class Animal
+│   ├──   public:
+│   ├──     Animal(const std::string &name) : name(name)
+│   ├──     virtual void speak() const
+│   ├──     virtual ~Animal()
+│   ├── protected:
+│   ├──     std::string name;
+│   ├── class Dog : public Animal
+│   ├──   public:
+│   ├──     Dog(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── class Cat : public Animal
+│   ├──   public:
+│   ├──     Cat(const std::string &name) : Animal(name)
+│   ├──     void speak() const override
+│   ├── class CatDog: public Animal, public Cat, public Dog
+│   ├──   public:
+│   ├──       CatDog(const std::string &name) : Animal(name)
+│   ├──       int meow_bark()
+│   ├── nb::bytes BuildRnnDescriptor(int input_size, int hidden_size, int
+│   │   num_layers,
+│   │                                int batch_size, int max_seq_length, float
+│   │   dropout,
+│   │                                bool bidirectional, bool cudnn_allow_tf32,
+│   │                                int workspace_size, int reserve_space_size)
+│   ├── int main()
+│   ├── enum ECarTypes
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV,
+│   ├──   Wagon
+│   ├── ECarTypes GetPreferredCarType()
+│   ├── enum ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── enum class ECarTypes : uint8_t
+│   ├──   Sedan,
+│   ├──   Hatchback,
+│   ├──   SUV = 254,
+│   ├──   Hybrid
+│   ├── void myFunction(string fname, int age)
+│   ├── template <typename T> T cos(T)
+│   ├── template <typename T> T sin(T)
+│   ├── template <typename T> T sqrt(T)
+│   ├── template<typename T> struct VLEN
+│   ├── template<typename T> class arr
+│   ├──   private:
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──     static T *ralloc(size_t num)
+│   ├──     static void dealloc(T *ptr)
+│   ├──   public:
+│   ├──     arr() : p(0), sz(0)
+│   ├──     arr(size_t n) : p(ralloc(n)), sz(n)
+│   ├──     arr(arr &&other)
+│   │         : p(other.p), sz(other.sz)
+│   ├──     ~arr()
+│   ├──     void resize(size_t n)
+│   ├──     T &operator[](size_t idx)
+│   ├──     T *data()
+│   ├──     size_t size() const
+│   ├── class Buffer
+│   ├──  private:
+│   ├──   void* ptr_;
+│   ├── std::tuple<array, array, array> quantize(
+│   │       const array& w,
+│   │       int group_size,
+│   │       int bits,
+│   │       StreamOrDevice s)
+│   ├── #define PY_SSIZE_T_CLEAN
+│   ├── #define PLATFORM_IS_X86
+│   ├── #define PLATFORM_WINDOWS
+│   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   ├── static int GetXCR0EAX()
+│   ├── #define GETCPUID(a, b, c, d, a_inp, c_inp)
+│   ├── static int GetXCR0EAX()
+│   ├──   asm("XGETBV" : "=a"(eax), "=d"(edx) : "c"(0))
+│   ├── static void ReportMissingCpuFeature(const char* name)
+│   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   ├── static PyObject *CheckCpuFeatures(PyObject *self, PyObject *args)
+│   ├── static PyMethodDef cpu_feature_guard_methods[]
+│   ├── static struct PyModuleDef cpu_feature_guard_module
+│   ├── #define EXPORT_SYMBOL __declspec(dllexport)
+│   ├── #define EXPORT_SYMBOL __attribute__ ((visibility("default")))
+│   ├── EXPORT_SYMBOL PyMODINIT_FUNC PyInit_cpu_feature_guard(void)
+│   ├── typedef struct
+│   ├──     GPT2Config config;
+│   ├──     ParameterTensors params;
+│   ├──     size_t param_sizes[NUM_PARAMETER_TENSORS];
+│   ├──     float* params_memory;
+│   ├──     size_t num_parameters;
+│   ├──     ParameterTensors grads;
+│   ├──     float* grads_memory;
+│   ├──     float* m_memory;
+│   ├──     float* v_memory;
+│   ├──     ActivationTensors acts;
+│   ├──     size_t act_sizes[NUM_ACTIVATION_TENSORS];
+│   ├──     float* acts_memory;
+│   ├──     size_t num_activations;
+│   ├──     ActivationTensors grads_acts;
+│   ├──     float* grads_acts_memory;
+│   ├──     int batch_size;
+│   ├──     int seq_len;
+│   ├──     int* inputs;
+│   ├──     int* targets;
+│   ├──     float mean_loss;
+│   └── } GPT2;
+├── 📄 cpp_examples_impl.cc (60 tokens, 10 lines)
+│   ├── PYBIND11_MODULE(cpp_examples, m)
+│   └──     m.def("add", &add<int>, "An example function to add two numbers.")
+├── 📄 cpp_examples_impl.cu (37 tokens, 10 lines)
+│   ├── template <typename T>
+│   │   T add(T a, T b)
+│   └── template <>
+│       int add<int>(int a, int b)
+├── 📄 cpp_examples_impl.h (22 tokens, 6 lines)
+│   ├── template <typename T>
+│   │   T add(T a, T b)
+│   └── template <>
+│       int add<int>(int, int)
+├── 📄 edge_case.hpp (426 tokens, 28 lines)
+├── 📄 fractal.thy (1,712 tokens, 147 lines)
+├── 📄 Microsoft.PowerShell_profile.ps1 (3,346 tokens, 497 lines)
+├── 📄 python_complex_class.py (10 tokens, 2 lines)
+│   └── class Box(Space[NDArray[Any]])
+├── 📄 ramda__cloneRegExp.js (173 tokens, 9 lines)
+│   └── export default function _cloneRegExp(pattern)
+├── 📄 ramda_prop.js (646 tokens, 85 lines)
+│   ├── /**
+│   │    * Returns a function that when supplied an object returns the indicated
+│   │    * property of that object, if it exists.
+│   │    * @category Object
+│   │    * @typedefn Idx = String | Int | Symbol
+│   │    * @sig Idx -> {s: a} -> a | Undefined
+│   │    * @param {String|Number} p The property name or array index
+│   │    * @param {Object} obj The object to query
+│   │    * @return {*} The value at `obj.p`.
+│   │    */
+│   │   var prop = _curry2(function prop(p, obj)
+│   ├── /**
+│   │    * Solves equations of the form a * x = b
+│   │    * @param {{
+│   │    *  z: number
+│   │    * }} x
+│   │    */
+│   │   function foo(x)
+│   ├── /**
+│   │    * Deconstructs an array field from the input documents to output a
+│   │   document for each element.
+│   │    * Each output document is the input document with the value of the
+│   │   array field replaced by the element.
+│   │    * @category Object
+│   │    * @sig String -> {k: [v]} -> [{k: v}]
+│   │    * @param {String} key The key to determine which property of the object
+│   │   should be unwound.
+│   │    * @param {Object} object The object containing the list to unwind at
+│   │   the property named by the key.
+│   │    * @return {List} A list of new objects, each having the given key
+│   │   associated to an item from the unwound list.
+│   │    */
+│   │   var unwind = _curry2(function(key, object)
+│   └──   return _map(function(item)
+├── 📄 tensorflow_flags.h (7,628 tokens, 668 lines)
+│   ├── #define TENSORFLOW_CORE_CONFIG_FLAG_DEFS_H_
+│   ├── class Flags
+│   ├──  public:
+│   ├── bool SetterForXlaAutoJitFlag(const string& value)
+│   ├── bool SetterForXlaCallModuleDisabledChecks(const string& value)
+│   ├── void AppendMarkForCompilationPassFlagsInternal(std::vector<Flag>*
+│   │   flag_list)
+│   ├── void AllocateAndParseJitRtFlags()
+│   ├── void AllocateAndParseFlags()
+│   ├── void ResetFlags()
+│   ├── bool SetXlaAutoJitFlagFromFlagString(const string& value)
+│   ├── BuildXlaOpsPassFlags* GetBuildXlaOpsPassFlags()
+│   ├── MarkForCompilationPassFlags* GetMarkForCompilationPassFlags()
+│   ├── XlaSparseCoreFlags* GetXlaSparseCoreFlags()
+│   ├── XlaDeviceFlags* GetXlaDeviceFlags()
+│   ├── XlaOpsCommonFlags* GetXlaOpsCommonFlags()
+│   ├── XlaCallModuleFlags* GetXlaCallModuleFlags()
+│   ├── MlirCommonFlags* GetMlirCommonFlags()
+│   ├── void ResetJitCompilerFlags()
+│   ├── const JitRtFlags& GetJitRtFlags()
+│   ├── ConfigProto::Experimental::MlirBridgeRollout GetMlirBridgeRolloutState(
+│   │       std::optional<const ConfigProto> config_proto)
+│   ├── void AppendMarkForCompilationPassFlags(std::vector<Flag>* flag_list)
+│   ├── void DisableXlaCompilation()
+│   ├── void EnableXlaCompilation()
+│   ├── bool FailOnXlaCompilation()
+│   ├── #define TF_PY_DECLARE_FLAG(flag_name)
+│   └── PYBIND11_MODULE(flags_pybind, m)
+├── 📄 test.f (181 tokens, 30 lines)
+├── 📄 torch.rst (60 tokens, 8 lines)
+│   ├── # libtorch (C++-only)
+│   └── - Building libtorch using Python
+└── 📄 yc.html (9,063 tokens, 169 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages_group7.txt b/tests/golden/legacy/trees_v1/more_languages_group7.txt
new file mode 100644
index 0000000..2f1fcc3
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group7.txt
@@ -0,0 +1,83 @@
+📁 group7 (1 folder, 5 files)
+├── 📄 absurdly_huge.jsonl (8,347 tokens, 126 lines)
+│   ├── SMILES: str
+│   ├── Yield: float
+│   ├── Temperature: int
+│   ├── Pressure: float
+│   ├── Solvent: str
+│   ├── Success: bool
+│   ├── Reaction_Conditions: dict
+│   ├── Products: list
+│   └── EdgeCasesMissed: None
+├── 📄 angular_crud.ts (1,192 tokens, 148 lines)
+│   ├── interface DBCommand<T = any>
+│   ├── export class IndexedDbService
+│   ├──     constructor()
+│   ├──     async create_connection({ db_name = 'client_db', table_name }:
+│   │   DBCommand)
+│   ├──                 upgrade(db)
+│   ├──     async create_model({ db_name, table_name, model }: DBCommand)
+│   ├──         verify_matching({ table_name, model })
+│   ├──     async read_key({ db_name, table_name, key }: DBCommand)
+│   ├──     async update_model({ db_name, table_name, model }: DBCommand)
+│   ├──         verify_matching({ table_name, model })
+│   ├──     async delete_key({ db_name, table_name, key }: DBCommand)
+│   ├──     async list_table({
+│   │           db_name,
+│   │           table_name,
+│   │           where,
+│   │       }: DBCommand & { where?: { [key: string]: string | number } })
+│   └──     async search_table(criteria: SearchCriteria)
+├── 📄 structure.py (400 tokens, 92 lines)
+│   ├── @runtime_checkable
+│   │   class DataClass(Protocol)
+│   ├──     __dataclass_fields__: dict
+│   ├── class MyInteger(Enum)
+│   ├──     ONE = 1
+│   ├──     TWO = 2
+│   ├──     THREE = 42
+│   ├── class MyString(Enum)
+│   ├──     AAA1 = "aaa"
+│   ├──     BB_B = """edge
+│   │   case"""
+│   ├── @dataclass(frozen=True, slots=True, kw_only=True)
+│   │   class Tool
+│   ├──     name: str
+│   ├──     description: str
+│   ├──     input_model: DataClass
+│   ├──     output_model: DataClass
+│   ├──     def execute(self, *args, **kwargs)
+│   ├──     @property
+│   │       def edge_case(self) -> str
+│   ├──     def should_still_see_me(self, x: bool = True) -> "Tool"
+│   ├── @dataclass
+│   │   class MyInput[T]
+│   ├──     name: str
+│   ├──     rank: MyInteger
+│   ├──     serial_n: int
+│   ├── @dataclass
+│   │   class Thingy
+│   ├──     is_edge_case: bool
+│   ├── @dataclass
+│   │   class MyOutput
+│   ├──     orders: str
+│   ├── class MyTools(Enum)
+│   ├──     TOOL_A = Tool(
+│   │           name="complicated",
+│   │           description="edge case!",
+│   │           input_model=MyInput[Thingy],
+│   │           output_model=MyOutput,
+│   │       )
+│   ├──     TOOL_B = Tool(
+│   │           name="""super
+│   │   complicated
+│   │   """,
+│   │           description="edge case!",
+│   │           input_model=MyInput,
+│   │           output_model=MyOutput,
+│   │       )
+│   ├── @final
+│   │   class dtype(Generic[_DTypeScalar_co])
+│   └──     names: None | tuple[builtins.str, ...]
+├── 📄 test.wgsl (528 tokens, 87 lines)
+└── 📄 test.metal (272 tokens, 34 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages_group_lisp.txt b/tests/golden/legacy/trees_v1/more_languages_group_lisp.txt
new file mode 100644
index 0000000..fefeb36
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group_lisp.txt
@@ -0,0 +1,5 @@
+📁 group_lisp (1 folder, 4 files)
+├── 📄 clojure_test.clj (682 tokens, 85 lines)
+├── 📄 LispTest.lisp (25 tokens, 6 lines)
+├── 📄 racket_struct.rkt (14 tokens, 1 line)
+└── 📄 test_scheme.scm (360 tokens, 44 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages_group_todo.txt b/tests/golden/legacy/trees_v1/more_languages_group_todo.txt
new file mode 100644
index 0000000..d13b512
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/more_languages_group_todo.txt
@@ -0,0 +1,13 @@
+📁 group_todo (1 folder, 12 files)
+├── 📄 AAPLShaders.metal (5,780 tokens, 566 lines)
+├── 📄 crystal_test.cr (48 tokens, 15 lines)
+├── 📄 dart_test.dart (108 tokens, 24 lines)
+├── 📄 elixir_test.exs (39 tokens, 10 lines)
+├── 📄 forward.frag (739 tokens, 87 lines)
+├── 📄 forward.vert (359 tokens, 48 lines)
+├── 📄 nodemon.json (118 tokens, 20 lines)
+├── 📄 sas_test.sas (97 tokens, 22 lines)
+├── 📄 test_setup_py.test (133 tokens, 24 lines)
+├── 📄 testTypings.d.ts (158 tokens, 23 lines)
+├── 📄 vba_test.bas (67 tokens, 16 lines)
+└── 📄 wgsl_test.wgsl (94 tokens, 17 lines)
diff --git a/tests/golden/legacy/trees_v1/multi_seed.txt b/tests/golden/legacy/trees_v1/multi_seed.txt
new file mode 100644
index 0000000..c23221a
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/multi_seed.txt
@@ -0,0 +1,199 @@
+🌵 Root (2 folders, 17 files)
+├── 📁 group1 (1 folder, 11 files)
+│   ├── 📄 addamt.cobol (441 tokens, 40 lines)
+│   ├── 📄 CUSTOMER-INVOICE.CBL (412 tokens, 60 lines)
+│   ├── 📄 JavaTest.java (578 tokens, 86 lines)
+│   ├── 📄 JuliaTest.jl (381 tokens, 63 lines)
+│   ├── 📄 KotlinTest.kt (974 tokens, 171 lines)
+│   ├── 📄 lesson.cbl (635 tokens, 78 lines)
+│   ├── 📄 LuaTest.lua (83 tokens, 16 lines)
+│   ├── 📄 ObjectiveCTest.m (62 tokens, 16 lines)
+│   ├── 📄 OcamlTest.ml (49 tokens, 12 lines)
+│   ├── 📄 test.js (757 tokens, 154 lines)
+│   │   ├── class MyClass
+│   │   ├──   myMethod()
+│   │   ├──   async asyncMethod(a, b)
+│   │   ├──   methodWithDefaultParameters(a = 5, b = 10)
+│   │   ├──   multilineMethod(
+│   │   │       c,
+│   │   │       d
+│   │   │     )
+│   │   ├──   multilineMethodWithDefaults(
+│   │   │       t = "tree",
+│   │   │       p = "plus"
+│   │   │     )
+│   │   ├── function myFunction(param1, param2)
+│   │   ├── function multilineFunction(
+│   │   │     param1,
+│   │   │     param2
+│   │   │   )
+│   │   ├── const arrowFunction = () =>
+│   │   ├── const parametricArrow = (a, b) =>
+│   │   ├── function ()
+│   │   ├── function outerFunction(outerParam)
+│   │   ├──   function innerFunction(innerParam)
+│   │   ├──   innerFunction("inner")
+│   │   ├── const myObject = {
+│   │   ├──   myMethod: function (stuff)
+│   │   ├── let myArrowObject = {
+│   │   ├──   myArrow: ({
+│   │   │       a,
+│   │   │       b,
+│   │   │       c,
+│   │   │     }) =>
+│   │   ├── const myAsyncArrowFunction = async () =>
+│   │   ├── function functionWithRestParameters(...args)
+│   │   ├── const namedFunctionExpression = function myNamedFunction()
+│   │   ├── const multilineArrowFunction = (
+│   │   │     a,
+│   │   │     b
+│   │   │   ) =>
+│   │   ├── function functionReturningFunction()
+│   │   ├──   return function ()
+│   │   ├── function destructuringOnMultipleLines({
+│   │   │     a,
+│   │   │     b,
+│   │   │   })
+│   │   ├── const arrowFunctionWithDestructuring = ({ a, b }) =>
+│   │   ├── const multilineDestructuringArrow = ({
+│   │   │     a,
+│   │   │     b,
+│   │   │   }) =>
+│   │   ├── async function asyncFunctionWithErrorHandling()
+│   │   ├── class Car
+│   │   ├──   constructor(brand)
+│   │   ├──   present()
+│   │   ├── class Model extends Car
+│   │   ├──   constructor(brand, mod)
+│   │   ├──     super(brand)
+│   │   └──   show()
+│   └── 📄 test.ts (832 tokens, 165 lines)
+│       ├── type MyType
+│       ├── interface MyInterface
+│       ├── class TsClass
+│       ├──   myMethod()
+│       ├──   myMethodWithArgs(param1: string, param2: number): void
+│       ├──   static myStaticMethod<T>(param: T): T
+│       ├──   multilineMethod(
+│       │       c: number,
+│       │       d: number
+│       │     ): number
+│       ├──   multilineMethodWithDefaults(
+│       │       t: string = "tree",
+│       │       p: string = "plus"
+│       │     ): string
+│       ├── export class AdvancedComponent implements MyInterface
+│       ├──   async myAsyncMethod(
+│       │       a: string,
+│       │       b: number,
+│       │       c: string
+│       │     ): Promise<void>
+│       ├──   genericMethod<T, U>(
+│       │       arg1: T,
+│       │       arg2: U
+│       │     ): [T, U]
+│       ├── export class TicketsComponent implements MyInterface
+│       ├──   async myAsyncMethod({ a, b, c }: { a: String; b: Number; c: String
+│       │   })
+│       ├── function tsFunction()
+│       ├── function tsFunctionSigned(
+│       │     param1: number,
+│       │     param2: number
+│       │   ): void
+│       ├── export default async function tsFunctionComplicated<A, B, C>({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string>
+│       ├── const tsArrowFunctionSigned = ({
+│       │     a,
+│       │     b,
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │   }) =>
+│       ├── export const tsComplicatedArrow = async ({
+│       │     a = 1 | 2,
+│       │     b = "bob",
+│       │     c = async () => "charlie",
+│       │   }: {
+│       │     a: number;
+│       │     b: string;
+│       │     c: () => Promise<string>;
+│       │   }): Promise<string> =>
+│       ├── const arrowFunction = () =>
+│       ├── const arrow = (a: String, b: Number) =>
+│       ├── const asyncArrowFunction = async () =>
+│       ├── const asyncArrow = async (a: String, b: Number) =>
+│       ├── let weirdArrow = () =>
+│       ├── const asyncPromiseArrow = async (): Promise<void> =>
+│       ├── let myWeirdArrowSigned = (x: number): number =>
+│       ├── class Person
+│       ├──   constructor(private firstName: string, private lastName: string)
+│       ├──   getFullName(): string
+│       ├──   describe(): string
+│       ├── class Employee extends Person
+│       ├──   constructor(
+│       │       firstName: string,
+│       │       lastName: string,
+│       │       private jobTitle: string
+│       │     )
+│       ├──     super(firstName, lastName)
+│       ├──   describe(): string
+│       ├── interface Shape
+│       └── interface Square extends Shape
+└── 📁 path_to_test (1 folder, 6 files)
+    ├── 📄 class_method_type.py (525 tokens, 101 lines)
+    │   ├── T = TypeVar("T")
+    │   ├── def parse_py(contents: str) -> List[str]
+    │   ├── class MyClass
+    │   ├──     @staticmethod
+    │   │       def physical_element_aval(dtype) -> core.ShapedArray
+    │   ├──     def my_method(self)
+    │   ├──     @staticmethod
+    │   │       def my_typed_method(obj: dict) -> int
+    │   ├──     def my_multiline_signature_method(
+    │   │           self,
+    │   │           alice: str = None,
+    │   │           bob: int = None,
+    │   │       ) -> tuple
+    │   ├── @lru_cache(maxsize=None)
+    │   │   def my_multiline_signature_function(
+    │   │       tree: tuple = (),
+    │   │       plus: str = "+",
+    │   │   ) -> tuple
+    │   ├── class LogLevelEnum(str, Enum)
+    │   ├──     CRITICAL = "CRITICAL"
+    │   ├──     GREETING = "GREETING"
+    │   ├──     WARNING = "WARNING"
+    │   ├──     ERROR = "ERROR"
+    │   ├──     DEBUG = "DEBUG"
+    │   ├──     INFO = "INFO"
+    │   ├──     OFF = "OFF"
+    │   ├── class Thingy(BaseModel)
+    │   ├──     metric: float
+    │   ├── @dataclass
+    │   │   class TestDataclass
+    │   ├──     tree: str
+    │   ├── A = TypeVar("A", str, bytes)
+    │   ├── def omega_yikes(file: str, expected: List[str]) -> bool
+    │   ├── def ice[T](args: Iterable[T] = ())
+    │   ├── class list[T]
+    │   ├──     def __getitem__(self, index: int, /) -> T
+    │   ├──     @classmethod
+    │   │       def from_code(cls, toolbox, code: bytes, score=None) -> "Thingy"
+    │   ├──     @classmethod
+    │   │       def from_str(cls, toolbox, string: str, score=None) -> "Thingy"
+    │   └── class Router(hk.Module)
+    ├── 📄 empty.py (0 tokens, 0 lines)
+    ├── 📄 file.md (11 tokens, 2 lines)
+    │   └── # Hello, world!
+    ├── 📄 file.py (18 tokens, 3 lines)
+    │   └── def hello_world()
+    ├── 📄 file.txt (10 tokens, 2 lines)
+    └── 📄 version.py (13 tokens, 2 lines)
+        └── __version__ = "1.2.3"
diff --git a/tests/golden/legacy/trees_v1/path_to_test.txt b/tests/golden/legacy/trees_v1/path_to_test.txt
new file mode 100644
index 0000000..c1f5bc7
--- /dev/null
+++ b/tests/golden/legacy/trees_v1/path_to_test.txt
@@ -0,0 +1,51 @@
+📁 path_to_test (1 folder, 6 files)
+├── 📄 class_method_type.py (525 tokens, 101 lines)
+│   ├── T = TypeVar("T")
+│   ├── def parse_py(contents: str) -> List[str]
+│   ├── class MyClass
+│   ├──     @staticmethod
+│   │       def physical_element_aval(dtype) -> core.ShapedArray
+│   ├──     def my_method(self)
+│   ├──     @staticmethod
+│   │       def my_typed_method(obj: dict) -> int
+│   ├──     def my_multiline_signature_method(
+│   │           self,
+│   │           alice: str = None,
+│   │           bob: int = None,
+│   │       ) -> tuple
+│   ├── @lru_cache(maxsize=None)
+│   │   def my_multiline_signature_function(
+│   │       tree: tuple = (),
+│   │       plus: str = "+",
+│   │   ) -> tuple
+│   ├── class LogLevelEnum(str, Enum)
+│   ├──     CRITICAL = "CRITICAL"
+│   ├──     GREETING = "GREETING"
+│   ├──     WARNING = "WARNING"
+│   ├──     ERROR = "ERROR"
+│   ├──     DEBUG = "DEBUG"
+│   ├──     INFO = "INFO"
+│   ├──     OFF = "OFF"
+│   ├── class Thingy(BaseModel)
+│   ├──     metric: float
+│   ├── @dataclass
+│   │   class TestDataclass
+│   ├──     tree: str
+│   ├── A = TypeVar("A", str, bytes)
+│   ├── def omega_yikes(file: str, expected: List[str]) -> bool
+│   ├── def ice[T](args: Iterable[T] = ())
+│   ├── class list[T]
+│   ├──     def __getitem__(self, index: int, /) -> T
+│   ├──     @classmethod
+│   │       def from_code(cls, toolbox, code: bytes, score=None) -> "Thingy"
+│   ├──     @classmethod
+│   │       def from_str(cls, toolbox, string: str, score=None) -> "Thingy"
+│   └── class Router(hk.Module)
+├── 📄 empty.py (0 tokens, 0 lines)
+├── 📄 file.md (11 tokens, 2 lines)
+│   └── # Hello, world!
+├── 📄 file.py (18 tokens, 3 lines)
+│   └── def hello_world()
+├── 📄 file.txt (10 tokens, 2 lines)
+└── 📄 version.py (13 tokens, 2 lines)
+    └── __version__ = "1.2.3"

From 5b41a12366d2f07b372394e148c2349518adc5bd Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 15:48:03 -0400
Subject: [PATCH 3/8] feat: Rust port of tree_plus (workspace: core + cli)

tree_plus_core: model, deterministic rich-compatible renderer, natural
sort (natsort os_sorted parity), fnmatch-compatible ignores, amortized
globs, wc-parity counting, and component extraction:
- tree-sitter extractors for Rust, Python, JS/TS, C/C++ with formatters
  that reproduce the legacy output (golden-parity tested, including
  catastrophic.c) and salvage ERROR regions on invalid syntax
- regex extractors (LazyLock, linear-time engine only) for markers,
  Markdown/RST/txt, .env, requirements, Makefile/Justfile, Angular
- native parsers for JSON/JSONL/YAML/TOML/CSV and SQLite (feature)

tree_plus_cli: clap binary as tree_plus + tprs alias (PATH collision
with the Python entry point), legacy flags, footer parity, real
terminal-width detection.

Parallel traversal with rayon behind a deterministic sort; arbitrary
bytes never panic (robustness suite); criterion benches included.
55-85x faster than the Python implementation end to end.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 .gitignore                                    |    4 +-
 Cargo.lock                                    | 1096 +++++++++++++++++
 Cargo.toml                                    |   29 +
 crates/tree_plus_cli/Cargo.toml               |   21 +
 crates/tree_plus_cli/src/main.rs              |  173 +++
 crates/tree_plus_cli/tests/cli.rs             |   92 ++
 crates/tree_plus_core/Cargo.toml              |   36 +
 .../tree_plus_core/benches/tree_plus_bench.rs |   78 ++
 crates/tree_plus_core/examples/dump_ast.rs    |   55 +
 crates/tree_plus_core/examples/extract.rs     |   16 +
 crates/tree_plus_core/src/config.rs           |   39 +
 crates/tree_plus_core/src/count.rs            |  203 +++
 crates/tree_plus_core/src/extract/data.rs     |  582 +++++++++
 crates/tree_plus_core/src/extract/markdown.rs |  180 +++
 crates/tree_plus_core/src/extract/markers.rs  |   60 +
 crates/tree_plus_core/src/extract/mod.rs      |  277 +++++
 crates/tree_plus_core/src/extract/simple.rs   |  216 ++++
 .../src/extract/treesitter/c_cpp.rs           |  591 +++++++++
 .../src/extract/treesitter/mod.rs             |   66 +
 .../src/extract/treesitter/python.rs          |  346 ++++++
 .../src/extract/treesitter/rust.rs            |  312 +++++
 .../src/extract/treesitter/typescript.rs      |  420 +++++++
 crates/tree_plus_core/src/ignore.rs           |  307 +++++
 crates/tree_plus_core/src/lib.rs              |   29 +
 crates/tree_plus_core/src/model.rs            |  125 ++
 crates/tree_plus_core/src/render.rs           |  347 ++++++
 crates/tree_plus_core/src/sort.rs             |  214 ++++
 crates/tree_plus_core/src/walk.rs             |  260 ++++
 crates/tree_plus_core/tests/golden_parity.rs  |  243 ++++
 crates/tree_plus_core/tests/robustness.rs     |   86 ++
 30 files changed, 6502 insertions(+), 1 deletion(-)
 create mode 100644 Cargo.lock
 create mode 100644 Cargo.toml
 create mode 100644 crates/tree_plus_cli/Cargo.toml
 create mode 100644 crates/tree_plus_cli/src/main.rs
 create mode 100644 crates/tree_plus_cli/tests/cli.rs
 create mode 100644 crates/tree_plus_core/Cargo.toml
 create mode 100644 crates/tree_plus_core/benches/tree_plus_bench.rs
 create mode 100644 crates/tree_plus_core/examples/dump_ast.rs
 create mode 100644 crates/tree_plus_core/examples/extract.rs
 create mode 100644 crates/tree_plus_core/src/config.rs
 create mode 100644 crates/tree_plus_core/src/count.rs
 create mode 100644 crates/tree_plus_core/src/extract/data.rs
 create mode 100644 crates/tree_plus_core/src/extract/markdown.rs
 create mode 100644 crates/tree_plus_core/src/extract/markers.rs
 create mode 100644 crates/tree_plus_core/src/extract/mod.rs
 create mode 100644 crates/tree_plus_core/src/extract/simple.rs
 create mode 100644 crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
 create mode 100644 crates/tree_plus_core/src/extract/treesitter/mod.rs
 create mode 100644 crates/tree_plus_core/src/extract/treesitter/python.rs
 create mode 100644 crates/tree_plus_core/src/extract/treesitter/rust.rs
 create mode 100644 crates/tree_plus_core/src/extract/treesitter/typescript.rs
 create mode 100644 crates/tree_plus_core/src/ignore.rs
 create mode 100644 crates/tree_plus_core/src/lib.rs
 create mode 100644 crates/tree_plus_core/src/model.rs
 create mode 100644 crates/tree_plus_core/src/render.rs
 create mode 100644 crates/tree_plus_core/src/sort.rs
 create mode 100644 crates/tree_plus_core/src/walk.rs
 create mode 100644 crates/tree_plus_core/tests/golden_parity.rs
 create mode 100644 crates/tree_plus_core/tests/robustness.rs

diff --git a/.gitignore b/.gitignore
index 8ce7999..775e954 100644
--- a/.gitignore
+++ b/.gitignore
@@ -55,4 +55,6 @@ tests/version_increments/renamed_dry_run_version.py
 .mcp_server.pid
 
 # using uv sometimes we get a build folder
-build/
\ No newline at end of file
+build/
+# Rust workspace
+/target/
diff --git a/Cargo.lock b/Cargo.lock
new file mode 100644
index 0000000..1510d6f
--- /dev/null
+++ b/Cargo.lock
@@ -0,0 +1,1096 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 4
+
+[[package]]
+name = "ahash"
+version = "0.8.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5a15f179cd60c4584b8a8c596927aadc462e27f2ca70c04e0071964a73ba7a75"
+dependencies = [
+ "cfg-if",
+ "once_cell",
+ "version_check",
+ "zerocopy",
+]
+
+[[package]]
+name = "aho-corasick"
+version = "1.1.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ddd31a130427c27518df266943a5308ed92d4b226cc639f5a8f1002816174301"
+dependencies = [
+ "memchr",
+]
+
+[[package]]
+name = "anes"
+version = "0.1.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4b46cbb362ab8752921c97e041f5e366ee6297bd428a31275b9fcf1e380f7299"
+
+[[package]]
+name = "anstream"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "824a212faf96e9acacdbd09febd34438f8f711fb84e09a8916013cd7815ca28d"
+dependencies = [
+ "anstyle",
+ "anstyle-parse",
+ "anstyle-query",
+ "anstyle-wincon",
+ "colorchoice",
+ "is_terminal_polyfill",
+ "utf8parse",
+]
+
+[[package]]
+name = "anstyle"
+version = "1.0.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "940b3a0ca603d1eade50a4846a2afffd5ef57a9feac2c0e2ec2e14f9ead76000"
+
+[[package]]
+name = "anstyle-parse"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "52ce7f38b242319f7cabaa6813055467063ecdc9d355bbb4ce0c68908cd8130e"
+dependencies = [
+ "utf8parse",
+]
+
+[[package]]
+name = "anstyle-query"
+version = "1.1.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "40c48f72fd53cd289104fc64099abca73db4166ad86ea0b4341abe65af83dadc"
+dependencies = [
+ "windows-sys",
+]
+
+[[package]]
+name = "anstyle-wincon"
+version = "3.0.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "291e6a250ff86cd4a820112fb8898808a366d8f9f58ce16d1f538353ad55747d"
+dependencies = [
+ "anstyle",
+ "once_cell_polyfill",
+ "windows-sys",
+]
+
+[[package]]
+name = "autocfg"
+version = "1.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f2032f911046de80f0a198e0901378627c33f59ea0ac00e363d481118bd70a53"
+
+[[package]]
+name = "bitflags"
+version = "2.13.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b4388bee8683e3d04af747c73422af53102d2bd24d9eadb6cbc100baef4b43f8"
+
+[[package]]
+name = "bumpalo"
+version = "3.20.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "72f5acc6cb2ba439de613abc23857ec3d78374d8ed5ac84e9d11336e87da8649"
+
+[[package]]
+name = "cast"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "37b2a672a2cb129a2e41c10b1224bb368f9f37a2b16b612598138befd7b37eb5"
+
+[[package]]
+name = "cc"
+version = "1.2.63"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "556e016178bb5662a08681bbe0f00f8e17631781a4dfc8c45e466e4b185ec27f"
+dependencies = [
+ "find-msvc-tools",
+ "shlex",
+]
+
+[[package]]
+name = "cfg-if"
+version = "1.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9330f8b2ff13f34540b44e946ef35111825727b38d33286ef986142615121801"
+
+[[package]]
+name = "ciborium"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "42e69ffd6f0917f5c029256a24d0161db17cea3997d185db0d35926308770f0e"
+dependencies = [
+ "ciborium-io",
+ "ciborium-ll",
+ "serde",
+]
+
+[[package]]
+name = "ciborium-io"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "05afea1e0a06c9be33d539b876f1ce3692f4afea2cb41f740e7743225ed1c757"
+
+[[package]]
+name = "ciborium-ll"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "57663b653d948a338bfb3eeba9bb2fd5fcfaecb9e199e87e1eda4d9e8b240fd9"
+dependencies = [
+ "ciborium-io",
+ "half",
+]
+
+[[package]]
+name = "clap"
+version = "4.6.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1ddb117e43bbf7dacf0a4190fef4d345b9bad68dfc649cb349e7d17d28428e51"
+dependencies = [
+ "clap_builder",
+ "clap_derive",
+]
+
+[[package]]
+name = "clap_builder"
+version = "4.6.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "714a53001bf66416adb0e2ef5ac857140e7dc3a0c48fb28b2f10762fc4b5069f"
+dependencies = [
+ "anstream",
+ "anstyle",
+ "clap_lex",
+ "strsim",
+]
+
+[[package]]
+name = "clap_derive"
+version = "4.6.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f2ce8604710f6733aa641a2b3731eaa1e8b3d9973d5e3565da11800813f997a9"
+dependencies = [
+ "heck",
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "clap_lex"
+version = "1.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c8d4a3bb8b1e0c1050499d1815f5ab16d04f0959b233085fb31653fbfc9d98f9"
+
+[[package]]
+name = "colorchoice"
+version = "1.0.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1d07550c9036bf2ae0c684c4297d503f838287c83c53686d05370d0e139ae570"
+
+[[package]]
+name = "criterion"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f2b12d017a929603d80db1831cd3a24082f8137ce19c69e6447f54f5fc8d692f"
+dependencies = [
+ "anes",
+ "cast",
+ "ciborium",
+ "clap",
+ "criterion-plot",
+ "is-terminal",
+ "itertools",
+ "num-traits",
+ "once_cell",
+ "oorandom",
+ "plotters",
+ "rayon",
+ "regex",
+ "serde",
+ "serde_derive",
+ "serde_json",
+ "tinytemplate",
+ "walkdir",
+]
+
+[[package]]
+name = "criterion-plot"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6b50826342786a51a89e2da3a28f1c32b06e387201bc2d19791f622c673706b1"
+dependencies = [
+ "cast",
+ "itertools",
+]
+
+[[package]]
+name = "crossbeam-deque"
+version = "0.8.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9dd111b7b7f7d55b72c0a6ae361660ee5853c9af73f70c3c2ef6858b950e2e51"
+dependencies = [
+ "crossbeam-epoch",
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "crossbeam-epoch"
+version = "0.9.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5b82ac4a3c2ca9c3460964f020e1402edd5753411d7737aa39c3714ad1b5420e"
+dependencies = [
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "crossbeam-utils"
+version = "0.8.21"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d0a5c400df2834b80a4c3327b3aad3a4c4cd4de0629063962b03235697506a28"
+
+[[package]]
+name = "crunchy"
+version = "0.2.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "460fbee9c2c2f33933d720630a6a0bac33ba7053db5344fac858d4b8952d77d5"
+
+[[package]]
+name = "either"
+version = "1.16.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "91622ff5e7162018101f2fea40d6ebf4a78bbe5a49736a2020649edf9693679e"
+
+[[package]]
+name = "equivalent"
+version = "1.0.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "877a4ace8713b0bcf2a4e7eec82529c029f1d0619886d18145fea96c3ffe5c0f"
+
+[[package]]
+name = "errno"
+version = "0.3.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "39cab71617ae0d63f51a36d69f866391735b51691dbda63cf6f96d042b63efeb"
+dependencies = [
+ "libc",
+ "windows-sys",
+]
+
+[[package]]
+name = "fallible-iterator"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2acce4a10f12dc2fb14a218589d4f1f62ef011b2d0cc4b3cb1bba8e94da14649"
+
+[[package]]
+name = "fallible-streaming-iterator"
+version = "0.1.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7360491ce676a36bf9bb3c56c1aa791658183a54d2744120f27285738d90465a"
+
+[[package]]
+name = "find-msvc-tools"
+version = "0.1.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5baebc0774151f905a1a2cc41989300b1e6fbb29aff0ceffa1064fdd3088d582"
+
+[[package]]
+name = "futures-core"
+version = "0.3.32"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7e3450815272ef58cec6d564423f6e755e25379b217b0bc688e295ba24df6b1d"
+
+[[package]]
+name = "futures-task"
+version = "0.3.32"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "037711b3d59c33004d3856fbdc83b99d4ff37a24768fa1be9ce3538a1cde4393"
+
+[[package]]
+name = "futures-util"
+version = "0.3.32"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "389ca41296e6190b48053de0321d02a77f32f8a5d2461dd38762c0593805c6d6"
+dependencies = [
+ "futures-core",
+ "futures-task",
+ "pin-project-lite",
+ "slab",
+]
+
+[[package]]
+name = "half"
+version = "2.7.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6ea2d84b969582b4b1864a92dc5d27cd2b77b622a8d79306834f1be5ba20d84b"
+dependencies = [
+ "cfg-if",
+ "crunchy",
+ "zerocopy",
+]
+
+[[package]]
+name = "hashbrown"
+version = "0.14.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e5274423e17b7c9fc20b6e7e208532f9b19825d82dfd615708b70edd83df41f1"
+dependencies = [
+ "ahash",
+]
+
+[[package]]
+name = "hashbrown"
+version = "0.17.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ed5909b6e89a2db4456e54cd5f673791d7eca6732202bbf2a9cc504fe2f9b84a"
+
+[[package]]
+name = "hashlink"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6ba4ff7128dee98c7dc9794b6a411377e1404dba1c97deb8d1a55297bd25d8af"
+dependencies = [
+ "hashbrown 0.14.5",
+]
+
+[[package]]
+name = "heck"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
+
+[[package]]
+name = "hermit-abi"
+version = "0.5.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fc0fef456e4baa96da950455cd02c081ca953b141298e41db3fc7e36b1da849c"
+
+[[package]]
+name = "indexmap"
+version = "2.14.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d466e9454f08e4a911e14806c24e16fba1b4c121d1ea474396f396069cf949d9"
+dependencies = [
+ "equivalent",
+ "hashbrown 0.17.1",
+]
+
+[[package]]
+name = "is-terminal"
+version = "0.4.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3640c1c38b8e4e43584d8df18be5fc6b0aa314ce6ebf51b53313d4306cca8e46"
+dependencies = [
+ "hermit-abi",
+ "libc",
+ "windows-sys",
+]
+
+[[package]]
+name = "is_terminal_polyfill"
+version = "1.70.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a6cb138bb79a146c1bd460005623e142ef0181e3d0219cb493e02f7d08a35695"
+
+[[package]]
+name = "itertools"
+version = "0.10.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b0fd2260e829bddf4cb6ea802289de2f86d6a7a690192fbe91b3f46e0f2c8473"
+dependencies = [
+ "either",
+]
+
+[[package]]
+name = "itoa"
+version = "1.0.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8f42a60cbdf9a97f5d2305f08a87dc4e09308d1276d28c869c684d7777685682"
+
+[[package]]
+name = "js-sys"
+version = "0.3.100"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f2025f20d7a4fa7785846e7b63d10a76d3f1cee98ee5cb79ea59703f95e42162"
+dependencies = [
+ "cfg-if",
+ "futures-util",
+ "wasm-bindgen",
+]
+
+[[package]]
+name = "libc"
+version = "0.2.186"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "68ab91017fe16c622486840e4c83c9a37afeff978bd239b5293d61ece587de66"
+
+[[package]]
+name = "libsqlite3-sys"
+version = "0.30.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2e99fb7a497b1e3339bc746195567ed8d3e24945ecd636e3619d20b9de9e9149"
+dependencies = [
+ "cc",
+ "pkg-config",
+ "vcpkg",
+]
+
+[[package]]
+name = "linux-raw-sys"
+version = "0.12.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "32a66949e030da00e8c7d4434b251670a91556f4144941d37452769c25d58a53"
+
+[[package]]
+name = "memchr"
+version = "2.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6b947ae49db0d222b1dbc6b113ce7248a3fc3a6ca21b696717bfc000ba4484d8"
+
+[[package]]
+name = "num-traits"
+version = "0.2.19"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "071dfc062690e90b734c0b2273ce72ad0ffa95f0c74596bc250dcfd960262841"
+dependencies = [
+ "autocfg",
+]
+
+[[package]]
+name = "once_cell"
+version = "1.21.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9f7c3e4beb33f85d45ae3e3a1792185706c8e16d043238c593331cc7cd313b50"
+
+[[package]]
+name = "once_cell_polyfill"
+version = "1.70.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "384b8ab6d37215f3c5301a95a4accb5d64aa607f1fcb26a11b5303878451b4fe"
+
+[[package]]
+name = "oorandom"
+version = "11.1.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d6790f58c7ff633d8771f42965289203411a5e5c68388703c06e14f24770b41e"
+
+[[package]]
+name = "pin-project-lite"
+version = "0.2.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a89322df9ebe1c1578d689c92318e070967d1042b512afbe49518723f4e6d5cd"
+
+[[package]]
+name = "pkg-config"
+version = "0.3.33"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "19f132c84eca552bf34cab8ec81f1c1dcc229b811638f9d283dceabe58c5569e"
+
+[[package]]
+name = "plotters"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5aeb6f403d7a4911efb1e33402027fc44f29b5bf6def3effcc22d7bb75f2b747"
+dependencies = [
+ "num-traits",
+ "plotters-backend",
+ "plotters-svg",
+ "wasm-bindgen",
+ "web-sys",
+]
+
+[[package]]
+name = "plotters-backend"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "df42e13c12958a16b3f7f4386b9ab1f3e7933914ecea48da7139435263a4172a"
+
+[[package]]
+name = "plotters-svg"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "51bae2ac328883f7acdfea3d66a7c35751187f870bc81f94563733a154d7a670"
+dependencies = [
+ "plotters-backend",
+]
+
+[[package]]
+name = "proc-macro2"
+version = "1.0.106"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8fd00f0bb2e90d81d1044c2b32617f68fcb9fa3bb7640c23e9c748e53fb30934"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "quote"
+version = "1.0.45"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "41f2619966050689382d2b44f664f4bc593e129785a36d6ee376ddf37259b924"
+dependencies = [
+ "proc-macro2",
+]
+
+[[package]]
+name = "rayon"
+version = "1.12.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fb39b166781f92d482534ef4b4b1b2568f42613b53e5b6c160e24cfbfa30926d"
+dependencies = [
+ "either",
+ "rayon-core",
+]
+
+[[package]]
+name = "rayon-core"
+version = "1.13.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "22e18b0f0062d30d4230b2e85ff77fdfe4326feb054b9783a3460d8435c8ab91"
+dependencies = [
+ "crossbeam-deque",
+ "crossbeam-utils",
+]
+
+[[package]]
+name = "regex"
+version = "1.12.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f1292b7759ae1cb9ec195452d1390a074f0cd8541ab7a5a8c31cd6db45d4a6ba"
+dependencies = [
+ "aho-corasick",
+ "memchr",
+ "regex-automata",
+ "regex-syntax",
+]
+
+[[package]]
+name = "regex-automata"
+version = "0.4.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6e1dd4122fc1595e8162618945476892eefca7b88c52820e74af6262213cae8f"
+dependencies = [
+ "aho-corasick",
+ "memchr",
+ "regex-syntax",
+]
+
+[[package]]
+name = "regex-syntax"
+version = "0.8.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d6f6ff9a378485b298a5286656da665ba74413d36db0979633275d2e708145d4"
+
+[[package]]
+name = "rusqlite"
+version = "0.32.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7753b721174eb8ff87a9a0e799e2d7bc3749323e773db92e0984debb00019d6e"
+dependencies = [
+ "bitflags",
+ "fallible-iterator",
+ "fallible-streaming-iterator",
+ "hashlink",
+ "libsqlite3-sys",
+ "smallvec",
+]
+
+[[package]]
+name = "rustix"
+version = "1.1.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b6fe4565b9518b83ef4f91bb47ce29620ca828bd32cb7e408f0062e9930ba190"
+dependencies = [
+ "bitflags",
+ "errno",
+ "libc",
+ "linux-raw-sys",
+ "windows-sys",
+]
+
+[[package]]
+name = "rustversion"
+version = "1.0.22"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b39cdef0fa800fc44525c84ccb54a029961a8215f9619753635a9c0d2538d46d"
+
+[[package]]
+name = "ryu"
+version = "1.0.23"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9774ba4a74de5f7b1c1451ed6cd5285a32eddb5cccb8cc655a4e50009e06477f"
+
+[[package]]
+name = "same-file"
+version = "1.0.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "93fc1dc3aaa9bfed95e02e6eadabb4baf7e3078b0bd1b4d7b6b0b68378900502"
+dependencies = [
+ "winapi-util",
+]
+
+[[package]]
+name = "serde"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9a8e94ea7f378bd32cbbd37198a4a91436180c5bb472411e48b5ec2e2124ae9e"
+dependencies = [
+ "serde_core",
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_core"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "41d385c7d4ca58e59fc732af25c3983b67ac852c1a25000afe1175de458b67ad"
+dependencies = [
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_derive"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d540f220d3187173da220f885ab66608367b6574e925011a9353e4badda91d79"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "serde_json"
+version = "1.0.150"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e8014e44b4736ed0538adeecded0fce2a272f22dc9578a7eb6b2d9993c74cfb9"
+dependencies = [
+ "indexmap",
+ "itoa",
+ "memchr",
+ "serde",
+ "serde_core",
+ "zmij",
+]
+
+[[package]]
+name = "serde_spanned"
+version = "0.6.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bf41e0cfaf7226dca15e8197172c295a782857fcb97fad1808a166870dee75a3"
+dependencies = [
+ "serde",
+]
+
+[[package]]
+name = "serde_yaml"
+version = "0.9.34+deprecated"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6a8b1a1a2ebf674015cc02edccce75287f1a0130d394307b36743c2f5d504b47"
+dependencies = [
+ "indexmap",
+ "itoa",
+ "ryu",
+ "serde",
+ "unsafe-libyaml",
+]
+
+[[package]]
+name = "shlex"
+version = "2.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f8fadd59c855ef2080decdef8ff161eb6661b86933c9d82e5ba29dc602a55aba"
+
+[[package]]
+name = "slab"
+version = "0.4.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0c790de23124f9ab44544d7ac05d60440adc586479ce501c1d6d7da3cd8c9cf5"
+
+[[package]]
+name = "smallvec"
+version = "1.15.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "67b1b7a3b5fe4f1376887184045fcf45c69e92af734b7aaddc05fb777b6fbd03"
+
+[[package]]
+name = "streaming-iterator"
+version = "0.1.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2b2231b7c3057d5e4ad0156fb3dc807d900806020c5ffa3ee6ff2c8c76fb8520"
+
+[[package]]
+name = "strsim"
+version = "0.11.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7da8b5736845d9f2fcb837ea5d9e2628564b3b043a70948a3f0b778838c5fb4f"
+
+[[package]]
+name = "syn"
+version = "2.0.117"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e665b8803e7b1d2a727f4023456bbbbe74da67099c585258af0ad9c5013b9b99"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "unicode-ident",
+]
+
+[[package]]
+name = "terminal_size"
+version = "0.4.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "230a1b821ccbd75b185820a1f1ff7b14d21da1e442e22c0863ea5f08771a8874"
+dependencies = [
+ "rustix",
+ "windows-sys",
+]
+
+[[package]]
+name = "thiserror"
+version = "2.0.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4288b5bcbc7920c07a1149a35cf9590a2aa808e0bc1eafaade0b80947865fbc4"
+dependencies = [
+ "thiserror-impl",
+]
+
+[[package]]
+name = "thiserror-impl"
+version = "2.0.18"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ebc4ee7f67670e9b64d05fa4253e753e016c6c95ff35b89b7941d6b856dec1d5"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "tinytemplate"
+version = "1.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "be4d6b5f19ff7664e8c98d03e2139cb510db9b0a60b55f8e8709b689d939b6bc"
+dependencies = [
+ "serde",
+ "serde_json",
+]
+
+[[package]]
+name = "toml"
+version = "0.8.23"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "dc1beb996b9d83529a9e75c17a1686767d148d70663143c7854d8b4a09ced362"
+dependencies = [
+ "indexmap",
+ "serde",
+ "serde_spanned",
+ "toml_datetime",
+ "toml_edit",
+]
+
+[[package]]
+name = "toml_datetime"
+version = "0.6.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "22cddaf88f4fbc13c51aebbf5f8eceb5c7c5a9da2ac40a13519eb5b0a0e8f11c"
+dependencies = [
+ "serde",
+]
+
+[[package]]
+name = "toml_edit"
+version = "0.22.27"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "41fe8c660ae4257887cf66394862d21dbca4a6ddd26f04a3560410406a2f819a"
+dependencies = [
+ "indexmap",
+ "serde",
+ "serde_spanned",
+ "toml_datetime",
+ "toml_write",
+ "winnow",
+]
+
+[[package]]
+name = "toml_write"
+version = "0.1.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5d99f8c9a7727884afe522e9bd5edbfc91a3312b36a77b5fb8926e4c31a41801"
+
+[[package]]
+name = "tree-sitter"
+version = "0.25.10"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "78f873475d258561b06f1c595d93308a7ed124d9977cb26b148c2084a4a3cc87"
+dependencies = [
+ "cc",
+ "regex",
+ "regex-syntax",
+ "serde_json",
+ "streaming-iterator",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree-sitter-c"
+version = "0.24.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a9b2eb57a55fed6b00812912e730b7a275cf4fe98bfd6a5d76263d4438371728"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree-sitter-cpp"
+version = "0.23.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "df2196ea9d47b4ab4a31b9297eaa5a5d19a0b121dceb9f118f6790ad0ab94743"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree-sitter-javascript"
+version = "0.25.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "68204f2abc0627a90bdf06e605f5c470aa26fdcb2081ea553a04bdad756693f5"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree-sitter-language"
+version = "0.1.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "009994f150cc0cd50ff54917d5bc8bffe8cad10ca10d81c34da2ec421ae61782"
+
+[[package]]
+name = "tree-sitter-python"
+version = "0.25.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6bf85fd39652e740bf60f46f4cda9492c3a9ad75880575bf14960f775cb74a1c"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree-sitter-rust"
+version = "0.24.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "439e577dbe07423ec2582ac62c7531120dbfccfa6e5f92406f93dd271a120e45"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree-sitter-typescript"
+version = "0.23.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6c5f76ed8d947a75cc446d5fccd8b602ebf0cde64ccf2ffa434d873d7a575eff"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
+[[package]]
+name = "tree_plus_cli"
+version = "2.0.0-alpha.1"
+dependencies = [
+ "clap",
+ "terminal_size",
+ "tree_plus_core",
+]
+
+[[package]]
+name = "tree_plus_core"
+version = "2.0.0-alpha.1"
+dependencies = [
+ "criterion",
+ "rayon",
+ "regex",
+ "rusqlite",
+ "serde",
+ "serde_json",
+ "serde_yaml",
+ "thiserror",
+ "toml",
+ "tree-sitter",
+ "tree-sitter-c",
+ "tree-sitter-cpp",
+ "tree-sitter-javascript",
+ "tree-sitter-python",
+ "tree-sitter-rust",
+ "tree-sitter-typescript",
+ "unicode-width",
+]
+
+[[package]]
+name = "unicode-ident"
+version = "1.0.24"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
+
+[[package]]
+name = "unicode-width"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b4ac048d71ede7ee76d585517add45da530660ef4390e49b098733c6e897f254"
+
+[[package]]
+name = "unsafe-libyaml"
+version = "0.2.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "673aac59facbab8a9007c7f6108d11f63b603f7cabff99fabf650fea5c32b861"
+
+[[package]]
+name = "utf8parse"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "06abde3611657adf66d383f00b093d7faecc7fa57071cce2578660c9f1010821"
+
+[[package]]
+name = "vcpkg"
+version = "0.2.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "accd4ea62f7bb7a82fe23066fb0957d48ef677f6eeb8215f372f52e48bb32426"
+
+[[package]]
+name = "version_check"
+version = "0.9.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0b928f33d975fc6ad9f86c8f283853ad26bdd5b10b7f1542aa2fa15e2289105a"
+
+[[package]]
+name = "walkdir"
+version = "2.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "29790946404f91d9c5d06f9874efddea1dc06c5efe94541a7d6863108e3a5e4b"
+dependencies = [
+ "same-file",
+ "winapi-util",
+]
+
+[[package]]
+name = "wasm-bindgen"
+version = "0.2.123"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a254a4b10c19a76f09a27640e7ffbf9bc30bf67e16a3bf28aaefa4920fe81563"
+dependencies = [
+ "cfg-if",
+ "once_cell",
+ "rustversion",
+ "wasm-bindgen-macro",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-macro"
+version = "0.2.123"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "24a40fc75b0ec6f3746ceb10d36f53a93dcd68a93b11b6445983945d79eba0dc"
+dependencies = [
+ "quote",
+ "wasm-bindgen-macro-support",
+]
+
+[[package]]
+name = "wasm-bindgen-macro-support"
+version = "0.2.123"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "908f34bd9b9ce3d4caf07b72dfab63d61504d156856c6bd3cd87fa350cf3985b"
+dependencies = [
+ "bumpalo",
+ "proc-macro2",
+ "quote",
+ "syn",
+ "wasm-bindgen-shared",
+]
+
+[[package]]
+name = "wasm-bindgen-shared"
+version = "0.2.123"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7acbf7616c27b194bbb550bf77ed0c2c3e5b7fd1260a93082b95fb7f47959b92"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "web-sys"
+version = "0.3.100"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6e0871acf327f283dc6da28a1696cdc64fb355ba9f935d052021fa77f35cce69"
+dependencies = [
+ "js-sys",
+ "wasm-bindgen",
+]
+
+[[package]]
+name = "winapi-util"
+version = "0.1.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c2a7b1c03c876122aa43f3020e6c3c3ee5c05081c9a00739faf7503aeba10d22"
+dependencies = [
+ "windows-sys",
+]
+
+[[package]]
+name = "windows-link"
+version = "0.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f0805222e57f7521d6a62e36fa9163bc891acd422f971defe97d64e70d0a4fe5"
+
+[[package]]
+name = "windows-sys"
+version = "0.61.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ae137229bcbd6cdf0f7b80a31df61766145077ddf49416a728b02cb3921ff3fc"
+dependencies = [
+ "windows-link",
+]
+
+[[package]]
+name = "winnow"
+version = "0.7.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "df79d97927682d2fd8adb29682d1140b343be4ac0f08fd68b7765d9c059d3945"
+dependencies = [
+ "memchr",
+]
+
+[[package]]
+name = "zerocopy"
+version = "0.8.50"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3b065d4f0e55f82fae73202e189638116a87c55ab6b8e6c2721e13dd9d854ad1"
+dependencies = [
+ "zerocopy-derive",
+]
+
+[[package]]
+name = "zerocopy-derive"
+version = "0.8.50"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0b631b19d36a892ab55420c92dbc83ccd79274f25be714855d3074aa71cab639"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "zmij"
+version = "1.0.21"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b8848ee67ecc8aedbaf3e4122217aff892639231befc6a1b58d29fff4c2cabaa"
diff --git a/Cargo.toml b/Cargo.toml
new file mode 100644
index 0000000..afff7ec
--- /dev/null
+++ b/Cargo.toml
@@ -0,0 +1,29 @@
+[workspace]
+resolver = "2"
+members = ["crates/tree_plus_core", "crates/tree_plus_cli"]
+
+[workspace.package]
+version = "2.0.0-alpha.1"
+edition = "2021"
+license = "MIT OR Apache-2.0"
+repository = "https://github.com/bionicles/tree_plus"
+
+[workspace.dependencies]
+regex = "1"
+serde = { version = "1", features = ["derive"] }
+serde_json = { version = "1", features = ["preserve_order"] }
+toml = { version = "0.8", features = ["preserve_order"] }
+serde_yaml = "0.9"
+unicode-width = "0.2"
+rayon = "1.10"
+thiserror = "2"
+clap = { version = "4.5", features = ["derive"] }
+tree-sitter = "0.25"
+tree-sitter-rust = "0.24"
+tree-sitter-python = "0.25"
+tree-sitter-javascript = "0.25"
+tree-sitter-typescript = "0.23"
+tree-sitter-c = "0.24"
+tree-sitter-cpp = "0.23"
+rusqlite = { version = "0.32", features = ["bundled"] }
+criterion = "0.5"
diff --git a/crates/tree_plus_cli/Cargo.toml b/crates/tree_plus_cli/Cargo.toml
new file mode 100644
index 0000000..44b5c50
--- /dev/null
+++ b/crates/tree_plus_cli/Cargo.toml
@@ -0,0 +1,21 @@
+[package]
+name = "tree_plus_cli"
+description = "A `tree` util enhanced with tokens, lines, and components."
+version.workspace = true
+edition.workspace = true
+license.workspace = true
+repository.workspace = true
+
+[[bin]]
+name = "tree_plus"
+path = "src/main.rs"
+
+# alias to avoid PATH collision with the Python tree_plus entry point
+[[bin]]
+name = "tprs"
+path = "src/main.rs"
+
+[dependencies]
+tree_plus_core = { path = "../tree_plus_core" }
+clap.workspace = true
+terminal_size = "0.4"
diff --git a/crates/tree_plus_cli/src/main.rs b/crates/tree_plus_cli/src/main.rs
new file mode 100644
index 0000000..f0f5af5
--- /dev/null
+++ b/crates/tree_plus_cli/src/main.rs
@@ -0,0 +1,173 @@
+//! tree_plus CLI: a `tree` util enhanced with tokens, lines, and components.
+//!
+//! Flag-compatible with the legacy Python click CLI for the version-1 scope
+//! (local filesystem mode). Web/HN modes are deferred; see
+//! docs/rust-port-differences.md.
+
+use std::time::Instant;
+
+use clap::Parser;
+
+use tree_plus_core::count::TokenizerName;
+use tree_plus_core::{from_seeds, render_to_string, TreePlusConfig};
+
+const VERSION: &str = env!("CARGO_PKG_VERSION");
+
+#[derive(Parser, Debug)]
+#[command(
+    name = "tree_plus",
+    disable_version_flag = true,
+    about = "A `tree` util enhanced with tokens, lines, and components.",
+    after_help = format!("v({VERSION}) --- https://github.com/bionicles/tree_plus/blob/main/README.md"),
+)]
+struct Cli {
+    /// Patterns to ignore, in quotes: -i "*.java"
+    #[arg(short = 'i', long = "ignore")]
+    ignore: Vec<String>,
+
+    /// Override DEFAULT_IGNORE (includes ignored content): -o -i "*.java"
+    #[arg(short = 'o', long = "override")]
+    override_ignore: bool,
+
+    /// Patterns to find, in quotes: -g "*.rs"
+    #[arg(short = 'g', long = "glob")]
+    glob: Vec<String>,
+
+    /// Print the version and exit.
+    #[arg(short = 'v', long = "version")]
+    version: bool,
+
+    /// Enables debug output.
+    #[arg(short = 'd', long = "debug")]
+    debug: bool,
+
+    /// DISABLE Syntax Highlighting. (The Rust port renders plain text; this
+    /// flag is accepted for compatibility and controls legacy markup escaping.)
+    #[arg(short = 's', long = "syntax")]
+    syntax: bool,
+
+    /// Omit module components. (False)
+    #[arg(short = 'c', long = "concise")]
+    concise: bool,
+
+    /// A shorthand for tiktoken with the 'gpt-4o' tokenizer (unsupported in
+    /// the Rust port: errors explicitly).
+    #[arg(short = 't', long = "tiktoken")]
+    tiktoken: bool,
+
+    /// Name of the tokenizer to use; only 'wc' is supported in the Rust port.
+    #[arg(short = 'T', long = "tokenizer-name")]
+    tokenizer_name: Option<String>,
+
+    /// Regex timeout in seconds (accepted for compatibility; the Rust port
+    /// does not need regex timeouts).
+    #[arg(long = "timeout")]
+    timeout: Option<f64>,
+
+    /// Paths or globs to map.
+    paths: Vec<String>,
+}
+
+fn main() {
+    let start = Instant::now();
+    let cli = Cli::parse();
+
+    if cli.version {
+        println!("{VERSION}");
+        return;
+    }
+
+    let tokenizer = match (cli.tiktoken, cli.tokenizer_name.as_deref()) {
+        (false, None) | (true, Some("wc")) | (false, Some("wc")) => TokenizerName::Wc,
+        (true, None) | (_, Some("gpt4o")) | (_, Some("gpt-4o")) => {
+            eprintln!(
+                "error: the gpt-4o tokenizer is not supported by the Rust port; \
+                 use the default word-count tokenizer (wc)"
+            );
+            std::process::exit(2);
+        }
+        (_, Some(other)) => {
+            eprintln!("error: unsupported tokenizer {other:?} (only 'wc' is supported)");
+            std::process::exit(2);
+        }
+    };
+
+    let config = TreePlusConfig {
+        ignore: cli.ignore.clone(),
+        override_ignore: cli.override_ignore,
+        globs: cli.glob.clone(),
+        concise: cli.concise,
+        tokenizer,
+        syntax: false,
+        max_tokens: tree_plus_core::config::MAX_TOKENS,
+    };
+    if cli.debug {
+        eprintln!(
+            "tree_plus main received paths={:?} ignore={:?} glob={:?}",
+            cli.paths, cli.ignore, cli.glob
+        );
+    }
+    let _ = cli.timeout; // compatibility no-op
+    let _ = cli.syntax; // plain-text output; highlighting deferred
+
+    let root = from_seeds(&cli.paths, &config);
+
+    let width = terminal_width();
+    print!("{}", render_to_string(&root, width));
+
+    let og_ignore = format_tuple(&cli.ignore);
+    let globs = format_tuple(&cli.glob);
+    let paths = format_tuple(&cli.paths);
+    let mut footer = format!("\ntree_plus v({VERSION}) ignore={og_ignore} globs={globs}");
+    if cli.concise {
+        footer.push_str(&format!(" concise=True paths={paths}"));
+    } else {
+        footer.push_str(&format!(" syntax={} paths={paths}", py_bool(!cli.syntax)));
+    }
+    footer.push_str(&format!(
+        "\n{} in {:.2} second(s).",
+        root.stats(),
+        start.elapsed().as_secs_f64()
+    ));
+    println!("{footer}");
+}
+
+/// Python tuple repr for the footer line, e.g. `('a',)` / `()` / `None`.
+fn format_tuple(items: &[String]) -> String {
+    if items.is_empty() {
+        return "()".to_string();
+    }
+    let inner: Vec<String> = items.iter().map(|s| format!("'{s}'")).collect();
+    if items.len() == 1 {
+        format!("({},)", inner[0])
+    } else {
+        format!("({})", inner.join(", "))
+    }
+}
+
+fn py_bool(b: bool) -> &'static str {
+    if b {
+        "True"
+    } else {
+        "False"
+    }
+}
+
+fn terminal_width() -> usize {
+    use std::io::IsTerminal;
+    // honor the legacy README-update width override
+    if std::env::var("TREE_PLUS_UPDATE_README").as_deref() == Ok("YES") {
+        return 128;
+    }
+    // piped output renders at the legacy capture width
+    if !std::io::stdout().is_terminal() {
+        return tree_plus_core::DEFAULT_WIDTH;
+    }
+    // real terminal width, like the legacy rich Console
+    if let Some((terminal_size::Width(w), _)) = terminal_size::terminal_size() {
+        if w > 0 {
+            return w as usize;
+        }
+    }
+    tree_plus_core::DEFAULT_WIDTH
+}
diff --git a/crates/tree_plus_cli/tests/cli.rs b/crates/tree_plus_cli/tests/cli.rs
new file mode 100644
index 0000000..50ecc6e
--- /dev/null
+++ b/crates/tree_plus_cli/tests/cli.rs
@@ -0,0 +1,92 @@
+//! CLI integration tests mirroring the README examples (local fs mode).
+
+use std::path::Path;
+use std::process::Command;
+
+fn repo_root() -> std::path::PathBuf {
+    Path::new(env!("CARGO_MANIFEST_DIR"))
+        .join("../..")
+        .canonicalize()
+        .unwrap()
+}
+
+fn tree_plus(args: &[&str]) -> std::process::Output {
+    Command::new(env!("CARGO_BIN_EXE_tree_plus"))
+        .args(args)
+        .current_dir(repo_root())
+        .output()
+        .expect("run tree_plus")
+}
+
+#[test]
+fn version_flag() {
+    let out = tree_plus(&["-v"]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(stdout.trim().starts_with("2."), "version: {stdout}");
+}
+
+#[test]
+fn renders_a_directory_with_stats_footer() {
+    let out = tree_plus(&["tests/path_to_test"]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(stdout.contains("📁 path_to_test (1 folder, 6 files)"));
+    assert!(stdout.contains("└── def hello_world()"));
+    assert!(stdout.contains("folder(s), 6 file(s),"));
+    assert!(stdout.contains("second(s)."));
+}
+
+#[test]
+fn ignore_flag_excludes_extension() {
+    let out = tree_plus(&["-i", "*.py", "tests/path_to_test"]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(!stdout.contains(".py ("));
+    assert!(stdout.contains("file.md"));
+}
+
+#[test]
+fn glob_flag_filters() {
+    let out = tree_plus(&["-g", "*.rs", "tests/more_languages"]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(stdout.contains(".rs ("));
+    assert!(!stdout.contains(".java"));
+}
+
+#[test]
+fn concise_mode_omits_components() {
+    let out = tree_plus(&["-c", "tests/path_to_test"]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(stdout.contains("📁 path_to_test"));
+    assert!(!stdout.contains("def hello_world()"));
+    assert!(stdout.contains("concise=True"));
+}
+
+#[test]
+fn glob_seed_pattern() {
+    let out = tree_plus(&["tree_plus_src/*.py"]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(stdout.contains("🌀 tree_plus_src/*.py ("));
+    assert!(stdout.contains("engine.py"));
+}
+
+#[test]
+fn unsupported_tokenizer_errors_explicitly() {
+    let out = tree_plus(&["-t", "tests/path_to_test"]);
+    assert!(!out.status.success());
+    let stderr = String::from_utf8_lossy(&out.stderr);
+    assert!(stderr.contains("not supported"));
+}
+
+#[test]
+fn renders_this_repository() {
+    let out = tree_plus(&["-c", "."]);
+    assert!(out.status.success());
+    let stdout = String::from_utf8_lossy(&out.stdout);
+    assert!(stdout.contains("📁 tree_plus ("));
+    assert!(stdout.contains("README.md"));
+}
diff --git a/crates/tree_plus_core/Cargo.toml b/crates/tree_plus_core/Cargo.toml
new file mode 100644
index 0000000..89fead3
--- /dev/null
+++ b/crates/tree_plus_core/Cargo.toml
@@ -0,0 +1,36 @@
+[package]
+name = "tree_plus_core"
+description = "Core library for tree_plus: map repositories into trees with components, tokens, and lines"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
+repository.workspace = true
+
+[features]
+default = ["sqlite"]
+sqlite = ["dep:rusqlite"]
+
+[dependencies]
+regex.workspace = true
+serde.workspace = true
+serde_json.workspace = true
+toml.workspace = true
+serde_yaml.workspace = true
+unicode-width.workspace = true
+rayon.workspace = true
+thiserror.workspace = true
+tree-sitter.workspace = true
+tree-sitter-rust.workspace = true
+tree-sitter-python.workspace = true
+tree-sitter-javascript.workspace = true
+tree-sitter-typescript.workspace = true
+tree-sitter-c.workspace = true
+tree-sitter-cpp.workspace = true
+rusqlite = { workspace = true, optional = true }
+
+[dev-dependencies]
+criterion.workspace = true
+
+[[bench]]
+name = "tree_plus_bench"
+harness = false
diff --git a/crates/tree_plus_core/benches/tree_plus_bench.rs b/crates/tree_plus_core/benches/tree_plus_bench.rs
new file mode 100644
index 0000000..1600a49
--- /dev/null
+++ b/crates/tree_plus_core/benches/tree_plus_bench.rs
@@ -0,0 +1,78 @@
+//! Criterion benchmarks: traversal+counting, extraction, and full renders.
+//!
+//! Run: cargo bench -p tree_plus_core
+//! Benchmark any repository: cargo bench -p tree_plus_core -- --quick
+//! (set TREE_PLUS_BENCH_PATH=/path/to/repo to point at a different tree).
+
+use std::path::{Path, PathBuf};
+
+use criterion::{criterion_group, criterion_main, Criterion};
+use std::hint::black_box;
+
+use tree_plus_core::{extract_components, from_seeds, render_to_string, TreePlusConfig};
+
+fn repo_root() -> PathBuf {
+    Path::new(env!("CARGO_MANIFEST_DIR"))
+        .join("../..")
+        .canonicalize()
+        .unwrap()
+}
+
+fn bench_target() -> PathBuf {
+    std::env::var("TREE_PLUS_BENCH_PATH")
+        .map(PathBuf::from)
+        .unwrap_or_else(|_| repo_root())
+}
+
+fn full_tree(c: &mut Criterion) {
+    let target = bench_target();
+    let seeds = vec![target.to_string_lossy().into_owned()];
+    let config = TreePlusConfig::default();
+    c.bench_function("full_tree_with_components", |b| {
+        b.iter(|| {
+            let tree = from_seeds(black_box(&seeds), &config);
+            black_box(render_to_string(&tree, 80))
+        })
+    });
+}
+
+fn concise_tree(c: &mut Criterion) {
+    let target = bench_target();
+    let seeds = vec![target.to_string_lossy().into_owned()];
+    let config = TreePlusConfig {
+        concise: true,
+        ..Default::default()
+    };
+    c.bench_function("concise_tree_traversal_counting", |b| {
+        b.iter(|| {
+            let tree = from_seeds(black_box(&seeds), &config);
+            black_box(render_to_string(&tree, 80))
+        })
+    });
+}
+
+fn single_file_extraction(c: &mut Criterion) {
+    let root = repo_root();
+    let cases = [
+        ("rust", "tests/more_languages/group4/rust_test.rs"),
+        ("python", "tests/path_to_test/class_method_type.py"),
+        ("typescript", "tests/more_languages/group1/test.ts"),
+        (
+            "c_pathological",
+            "tests/more_languages/group6/catastrophic.c",
+        ),
+        ("markdown", "tests/more_languages/group_todo/todo.md"),
+    ];
+    for (name, rel) in cases {
+        let path = root.join(rel);
+        if !path.exists() {
+            continue;
+        }
+        c.bench_function(&format!("extract_{name}"), |b| {
+            b.iter(|| black_box(extract_components(black_box(&path), false)))
+        });
+    }
+}
+
+criterion_group!(benches, full_tree, concise_tree, single_file_extraction);
+criterion_main!(benches);
diff --git a/crates/tree_plus_core/examples/dump_ast.rs b/crates/tree_plus_core/examples/dump_ast.rs
new file mode 100644
index 0000000..436cea7
--- /dev/null
+++ b/crates/tree_plus_core/examples/dump_ast.rs
@@ -0,0 +1,55 @@
+//! Debug helper: dump a tree-sitter AST outline.
+//! Usage: cargo run -p tree_plus_core --example dump_ast -- <file> [start_line] [end_line]
+
+fn main() {
+    let args: Vec<String> = std::env::args().skip(1).collect();
+    let path = &args[0];
+    let lo: usize = args.get(1).and_then(|s| s.parse().ok()).unwrap_or(0);
+    let hi: usize = args
+        .get(2)
+        .and_then(|s| s.parse().ok())
+        .unwrap_or(usize::MAX);
+    let content = std::fs::read_to_string(path).unwrap();
+    let ext = format!(
+        ".{}",
+        std::path::Path::new(path)
+            .extension()
+            .unwrap()
+            .to_string_lossy()
+    );
+    let language: tree_sitter::Language = match ext.as_str() {
+        ".cpp" | ".cc" | ".cu" | ".hpp" => tree_sitter_cpp::LANGUAGE.into(),
+        ".c" | ".h" => tree_sitter_c::LANGUAGE.into(),
+        ".py" => tree_sitter_python::LANGUAGE.into(),
+        ".rs" => tree_sitter_rust::LANGUAGE.into(),
+        ".js" | ".ts" => tree_sitter_typescript::LANGUAGE_TYPESCRIPT.into(),
+        other => panic!("no grammar for {other}"),
+    };
+    let mut parser = tree_sitter::Parser::new();
+    parser.set_language(&language).unwrap();
+    let tree = parser.parse(&content, None).unwrap();
+    let mut stack = vec![(tree.root_node(), 0usize)];
+    while let Some((node, depth)) = stack.pop() {
+        let row = node.start_position().row + 1;
+        if row >= lo && row <= hi {
+            let text: String = content[node.byte_range()]
+                .chars()
+                .take(60)
+                .collect::<String>()
+                .replace('\n', "\\n");
+            println!(
+                "{}{} [{}..{}] {:?}",
+                "  ".repeat(depth),
+                node.kind(),
+                row,
+                node.end_position().row + 1,
+                text
+            );
+        }
+        let mut cursor = node.walk();
+        let children: Vec<_> = node.named_children(&mut cursor).collect();
+        for child in children.into_iter().rev() {
+            stack.push((child, depth + 1));
+        }
+    }
+}
diff --git a/crates/tree_plus_core/examples/extract.rs b/crates/tree_plus_core/examples/extract.rs
new file mode 100644
index 0000000..2152e15
--- /dev/null
+++ b/crates/tree_plus_core/examples/extract.rs
@@ -0,0 +1,16 @@
+//! Debug helper: print extracted components as JSON for given paths.
+//! Usage: cargo run -p tree_plus_core --example extract -- <path>...
+
+fn main() {
+    for arg in std::env::args().skip(1) {
+        let components = tree_plus_core::extract_components(std::path::Path::new(&arg), false);
+        println!(
+            "{}",
+            serde_json::to_string_pretty(&serde_json::json!({
+                "path": arg,
+                "components": components,
+            }))
+            .unwrap()
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/config.rs b/crates/tree_plus_core/src/config.rs
new file mode 100644
index 0000000..346d50c
--- /dev/null
+++ b/crates/tree_plus_core/src/config.rs
@@ -0,0 +1,39 @@
+//! Configuration for tree building (CLI flags map onto this).
+
+use crate::count::TokenizerName;
+
+/// Maximum token count before a file's components are skipped
+/// (legacy `MAX_TOKENS`).
+pub const MAX_TOKENS: u64 = 1_000_000;
+
+#[derive(Debug, Clone)]
+pub struct TreePlusConfig {
+    /// User ignore patterns (unioned with defaults unless `override_ignore`).
+    pub ignore: Vec<String>,
+    /// Use only user ignore patterns.
+    pub override_ignore: bool,
+    /// Glob patterns to keep (e.g. `*.rs`).
+    pub globs: Vec<String>,
+    /// Skip component extraction entirely.
+    pub concise: bool,
+    /// Tokenizer (only `Wc` is supported in the Rust port).
+    pub tokenizer: TokenizerName,
+    /// Legacy syntax-highlighting flag; affects Rust enum markup escaping.
+    pub syntax: bool,
+    /// Maximum tokens before skipping component extraction.
+    pub max_tokens: u64,
+}
+
+impl Default for TreePlusConfig {
+    fn default() -> Self {
+        TreePlusConfig {
+            ignore: Vec::new(),
+            override_ignore: false,
+            globs: Vec::new(),
+            concise: false,
+            tokenizer: TokenizerName::Wc,
+            syntax: false,
+            max_tokens: MAX_TOKENS,
+        }
+    }
+}
diff --git a/crates/tree_plus_core/src/count.rs b/crates/tree_plus_core/src/count.rs
new file mode 100644
index 0000000..dcc21be
--- /dev/null
+++ b/crates/tree_plus_core/src/count.rs
@@ -0,0 +1,203 @@
+//! Token and line counting, matching the legacy `wc -ml`-based counter.
+//!
+//! Legacy default tokenizer ("wc"): `n_lines` = newline count, `n_tokens` =
+//! character count / 4 (1 token ~= 4 chars). Characters are Unicode scalar
+//! values for valid UTF-8; invalid bytes each count as one character, which
+//! matches `wc -m` behavior in a UTF-8 locale closely enough for parity.
+
+use std::collections::HashSet;
+use std::path::Path;
+use std::sync::LazyLock;
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Default)]
+pub struct TokenLineCount {
+    pub n_tokens: u64,
+    pub n_lines: u64,
+}
+
+/// Tokenizer selection mirroring the legacy `TokenizerName` enum.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Default)]
+pub enum TokenizerName {
+    #[default]
+    Wc,
+    /// Recognized but unsupported in the Rust port; selecting it is an
+    /// explicit error per the port requirements.
+    Gpt4o,
+}
+
+/// Extensions whose contents are never counted (legacy `extensions_not_to_count`).
+static EXTENSIONS_NOT_TO_COUNT: LazyLock<HashSet<&'static str>> = LazyLock::new(|| {
+    [
+        ".stl",
+        ".obj",
+        ".gcode",
+        ".3mf",
+        ".amf",
+        ".ply",
+        ".f3d",
+        ".iges",
+        ".igs",
+        "step",
+        ".stp",
+        ".vrml",
+        ".wrl",
+        ".7z",
+        ".aac",
+        ".ai",
+        ".avi",
+        ".bak",
+        ".bin",
+        ".bz2",
+        ".chk",
+        ".class",
+        ".csv",
+        ".d",
+        ".dat",
+        ".db",
+        ".dll",
+        ".doc",
+        ".docx",
+        ".dylib",
+        ".ear",
+        ".eps",
+        ".exe",
+        ".flac",
+        ".flv",
+        ".framework",
+        ".gdoc",
+        ".gif",
+        ".gsheet",
+        ".gz",
+        ".img",
+        ".ipa",
+        ".iso",
+        ".jar",
+        ".jpg",
+        ".jpeg",
+        ".lock",
+        ".log",
+        ".mov",
+        ".mp3",
+        ".mp4",
+        ".nib",
+        ".node",
+        ".o",
+        ".odg",
+        ".pack",
+        ".pdf",
+        ".png",
+        ".ppt",
+        ".pptx",
+        ".psd",
+        ".pyc",
+        ".pyo",
+        ".pyd",
+        ".rar",
+        ".rlib",
+        ".rmeta",
+        ".so",
+        ".sqlite",
+        ".storyboardc",
+        ".swp",
+        ".tar",
+        ".tml",
+        ".wav",
+        ".war",
+        ".wmv",
+        ".xcarchive",
+        ".xlsx",
+        ".zip",
+        ".zst",
+    ]
+    .into_iter()
+    .collect()
+});
+
+/// Lowercased extension including the dot, like `os.path.splitext` (lowered).
+pub fn dot_extension(path: &Path) -> String {
+    let name = path.file_name().and_then(|n| n.to_str()).unwrap_or("");
+    // os.path.splitext: leading dots do not start an extension
+    let trimmed = name.trim_start_matches('.');
+    let n_leading = name.len() - trimmed.len();
+    match trimmed.rfind('.') {
+        Some(i) => name[n_leading + i..].to_lowercase(),
+        None => String::new(),
+    }
+}
+
+/// Count tokens and lines for raw file bytes.
+pub fn count_tokens_lines_bytes(bytes: &[u8]) -> TokenLineCount {
+    let n_lines = bytes.iter().filter(|&&b| b == b'\n').count() as u64;
+    // count UTF-8 scalar values; invalid bytes count 1 each (lossy)
+    let mut n_chars: u64 = 0;
+    let mut rest = bytes;
+    while !rest.is_empty() {
+        match std::str::from_utf8(rest) {
+            Ok(s) => {
+                n_chars += s.chars().count() as u64;
+                break;
+            }
+            Err(e) => {
+                let valid = e.valid_up_to();
+                n_chars += std::str::from_utf8(&rest[..valid])
+                    .map(|s| s.chars().count() as u64)
+                    .unwrap_or(0);
+                let skip = e.error_len().unwrap_or(rest.len() - valid).max(1);
+                n_chars += 1; // the invalid sequence counts as one char
+                rest = &rest[valid + skip..];
+            }
+        }
+    }
+    TokenLineCount {
+        n_tokens: n_chars / 4,
+        n_lines,
+    }
+}
+
+/// Whether counting is skipped for this path (legacy `extensions_not_to_count`).
+pub fn skip_counting(path: &Path) -> bool {
+    EXTENSIONS_NOT_TO_COUNT.contains(dot_extension(path).as_str())
+}
+
+/// Count tokens and lines in a file; `None` when counting is skipped.
+pub fn count_tokens_lines(path: &Path) -> Option<TokenLineCount> {
+    if path.is_dir() || skip_counting(path) {
+        return None;
+    }
+    let bytes = std::fs::read(path).ok()?;
+    Some(count_tokens_lines_bytes(&bytes))
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn wc_style_counts() {
+        let c = count_tokens_lines_bytes(b"def hello_world():\n    print(\"hello world\")\n");
+        // 44 chars // 4 = 11 tokens, 2 newlines
+        assert_eq!(c.n_lines, 2);
+        assert_eq!(c.n_tokens, 11);
+    }
+
+    #[test]
+    fn empty_is_zero() {
+        let c = count_tokens_lines_bytes(b"");
+        assert_eq!(c, TokenLineCount::default());
+    }
+
+    #[test]
+    fn extension_skips() {
+        assert!(skip_counting(Path::new("x/a.csv")));
+        assert!(skip_counting(Path::new("a.SQLITE")));
+        assert!(!skip_counting(Path::new("a.py")));
+    }
+
+    #[test]
+    fn dot_extension_rules() {
+        assert_eq!(dot_extension(Path::new("a.PY")), ".py");
+        assert_eq!(dot_extension(Path::new(".env")), "");
+        assert_eq!(dot_extension(Path::new(".env.test")), ".test");
+        assert_eq!(dot_extension(Path::new("Makefile")), "");
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/data.rs b/crates/tree_plus_core/src/extract/data.rs
new file mode 100644
index 0000000..b9fdeb2
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/data.rs
@@ -0,0 +1,582 @@
+//! Structured data extractors: JSON family, JSONL, YAML, TOML, CSV, SQLite.
+//! Output formatting matches the legacy Python f-strings (including Python
+//! `str()` / `repr()` rendering of scalars where the legacy code relied on it).
+
+use serde_json::Value as Json;
+
+use super::{simple::splitlines, ExtractError, ExtractResult};
+
+fn parse_err<E: std::fmt::Display>(e: E) -> ExtractError {
+    ExtractError::Parse(e.to_string())
+}
+
+/// Python `str()` of a JSON scalar/value (used inside legacy f-strings).
+fn py_str(v: &Json) -> String {
+    match v {
+        Json::Null => "None".to_string(),
+        Json::Bool(b) => if *b { "True" } else { "False" }.to_string(),
+        Json::Number(n) => n.to_string(),
+        Json::String(s) => s.clone(),
+        other => py_repr(other),
+    }
+}
+
+/// Python `repr()` of a JSON value (lists/dicts inside legacy f-strings).
+fn py_repr(v: &Json) -> String {
+    match v {
+        Json::Null => "None".to_string(),
+        Json::Bool(b) => if *b { "True" } else { "False" }.to_string(),
+        Json::Number(n) => n.to_string(),
+        Json::String(s) => format!("'{}'", s.replace('\\', "\\\\").replace('\'', "\\'")),
+        Json::Array(items) => {
+            let inner: Vec<String> = items.iter().map(py_repr).collect();
+            format!("[{}]", inner.join(", "))
+        }
+        Json::Object(map) => {
+            let inner: Vec<String> = map
+                .iter()
+                .map(|(k, v)| format!("'{}': {}", k, py_repr(v)))
+                .collect();
+            format!("{{{}}}", inner.join(", "))
+        }
+    }
+}
+
+/// Legacy `parse_jsonl`: first line's keys with Python type names.
+pub fn extract_jsonl(content: &str) -> ExtractResult {
+    let first_line = content.split('\n').next().unwrap_or("");
+    let data: Json = serde_json::from_str(first_line).map_err(parse_err)?;
+    let Json::Object(map) = data else {
+        return Err(ExtractError::Parse("jsonl root is not an object".into()));
+    };
+    Ok(map
+        .iter()
+        .map(|(key, value)| match value {
+            Json::Array(_) => format!("{key}: list"),
+            Json::Object(_) => format!("{key}: dict"),
+            Json::Null => format!("{key}: None"),
+            Json::Bool(_) => format!("{key}: bool"),
+            Json::String(_) => format!("{key}: str"),
+            Json::Number(n) => {
+                if n.is_f64() {
+                    format!("{key}: float")
+                } else {
+                    format!("{key}: int")
+                }
+            }
+        })
+        .collect())
+}
+
+/// Legacy `parse_json_schema`: $schema/type/title/description keys.
+pub fn extract_json_schema(content: &str) -> ExtractResult {
+    let parsed: Json = serde_json::from_str(content).map_err(parse_err)?;
+    let mut keepers = Vec::new();
+    for key in ["$schema", "type", "title", "description"] {
+        if let Some(v) = parsed.get(key) {
+            keepers.push(format!("{key}: {}", py_str(v)));
+        }
+    }
+    Ok(keepers)
+}
+
+/// Legacy `parse_package_json`: name, version, scripts.
+pub fn extract_package_json(content: &str) -> ExtractResult {
+    let parsed: Json = serde_json::from_str(content).map_err(parse_err)?;
+    let mut keepers = Vec::new();
+    keepers.push(match parsed.get("name") {
+        Some(name) => format!("name: '{}'", py_str(name)),
+        None => "name: ?".to_string(),
+    });
+    keepers.push(match parsed.get("version") {
+        Some(version) => format!("version: {}", py_str(version)),
+        None => "version: ?".to_string(),
+    });
+    if let Some(Json::Object(scripts)) = parsed.get("scripts") {
+        if !scripts.is_empty() {
+            keepers.push("scripts:".to_string());
+            for (key, value) in scripts {
+                keepers.push(format!("    {key}: '{}'", py_str(value)));
+            }
+        }
+    }
+    Ok(keepers)
+}
+
+/// Legacy `parse_json_rpc`.
+pub fn extract_json_rpc(content: &str) -> ExtractResult {
+    let data: Json = serde_json::from_str(content).map_err(parse_err)?;
+    let get = |k: &str| data.get(k).map(py_str).unwrap_or_else(|| "N/A".to_string());
+    let mut components = vec![
+        format!("jsonrpc: {}", get("jsonrpc")),
+        format!("method: {}", get("method")),
+        "params:".to_string(),
+    ];
+    if let Some(Json::Object(params)) = data.get("params") {
+        for (param, value) in params {
+            components.push(format!("    {param}: {}", py_str(value)));
+        }
+    }
+    components.push(format!("id: {}", get("id")));
+    Ok(components)
+}
+
+/// Legacy `parse_openrpc_json`.
+pub fn extract_openrpc_json(content: &str) -> ExtractResult {
+    let data: Json = serde_json::from_str(content).map_err(parse_err)?;
+    let str_or =
+        |v: Option<&Json>, default: &str| v.map(py_str).unwrap_or_else(|| default.to_string());
+    let mut components = vec![
+        format!("openrpc: {}", str_or(data.get("openrpc"), "N/A")),
+        "info:".to_string(),
+    ];
+    let info = data.get("info");
+    components.push(format!(
+        "    title: {}",
+        str_or(info.and_then(|i| i.get("title")), "N/A")
+    ));
+    components.push(format!(
+        "    version: {}",
+        str_or(info.and_then(|i| i.get("version")), "N/A")
+    ));
+    components.push("methods:".to_string());
+    if let Some(Json::Array(methods)) = data.get("methods") {
+        for method in methods {
+            let name = str_or(method.get("name"), "None");
+            let desc = str_or(method.get("description"), "No description");
+            components.push(format!("    {name}: {desc}"));
+            components.push("        params:".to_string());
+            if let Some(Json::Array(params)) = method.get("params") {
+                for param in params {
+                    let ptype = str_or(param.get("schema").and_then(|s| s.get("type")), "N/A");
+                    let pname = str_or(param.get("name"), "None");
+                    components.push(format!("            - {pname}: {ptype}"));
+                }
+            }
+            let result = str_or(method.get("result").and_then(|r| r.get("name")), "N/A");
+            let result_desc = str_or(
+                method.get("result").and_then(|r| r.get("description")),
+                "No description",
+            );
+            components.push(format!("        result: {result} = {result_desc}"));
+        }
+    }
+    Ok(components)
+}
+
+/// Legacy `parse_csv`: header columns, or a count when there are > 11.
+pub fn extract_csv(content: &str) -> ExtractResult {
+    let rows = splitlines(content);
+    let first = rows
+        .first()
+        .ok_or_else(|| ExtractError::Parse("empty csv".into()))?;
+    let columns: Vec<String> = first.split(',').map(|c| c.to_string()).collect();
+    if columns.len() <= 11 {
+        Ok(columns)
+    } else {
+        Ok(vec![format!("{} columns", columns.len())])
+    }
+}
+
+// --- YAML ---------------------------------------------------------------
+
+use serde::Deserialize;
+use serde_yaml::Value as Yaml;
+
+fn yaml_str(v: &Yaml) -> String {
+    match v {
+        Yaml::Null => "None".to_string(),
+        Yaml::Bool(b) => if *b { "True" } else { "False" }.to_string(),
+        Yaml::Number(n) => n.to_string(),
+        Yaml::String(s) => s.clone(),
+        other => format!("{other:?}"), // legacy would have raised; rare
+    }
+}
+
+fn yaml_get<'a>(v: &'a Yaml, key: &str) -> Option<&'a Yaml> {
+    v.as_mapping()
+        .and_then(|m| m.get(Yaml::String(key.to_string())))
+}
+
+fn yaml_has(v: &Yaml, key: &str) -> bool {
+    yaml_get(v, key).is_some()
+}
+
+/// Legacy `parse_yml` with category detection.
+pub fn extract_yml(content: &str) -> ExtractResult {
+    let mut docs: Vec<Yaml> = Vec::new();
+    for doc in serde_yaml::Deserializer::from_str(content) {
+        let value = Yaml::deserialize(doc).map_err(parse_err)?;
+        docs.push(value);
+    }
+    // python yaml.safe_load_all yields None for empty docs; an empty file
+    // yields no docs at all
+    if docs.is_empty() {
+        return Ok(Vec::new());
+    }
+    let first = &docs[0];
+    if yaml_has(first, "apiVersion") && yaml_has(first, "kind") && yaml_has(first, "metadata") {
+        return Ok(extract_k8s(&docs).unwrap_or_else(unsupported_yaml));
+    }
+    if first
+        .as_sequence()
+        .map(|seq| {
+            seq.iter()
+                .any(|item| item.as_mapping().is_some_and(|_| yaml_has(item, "name")))
+        })
+        .unwrap_or(false)
+    {
+        return Ok(extract_ansible(&docs).unwrap_or_else(unsupported_yaml));
+    }
+    if yaml_has(first, "name") && yaml_has(first, "jobs") {
+        return Ok(extract_github(&docs).unwrap_or_else(unsupported_yaml));
+    }
+    if yaml_has(first, "openapi") || yaml_has(first, "swagger") {
+        return Ok(extract_openapi(&docs).unwrap_or_else(unsupported_yaml));
+    }
+    Ok(vec!["Unsupported YAML Category".to_string()])
+}
+
+fn unsupported_yaml(_: ()) -> Vec<String> {
+    vec!["Unsupported YAML Category".to_string()]
+}
+
+fn extract_k8s(docs: &[Yaml]) -> Result<Vec<String>, ()> {
+    let mut result = Vec::new();
+    for yml in docs {
+        let api = yaml_get(yml, "apiVersion").ok_or(())?;
+        let kind = yaml_get(yml, "kind").ok_or(())?;
+        let name = yaml_get(yml, "metadata")
+            .and_then(|m| yaml_get(m, "name"))
+            .ok_or(())?;
+        result.push(format!(
+            "{}.{} -> {}",
+            yaml_str(api),
+            yaml_str(kind),
+            yaml_str(name)
+        ));
+    }
+    Ok(result)
+}
+
+fn extract_ansible(docs: &[Yaml]) -> Result<Vec<String>, ()> {
+    let mut result = Vec::new();
+    for yml in docs {
+        let seq = yml.as_sequence().ok_or(())?;
+        for item in seq {
+            if item.as_mapping().is_some() {
+                let name = yaml_get(item, "name").map(yaml_str).unwrap_or_default();
+                result.push(name);
+            }
+        }
+    }
+    Ok(result)
+}
+
+fn extract_github(docs: &[Yaml]) -> Result<Vec<String>, ()> {
+    let yml = &docs[0];
+    let mut result = vec![yaml_str(yaml_get(yml, "name").ok_or(())?)];
+    let jobs = yaml_get(yml, "jobs")
+        .and_then(|j| j.as_mapping())
+        .ok_or(())?;
+    for (job_key, job) in jobs {
+        result.push(format!("  job: {}", yaml_str(job_key)));
+        if let Some(steps) = yaml_get(job, "steps").and_then(|s| s.as_sequence()) {
+            for step in steps {
+                if let Some(name) = yaml_get(step, "name") {
+                    result.push(format!("    - {}", yaml_str(name)));
+                }
+            }
+        }
+    }
+    Ok(result)
+}
+
+fn extract_openapi(docs: &[Yaml]) -> Result<Vec<String>, ()> {
+    let yml = &docs[0];
+    let mut components = Vec::new();
+    let openapi = yaml_get(yml, "openapi")
+        .or_else(|| yaml_get(yml, "swagger"))
+        .map(yaml_str)
+        .unwrap_or_else(|| "N/A".to_string());
+    components.push(format!("openapi: {openapi}"));
+    let info = yaml_get(yml, "info");
+    let title = info
+        .and_then(|i| yaml_get(i, "title"))
+        .map(yaml_str)
+        .unwrap_or_else(|| "N/A".to_string());
+    components.push(format!("    title: {title}"));
+    let description = info
+        .and_then(|i| yaml_get(i, "description"))
+        .map(yaml_str)
+        .unwrap_or_default();
+    let first_sentence = if description.is_empty() {
+        "N/A".to_string()
+    } else {
+        // legacy: regex.split(r"\.\s|\.\n", description, maxsplit=1)[0] + "."
+        let split_at = description
+            .char_indices()
+            .find(|&(i, c)| {
+                c == '.'
+                    && description[i + 1..]
+                        .chars()
+                        .next()
+                        .is_some_and(|n| n.is_whitespace())
+            })
+            .map(|(i, _)| i);
+        match split_at {
+            Some(i) => format!("{}.", &description[..i]),
+            None => format!("{description}."),
+        }
+    };
+    components.push(format!("    description: {first_sentence}"));
+    let version = info
+        .and_then(|i| yaml_get(i, "version"))
+        .map(yaml_str)
+        .unwrap_or_else(|| "N/A".to_string());
+    components.push(format!("    version: {version}"));
+    if let Some(servers) = yaml_get(yml, "servers").and_then(|s| s.as_sequence()) {
+        if !servers.is_empty() {
+            components.push("servers:".to_string());
+            for server in servers {
+                let url = yaml_get(server, "url")
+                    .map(yaml_str)
+                    .unwrap_or_else(|| "N/A".to_string());
+                components.push(format!("    - url: {url}"));
+            }
+        }
+    }
+    if let Some(paths) = yaml_get(yml, "paths").and_then(|p| p.as_mapping()) {
+        if !paths.is_empty() {
+            components.push("paths:".to_string());
+            for (path, methods) in paths {
+                components.push(format!("    '{}':", yaml_str(path)));
+                if let Some(methods) = methods.as_mapping() {
+                    for (method, details) in methods {
+                        let op = yaml_get(details, "operationId")
+                            .map(yaml_str)
+                            .unwrap_or_else(|| "N/A".to_string());
+                        let summary = yaml_get(details, "summary")
+                            .map(yaml_str)
+                            .unwrap_or_else(|| "N/A".to_string());
+                        components.push(format!(
+                            "        {} ({op}): {summary}",
+                            yaml_str(method).to_uppercase()
+                        ));
+                    }
+                }
+            }
+        }
+    }
+    Ok(components)
+}
+
+// --- TOML ---------------------------------------------------------------
+
+use toml::Value as Toml;
+
+fn toml_str(v: &Toml) -> String {
+    match v {
+        Toml::String(s) => s.clone(),
+        Toml::Integer(i) => i.to_string(),
+        Toml::Float(f) => f.to_string(),
+        Toml::Boolean(b) => if *b { "True" } else { "False" }.to_string(),
+        other => other.to_string(),
+    }
+}
+
+/// Legacy `format_dependency`.
+fn format_dependency(name: &str, details: &Toml) -> String {
+    match details {
+        Toml::String(s) => format!("  {name} {s}"),
+        Toml::Table(t) => {
+            let version = t.get("version").map(toml_str).unwrap_or_default();
+            let features: Vec<String> = t
+                .get("features")
+                .and_then(|f| f.as_array())
+                .map(|arr| arr.iter().map(toml_str).collect())
+                .unwrap_or_default();
+            if features.is_empty() {
+                format!("  {name} {version}")
+            } else {
+                format!("  {name} {version} (features: {})", features.join(", "))
+            }
+        }
+        _ => format!("  {name}"),
+    }
+}
+
+/// Legacy `parse_cargo_toml`.
+pub fn extract_cargo_toml(content: &str) -> ExtractResult {
+    let data: Toml = content.parse().map_err(parse_err)?;
+    let mut components = Vec::new();
+    if let Some(package) = data.get("package") {
+        for key in ["name", "version", "description", "license"] {
+            let value = package
+                .get(key)
+                .map(toml_str)
+                .unwrap_or_else(|| "N/A".to_string());
+            components.push(format!("{key}: {value}"));
+        }
+    }
+    if let Some(Toml::Table(deps)) = data.get("dependencies") {
+        components.push("dependencies:".to_string());
+        for (dep, details) in deps {
+            components.push(format_dependency(dep, details));
+        }
+    }
+    Ok(components)
+}
+
+/// Legacy `parse_pyproject_toml`.
+pub fn extract_pyproject_toml(content: &str) -> ExtractResult {
+    let data: Toml = content.parse().map_err(parse_err)?;
+    let mut components = Vec::new();
+    let project = data.get("project");
+    if let Some(project) = project {
+        for key in ["name", "version", "description"] {
+            let value = project
+                .get(key)
+                .map(toml_str)
+                .unwrap_or_else(|| "N/A".to_string());
+            components.push(format!("{key}: {value}"));
+        }
+        if let Some(classifiers) = project.get("classifiers").and_then(|c| c.as_array()) {
+            for classifier in classifiers {
+                let c = toml_str(classifier);
+                if c.contains("License ::") {
+                    components.push(c);
+                }
+            }
+        }
+        if let Some(deps) = project.get("dependencies").and_then(|d| d.as_array()) {
+            components.push("dependencies:".to_string());
+            for dep in deps {
+                components.push(format!("    {}", toml_str(dep)));
+            }
+        }
+    }
+    Ok(components)
+}
+
+// --- SQLite ---------------------------------------------------------------
+
+/// Legacy `parse_db`: tables and columns from a SQLite database.
+#[cfg(feature = "sqlite")]
+pub fn extract_sqlite(path: &std::path::Path) -> ExtractResult {
+    let conn =
+        rusqlite::Connection::open_with_flags(path, rusqlite::OpenFlags::SQLITE_OPEN_READ_ONLY)
+            .map_err(parse_err)?;
+    let mut stmt = conn
+        .prepare("SELECT name FROM sqlite_master WHERE type='table';")
+        .map_err(parse_err)?;
+    let tables: Vec<String> = stmt
+        .query_map([], |row| row.get::<_, String>(0))
+        .map_err(parse_err)?
+        .filter_map(Result::ok)
+        .collect();
+    let mut components = Vec::new();
+    for table in tables {
+        components.push(format!("{table} table:"));
+        let mut info = conn
+            .prepare(&format!("PRAGMA table_info({table});"))
+            .map_err(parse_err)?;
+        let rows = info
+            .query_map([], |row| {
+                Ok((
+                    row.get::<_, String>(1)?, // name
+                    row.get::<_, String>(2)?, // type
+                    row.get::<_, i64>(3)?,    // notnull
+                    row.get::<_, i64>(5)?,    // pk
+                ))
+            })
+            .map_err(parse_err)?;
+        for row in rows.filter_map(Result::ok) {
+            let (name, ctype, not_null, pk) = row;
+            let mut desc = format!("   {name} {}", ctype.to_lowercase());
+            if pk == 1 {
+                desc.push_str(" primary key");
+            }
+            if not_null == 1 && pk != 1 {
+                desc.push_str(" not null");
+            }
+            components.push(desc);
+        }
+    }
+    Ok(components)
+}
+
+#[cfg(not(feature = "sqlite"))]
+pub fn extract_sqlite(_path: &std::path::Path) -> ExtractResult {
+    Ok(Vec::new())
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn jsonl_types() {
+        let line = r#"{"a": 1, "b": "x", "c": [1], "d": {"k": 1}, "e": null, "f": 1.5, "g": true}"#;
+        assert_eq!(
+            extract_jsonl(line).unwrap(),
+            vec!["a: int", "b: str", "c: list", "d: dict", "e: None", "f: float", "g: bool"]
+        );
+    }
+
+    #[test]
+    fn package_json() {
+        let content = r#"{"name": "pkg", "version": "1.0.0", "scripts": {"test": "jest"}}"#;
+        assert_eq!(
+            extract_package_json(content).unwrap(),
+            vec![
+                "name: 'pkg'",
+                "version: 1.0.0",
+                "scripts:",
+                "    test: 'jest'"
+            ]
+        );
+    }
+
+    #[test]
+    fn csv_columns() {
+        assert_eq!(extract_csv("a,b,c\n1,2,3\n").unwrap(), vec!["a", "b", "c"]);
+        let wide = (0..12).map(|i| i.to_string()).collect::<Vec<_>>().join(",");
+        assert_eq!(extract_csv(&wide).unwrap(), vec!["12 columns"]);
+    }
+
+    #[test]
+    fn github_yaml() {
+        let content =
+            "name: CI\njobs:\n  build:\n    steps:\n      - name: checkout\n      - run: make\n";
+        assert_eq!(
+            extract_yml(content).unwrap(),
+            vec!["CI", "  job: build", "    - checkout"]
+        );
+    }
+
+    #[test]
+    fn unsupported_yaml_category() {
+        assert_eq!(
+            extract_yml("just: a mapping\n").unwrap(),
+            vec!["Unsupported YAML Category"]
+        );
+    }
+
+    #[test]
+    fn cargo_toml() {
+        let content = "[package]\nname = \"x\"\nversion = \"0.1.0\"\n[dependencies]\nserde = { version = \"1\", features = [\"derive\"] }\nregex = \"1\"\n";
+        assert_eq!(
+            extract_cargo_toml(content).unwrap(),
+            vec![
+                "name: x",
+                "version: 0.1.0",
+                "description: N/A",
+                "license: N/A",
+                "dependencies:",
+                "  serde 1 (features: derive)",
+                "  regex 1",
+            ]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/markdown.rs b/crates/tree_plus_core/src/extract/markdown.rs
new file mode 100644
index 0000000..2278f25
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/markdown.rs
@@ -0,0 +1,180 @@
+//! Markdown and plain-text extraction (legacy `parse_md`, `parse_txt`).
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+
+const MARKDOWN_LANGUAGES: &[&str] = &["md", "markdown", "mdx", "mdc"];
+
+static TASK_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"(-\s*\[ *[xX]?\])\s*(.*)").unwrap());
+static URL_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r#"\s*\(<a href=".*">.+</a>\)|<a href=".*">.+</a>"#).unwrap());
+static LINK_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"\(.*?\)").unwrap());
+
+/// Extract headers and tasks from Markdown (legacy `parse_md`).
+pub fn extract_md(content: &str) -> Vec<String> {
+    let mut headers_and_tasks: Vec<String> = Vec::new();
+    // (task_text, include_flag) of checked ancestors
+    let mut checked_ancestors: Vec<(String, bool)> = Vec::new();
+    let mut code_block_stack: Vec<String> = Vec::new();
+
+    for line in content.split('\n') {
+        let line = line.strip_suffix('\r').unwrap_or(line);
+        let stripped_line = line.trim();
+        if let Some(after_fence) = stripped_line.strip_prefix("```") {
+            if !code_block_stack.is_empty() {
+                if stripped_line == "```" {
+                    code_block_stack.pop();
+                } else {
+                    code_block_stack.push(after_fence.trim().to_string());
+                }
+            } else {
+                code_block_stack.push(after_fence.trim().to_string());
+            }
+            continue;
+        }
+
+        let in_opaque_block = code_block_stack
+            .last()
+            .map(|lang| !MARKDOWN_LANGUAGES.contains(&lang.as_str()))
+            .unwrap_or(false);
+        if in_opaque_block {
+            continue;
+        }
+
+        if line.starts_with('#') {
+            let line = line.trim_start();
+            let clean = URL_RE.replace_all(line, "");
+            let clean = LINK_RE.replace_all(&clean, "");
+            let clean = clean.trim();
+            if clean.trim_start_matches('#').trim().is_empty() {
+                continue;
+            }
+            headers_and_tasks.push(clean.to_string());
+        } else if TASK_RE.is_match(line.trim_start())
+            && TASK_RE
+                .captures(line.trim_start())
+                .map(|c| c.get(0).map(|m| m.start() == 0).unwrap_or(false))
+                .unwrap_or(false)
+        {
+            let lstripped = line.trim_start();
+            let indent_level = line.len() - lstripped.len();
+            if let Some(caps) = TASK_RE.captures(lstripped) {
+                let task_text = caps.get(2).map(|m| m.as_str()).unwrap_or("");
+                let is_checked = line.contains("[x]") || line.contains("[X]");
+                let task = format!(
+                    "{}{}{}",
+                    " ".repeat(indent_level),
+                    if is_checked { "- [x] " } else { "- [ ] " },
+                    task_text
+                );
+                checked_ancestors.retain(|a| a.0.len() < task.len());
+                if is_checked {
+                    checked_ancestors.push((task, false));
+                } else {
+                    for a in checked_ancestors.iter_mut() {
+                        a.1 = true;
+                    }
+                    headers_and_tasks.extend(checked_ancestors.iter().map(|a| a.0.clone()));
+                    headers_and_tasks.push(task);
+                }
+            }
+        }
+    }
+    headers_and_tasks
+}
+
+static CHECKBOX_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"-\s*\[\s*([^Xx])?\s*\]\s*(.+)").unwrap());
+
+/// Extract unchecked checkboxes from plain text (legacy `parse_txt`).
+pub fn extract_txt(content: &str) -> Vec<String> {
+    content
+        .split('\n')
+        .filter_map(|line| {
+            CHECKBOX_RE
+                .captures(line)
+                .and_then(|caps| caps.get(2))
+                .map(|m| format!("- [ ] {}", m.as_str().trim()))
+        })
+        .collect()
+}
+
+/// Extract RST headings (legacy `parse_rst`): a content line followed by an
+/// underline of `=` (heading, `# `) or `-` (subheading, `- `).
+pub fn extract_rst(content: &str) -> Vec<String> {
+    let lines: Vec<&str> = content.split('\n').collect();
+    let mut components = Vec::new();
+    let mut i = 0;
+    while i + 1 < lines.len() {
+        let underline = lines[i + 1].trim_end_matches('\r');
+        let is_sub = !underline.is_empty() && underline.chars().all(|c| c == '-');
+        let is_head = !underline.is_empty() && underline.chars().all(|c| c == '=');
+        if is_sub || is_head {
+            let text = lines[i].trim_end_matches('\r');
+            let prefix = if is_head { "# " } else { "- " };
+            components.push(format!("{prefix}{text}"));
+            i += 2;
+        } else {
+            i += 1;
+        }
+    }
+    components
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn headers_and_code_blocks() {
+        let content = "# Title\n```rust\n# not a header\n```\n## Sub\n";
+        assert_eq!(extract_md(content), vec!["# Title", "## Sub"]);
+    }
+
+    #[test]
+    fn markdown_code_blocks_stay_transparent() {
+        let content = "# A\n```md\n# Inner\n```\n";
+        assert_eq!(extract_md(content), vec!["# A", "# Inner"]);
+    }
+
+    #[test]
+    fn checked_ancestors_logic() {
+        let content = "- [x] done parent\n  - [ ] open child\n";
+        assert_eq!(
+            extract_md(content),
+            vec!["- [x] done parent", "  - [ ] open child"]
+        );
+    }
+
+    #[test]
+    fn checked_without_open_children_hidden() {
+        let content = "- [x] done alone\n- [ ] open\n";
+        assert_eq!(extract_md(content), vec!["- [ ] open"]);
+    }
+
+    #[test]
+    fn header_links_removed() {
+        assert_eq!(
+            extract_md("# Hello [link](https://x.y) world\n"),
+            vec!["# Hello [link] world"]
+        );
+    }
+
+    #[test]
+    fn txt_checkboxes() {
+        assert_eq!(
+            extract_txt("- [ ] do it\n- [x] done\ntext\n"),
+            vec!["- [ ] do it"]
+        );
+    }
+
+    #[test]
+    fn rst_headings() {
+        assert_eq!(
+            extract_rst("Title\n=====\n\nSub\n---\n"),
+            vec!["# Title", "- Sub"]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/markers.rs b/crates/tree_plus_core/src/extract/markers.rs
new file mode 100644
index 0000000..ed065dc
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/markers.rs
@@ -0,0 +1,60 @@
+//! TODO / BUG / NOTE marker extraction (legacy `parse_markers`).
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+
+static MARKER_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"(?P<mark>BUG|TODO|NOTE)(?P<mention> ?\([@\w ]+\) ?)?: (?P<msg>.*)").unwrap()
+});
+
+/// Extract `MARK: message` lines from content.
+pub fn extract_markers(content: &str) -> Vec<String> {
+    MARKER_RE
+        .captures_iter(content)
+        .filter_map(|caps| {
+            let mark = caps.name("mark")?.as_str();
+            let msg = caps.name("msg")?.as_str();
+            if msg.is_empty() {
+                // legacy: falsy msg ("" only, since trailing newline is excluded)
+                return None;
+            }
+            Some(format!("{}: {}", mark, msg.trim()))
+        })
+        .collect()
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn bug_todo_note() {
+        let content = "BUG: This is a bug.\nTODO: Fix this soon.\nNOTE: Interesting observation.";
+        assert_eq!(
+            extract_markers(content),
+            vec![
+                "BUG: This is a bug.",
+                "TODO: Fix this soon.",
+                "NOTE: Interesting observation.",
+            ]
+        );
+    }
+
+    #[test]
+    fn mention_is_dropped() {
+        assert_eq!(
+            extract_markers("TODO (@bion): hi there"),
+            vec!["TODO: hi there"]
+        );
+    }
+
+    #[test]
+    fn empty_msg_skipped() {
+        // ": " is part of the pattern, so "TODO: \n" has an empty msg
+        assert!(extract_markers("TODO: \n").is_empty());
+        // msg=" " (extra space) is truthy; strip() empties it -> "TODO: "
+        assert_eq!(extract_markers("TODO:  \n"), vec!["TODO: "]);
+        assert!(extract_markers("TODO:\n").is_empty());
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/mod.rs b/crates/tree_plus_core/src/extract/mod.rs
new file mode 100644
index 0000000..c3572ff
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/mod.rs
@@ -0,0 +1,277 @@
+//! Component extraction: dispatch a file to the right extractor.
+//!
+//! This is the Rust analog of the legacy `parse_file` (renamed: it does not
+//! parse files generically, it extracts displayable component labels).
+//! Dispatch order and special cases deliberately mirror
+//! `tree_plus_src/parse_file.py` because legacy behavior depends on them.
+
+pub mod data;
+pub mod markdown;
+pub mod markers;
+pub mod simple;
+pub mod treesitter;
+
+use std::path::Path;
+
+/// Errors during extraction. Legacy behavior maps any extractor error to an
+/// empty component list at the engine layer (including the marker pass).
+#[derive(Debug, thiserror::Error)]
+pub enum ExtractError {
+    #[error("io: {0}")]
+    Io(#[from] std::io::Error),
+    #[error("parse: {0}")]
+    Parse(String),
+}
+
+pub type ExtractResult = Result<Vec<String>, ExtractError>;
+
+const BINARY_CHECK_SIZE: usize = 1024;
+
+/// Legacy `is_binary_string`: any byte outside the text set means binary.
+/// Text bytes: {7, 8, 9, 10, 12, 13, 27} | [0x20, 0xFF] - {0x7F}.
+pub fn is_binary_bytes(data: &[u8]) -> bool {
+    data.iter().any(|&b| {
+        !(matches!(b, 7 | 8 | 9 | 10 | 12 | 13 | 27) || (0x20..=0xFF).contains(&b) && b != 0x7F)
+    })
+}
+
+/// Check the first KiB of a file (legacy `is_binary`).
+pub fn is_binary(path: &Path) -> bool {
+    use std::io::Read;
+    let Ok(mut f) = std::fs::File::open(path) else {
+        return false;
+    };
+    let mut buf = [0u8; BINARY_CHECK_SIZE];
+    let mut filled = 0;
+    while filled < buf.len() {
+        match f.read(&mut buf[filled..]) {
+            Ok(0) => break,
+            Ok(n) => filled += n,
+            Err(_) => return false,
+        }
+    }
+    is_binary_bytes(&buf[..filled])
+}
+
+/// `os.path.splitext`-compatible split: returns (stem path, lowered ".ext").
+fn splitext_lower(path: &str) -> (String, String) {
+    let (dir, name) = match path.rfind('/') {
+        Some(i) => (&path[..=i], &path[i + 1..]),
+        None => ("", path),
+    };
+    let trimmed = name.trim_start_matches('.');
+    let n_leading = name.len() - trimmed.len();
+    match trimmed.rfind('.') {
+        Some(i) => {
+            let split_at = n_leading + i;
+            (
+                format!("{dir}{}", &name[..split_at]),
+                name[split_at..].to_lowercase(),
+            )
+        }
+        None => (path.to_string(), String::new()),
+    }
+}
+
+/// Read the first `n` lines like the legacy `read_file(n_lines=n)`:
+/// lines keep their trailing newline and are joined with an extra `\n`.
+/// Returns an empty string when the file has fewer than `n` lines
+/// (legacy StopIteration behavior).
+fn read_first_lines(content: &str, n: usize) -> String {
+    let mut lines = Vec::with_capacity(n);
+    let mut rest = content;
+    for _ in 0..n {
+        match rest.find('\n') {
+            Some(i) => {
+                lines.push(&rest[..=i]);
+                rest = &rest[i + 1..];
+            }
+            None => {
+                if rest.is_empty() {
+                    return String::new(); // StopIteration -> ""
+                }
+                lines.push(rest);
+                rest = "";
+            }
+        }
+    }
+    lines.join("\n")
+}
+
+/// JS-family extensions (legacy `JS_EXTENSIONS`).
+const JS_EXTENSIONS: &[&str] = &[".js", ".jsx", ".ts", ".tsx"];
+/// C-family extensions (legacy `C_EXTENSIONS`).
+const C_EXTENSIONS: &[&str] = &[".c", ".cpp", ".cc", ".h", ".cu", ".cuh", ".hpp"];
+const PYTHON_EXTENSIONS: &[&str] = &[".py", ".pyi"];
+const MARKDOWN_EXTENSIONS: &[&str] = &[".md", ".markdown", ".mdx", ".mdc"];
+
+/// Extensions recognized by the legacy implementation but deferred in the
+/// Rust port version 1. Files still get TODO/BUG/NOTE markers; component
+/// extraction is tracked in docs/language-roadmap.md.
+const DEFERRED_EXTENSIONS: &[&str] = &[
+    ".php", ".kt", ".swift", ".go", ".sh", ".ps1", ".zig", ".rb", ".sql", ".graphql", ".cs", ".jl",
+    ".scala", ".java", ".pl", ".hs", ".fs", ".lisp", ".clj", ".scm", ".el", ".rkt", ".erl", ".hrl",
+    ".capnp", ".proto", ".tex", ".lean", ".f", ".for", ".f77", ".f90", ".f95", ".f03", ".f08",
+    ".tf", ".thy", ".lua", ".tcl", ".m", ".r", ".nb", ".wl", ".matlab", ".ml", ".cbl", ".cobol",
+    ".apl", ".metal", ".wgsl", ".html",
+];
+
+/// Whether this extension is deferred (legacy support, no Rust port yet).
+pub fn is_deferred_extension(ext: &str) -> bool {
+    DEFERRED_EXTENSIONS.contains(&ext)
+}
+
+/// Extract displayable component labels from a file.
+///
+/// `syntax` mirrors the legacy flag: when false, Rust enum components are
+/// rich-markup-escaped exactly like the legacy renderer expected.
+pub fn extract_components(path: &Path, syntax: bool) -> Vec<String> {
+    try_extract_components(path, syntax).unwrap_or_default()
+}
+
+fn try_extract_components(path: &Path, syntax: bool) -> ExtractResult {
+    let path_str = path.to_string_lossy().replace('\\', "/");
+    let (base_path, ext) = splitext_lower(&path_str);
+    let file_name = base_path
+        .rsplit('/')
+        .next()
+        .unwrap_or(&base_path)
+        .to_string();
+
+    // sqlite databases are handled before reading the file
+    if ext == ".db" || ext == ".sqlite" {
+        return data::extract_sqlite(path);
+    }
+
+    if is_binary(path) {
+        return Ok(Vec::new());
+    }
+
+    let raw = std::fs::read(path)?;
+    let full_content = String::from_utf8_lossy(&raw).into_owned();
+    // legacy read_file uses errors="strict": undecodable files read as ""
+    let full_content = if std::str::from_utf8(&raw).is_ok() {
+        full_content
+    } else {
+        String::new()
+    };
+
+    // big data files only read a few lines
+    let content: String = match ext.as_str() {
+        ".csv" => read_first_lines(&full_content, 3),
+        ".jsonl" => read_first_lines(&full_content, 2),
+        _ => full_content,
+    };
+
+    let components: Vec<String> = match ext.as_str() {
+        e if JS_EXTENSIONS.contains(&e) => {
+            if path_str.ends_with(".d.ts") {
+                Vec::new() // legacy parse_d_dot_ts returns [] (it never appends)
+            } else {
+                let mut components =
+                    treesitter::typescript::extract(&content, e == ".tsx" || e == ".jsx")?;
+                if path_str.contains("spec.ts") {
+                    components = simple::angular_spec(&content);
+                }
+                if path_str.contains("app-routing.module") {
+                    let mut routes = simple::angular_routes(&content);
+                    routes.extend(components);
+                    components = routes;
+                } else if path_str.contains("app.module") {
+                    let mut module = simple::angular_app_module(&content);
+                    module.extend(components);
+                    components = module;
+                } else if file_name == "environment" || file_name.contains("environment.") {
+                    components = simple::environment_ts(&content);
+                }
+                components
+            }
+        }
+        e if PYTHON_EXTENSIONS.contains(&e) => treesitter::python::extract(&content)?,
+        e if MARKDOWN_EXTENSIONS.contains(&e) => markdown::extract_md(&content),
+        ".rst" => markdown::extract_rst(&content),
+        ".json" => {
+            if path_str.to_lowercase().contains("package.json") {
+                data::extract_package_json(&content)?
+            } else if content.contains("$schema") {
+                data::extract_json_schema(&content)?
+            } else if content.contains("jsonrpc\": \"2") {
+                data::extract_json_rpc(&content)?
+            } else if content.contains("openrpc\": \"") {
+                data::extract_openrpc_json(&content)?
+            } else {
+                Vec::new()
+            }
+        }
+        ".yml" | ".yaml" => data::extract_yml(&content)?,
+        e if C_EXTENSIONS.contains(&e) => treesitter::c_cpp::extract(&content, e)?,
+        ".rs" => treesitter::rust::extract(&content, syntax)?,
+        ".jsonl" => data::extract_jsonl(&content)?,
+        ".env" => simple::dot_env(&content),
+        ".txt" => {
+            if file_name.contains("requirements") {
+                simple::requirements_txt(&content)
+            } else {
+                markdown::extract_txt(&content)
+            }
+        }
+        ".csv" => data::extract_csv(&content)?,
+        "" if file_name == "Makefile" || file_name == "Justfile" => simple::makefile(&content),
+        _ if file_name.starts_with(".env") => simple::dot_env(&content),
+        _ if path_str.ends_with("Cargo.toml") => data::extract_cargo_toml(&content)?,
+        _ if path_str.ends_with("pyproject.toml") => data::extract_pyproject_toml(&content)?,
+        _ => Vec::new(),
+    };
+
+    // markers skip exactly .txt and .md (legacy quirk: not .markdown/.mdx)
+    if ext == ".txt" || ext == ".md" {
+        Ok(components)
+    } else {
+        let mut total = markers::extract_markers(&content);
+        total.extend(components);
+        Ok(total)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn binary_detection() {
+        assert!(is_binary_bytes(&[0x00, 0x01]));
+        assert!(!is_binary_bytes(b"hello world\n"));
+        assert!(!is_binary_bytes("héllo".as_bytes())); // UTF-8 high bytes are text
+        assert!(is_binary_bytes(&[0x7F]));
+    }
+
+    #[test]
+    fn splitext_examples() {
+        assert_eq!(
+            splitext_lower("a/b/file.PY"),
+            ("a/b/file".to_string(), ".py".to_string())
+        );
+        assert_eq!(
+            splitext_lower(".env.test"),
+            (".env".to_string(), ".test".to_string())
+        );
+        assert_eq!(
+            splitext_lower("dir/.env"),
+            ("dir/.env".to_string(), String::new())
+        );
+        assert_eq!(
+            splitext_lower("Makefile"),
+            ("Makefile".to_string(), String::new())
+        );
+    }
+
+    #[test]
+    fn first_lines_eof_means_empty() {
+        assert_eq!(
+            read_first_lines("a,b\nc,d\ne,f\ng\n", 3),
+            "a,b\n\nc,d\n\ne,f\n"
+        );
+        assert_eq!(read_first_lines("a,b\n", 3), "");
+        assert_eq!(read_first_lines("", 2), "");
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/simple.rs b/crates/tree_plus_core/src/extract/simple.rs
new file mode 100644
index 0000000..38a00e1
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/simple.rs
@@ -0,0 +1,216 @@
+//! Line-oriented extractors: .env, requirements.txt, Makefile/Justfile,
+//! and the Angular special cases from the legacy TypeScript dispatch.
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+
+static ENV_KEY_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"([A-Z|_]*)=").unwrap());
+
+/// Python-`str.splitlines()`-alike for `\n` / `\r\n` / `\r` (no trailing empty).
+pub fn splitlines(content: &str) -> Vec<&str> {
+    let mut out = Vec::new();
+    let mut start = 0;
+    let bytes = content.as_bytes();
+    let mut i = 0;
+    while i < bytes.len() {
+        match bytes[i] {
+            b'\n' => {
+                out.push(&content[start..i]);
+                i += 1;
+                start = i;
+            }
+            b'\r' => {
+                out.push(&content[start..i]);
+                i += 1;
+                if i < bytes.len() && bytes[i] == b'\n' {
+                    i += 1;
+                }
+                start = i;
+            }
+            _ => i += 1,
+        }
+    }
+    if start < bytes.len() {
+        out.push(&content[start..]);
+    }
+    out
+}
+
+/// Legacy `parse_dot_env`: first `KEY=` match per non-comment line.
+pub fn dot_env(content: &str) -> Vec<String> {
+    splitlines(content)
+        .into_iter()
+        .filter(|line| !line.starts_with('#'))
+        .filter_map(|line| {
+            ENV_KEY_RE
+                .captures(line)
+                .and_then(|c| c.get(1))
+                .map(|m| m.as_str().to_string())
+        })
+        .collect()
+}
+
+/// Legacy `parse_requirements_txt`: every line not starting with `#`.
+pub fn requirements_txt(content: &str) -> Vec<String> {
+    splitlines(content)
+        .into_iter()
+        .filter(|line| !line.starts_with('#'))
+        .map(|line| line.to_string())
+        .collect()
+}
+
+/// Legacy `parse_makefile`: targets, includes, defines, .PHONY lines.
+pub fn makefile(content: &str) -> Vec<String> {
+    content
+        .split('\n')
+        .filter(|line| {
+            let stripped = line.trim();
+            !stripped.is_empty()
+                && !line.starts_with('\t')
+                && !stripped.starts_with('#')
+                && (line.starts_with(".PHONY")
+                    || line.contains(':')
+                    || line.starts_with("include")
+                    || line.starts_with("define"))
+        })
+        .map(|line| line.trim().trim_end_matches(':').to_string())
+        .collect()
+}
+
+static DESCRIBE_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"(\t*| *)describe\('(.*)'").unwrap());
+static IT_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r#"(\t*| *)it\(('|")(.*)('|")"#).unwrap());
+
+/// Legacy `parse_angular_spec`: describe/it lines.
+pub fn angular_spec(content: &str) -> Vec<String> {
+    let mut components = Vec::new();
+    for line in content.split('\n') {
+        if let Some(caps) = DESCRIBE_RE.captures(line) {
+            components.push(format!(
+                "{}describe '{}'",
+                caps.get(1).map_or("", |m| m.as_str()),
+                caps.get(2).map_or("", |m| m.as_str())
+            ));
+        } else if let Some(caps) = IT_RE.captures(line) {
+            let statement = caps
+                .get(3)
+                .map_or("", |m| m.as_str())
+                .replace('\\', "")
+                .replace('"', "'");
+            components.push(format!(
+                "{}it {}",
+                caps.get(1).map_or("", |m| m.as_str()),
+                statement
+            ));
+        }
+    }
+    components
+}
+
+static ROUTES_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"(?s)(const routes: Routes = \[\n.*?\];)").unwrap());
+
+/// Legacy `parse_angular_routes`.
+pub fn angular_routes(content: &str) -> Vec<String> {
+    ROUTES_RE
+        .captures(content)
+        .and_then(|c| c.get(1))
+        .map(|m| vec![m.as_str().to_string()])
+        .unwrap_or_default()
+}
+
+static NG_MODULE_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"(@NgModule\(\{\n\s*declarations: \[((\n\s*)[a-zA-Z]*,?)*)\n").unwrap()
+});
+
+/// Legacy `parse_angular_app_module`.
+pub fn angular_app_module(content: &str) -> Vec<String> {
+    NG_MODULE_RE
+        .captures(content)
+        .and_then(|c| c.get(1))
+        .map(|m| vec![m.as_str().to_string()])
+        .unwrap_or_default()
+}
+
+static ENV_TS_KEY_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"(\w+):").unwrap());
+
+/// Legacy `parse_environment_ts`: keys of `export const environment = {...}`.
+pub fn environment_ts(content: &str) -> Vec<String> {
+    let mut parsing = false;
+    let mut keepers: Vec<String> = Vec::new();
+    for line in content.split('\n') {
+        let stripped = line.trim();
+        if stripped.starts_with("export const environment = {") {
+            parsing = true;
+            continue;
+        }
+        if stripped.starts_with("};") {
+            break;
+        }
+        if stripped.starts_with("//") || stripped.starts_with("/*") || stripped.starts_with('*') {
+            continue;
+        }
+        if parsing {
+            if let Some(caps) = ENV_TS_KEY_RE.captures(line) {
+                keepers.push(format!("   {}", caps.get(1).map_or("", |m| m.as_str())));
+            }
+        }
+    }
+    if keepers.is_empty() {
+        keepers
+    } else {
+        let mut out = vec!["environment:".to_string()];
+        out.extend(keepers);
+        out
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn env_keys() {
+        assert_eq!(
+            dot_env("API_KEY=abc\n# comment SECRET=no\nDB_URL=postgres\n"),
+            vec!["API_KEY", "DB_URL"]
+        );
+    }
+
+    #[test]
+    fn env_lowercase_key_yields_empty_match() {
+        // legacy quirk: ([A-Z|_]*) matches empty before '='
+        assert_eq!(dot_env("lower=1\n"), vec![""]);
+    }
+
+    #[test]
+    fn makefile_targets() {
+        let content =
+            "SHELL := /bin/bash\n.PHONY: test\ntest:\n\tpytest\n# comment:\ninclude foo.mk\n";
+        assert_eq!(
+            makefile(content),
+            vec![
+                "SHELL := /bin/bash",
+                ".PHONY: test",
+                "test",
+                "include foo.mk"
+            ]
+        );
+    }
+
+    #[test]
+    fn requirements_keeps_blank_lines() {
+        assert_eq!(requirements_txt("a==1\n# c\n\nb\n"), vec!["a==1", "", "b"]);
+    }
+
+    #[test]
+    fn spec_describe_it() {
+        let content = "describe('AppComponent', () => {\n  it('should work', () => {});\n});\n";
+        assert_eq!(
+            angular_spec(content),
+            vec!["describe 'AppComponent'", "  it should work"]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs b/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
new file mode 100644
index 0000000..58a4839
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
@@ -0,0 +1,591 @@
+//! C and C++ component extraction (tree-sitter).
+//!
+//! Emits source-text slices in legacy `parse_c` style:
+//! - function definitions (signature through initializer list, no body);
+//! - struct/class/enum definitions with their fields and enumerators;
+//! - access specifiers, typedef struct + `} name;` closers, #define macros,
+//!   template declarations.
+//!
+//! Intentional differences from the legacy regex (documented in
+//! docs/rust-port-differences.md): control-flow noise (e.g. `while(...)`
+//! lines), string contents, and the TensorFlow flag special case are not
+//! emitted.
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+use tree_sitter::Node;
+
+use super::rust::strip_c_comments;
+use super::{parse, ExtractResult};
+
+/// Legacy "function" form: `^ *(modifier )?(return_type )(name)\(args\)` with
+/// at most two words before the name and no `)` inside the args.
+static FUNCTION_FORM_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"^ *(?:[\w:]+ )?[\w:*&]+(?:\s?<[^>]*>\s?)? [\w*&\[\]]+\([^)]*\)$").unwrap()
+});
+
+/// Legacy "method" form prefix (before the parameter list): indented,
+/// optional virtual/static, optional single-word return type, `~?[*\w]+`.
+static METHOD_PREFIX_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^\s+(?:virtual |static )?(?:\w+ )?~?[*\w]+$").unwrap());
+
+/// Extract C/C++ components. `ext` picks the grammar (.c/.h use C unless C++
+/// constructs are likely; .cpp/.cc/.cu/.cuh/.hpp use C++).
+pub fn extract(content: &str, ext: &str) -> ExtractResult {
+    let use_cpp =
+        matches!(ext, ".cpp" | ".cc" | ".cu" | ".cuh" | ".hpp") || looks_like_cpp(content);
+    let language: tree_sitter::Language = if use_cpp {
+        tree_sitter_cpp::LANGUAGE.into()
+    } else {
+        tree_sitter_c::LANGUAGE.into()
+    };
+    let tree = parse(content, &language)?;
+    let mut extractor = CExtractor {
+        content,
+        components: Vec::new(),
+        suppress_plain_fields: 0,
+    };
+    extractor.visit(tree.root_node());
+    Ok(extractor.components)
+}
+
+/// Cheap heuristic so C++ headers using .h still parse with the C++ grammar.
+fn looks_like_cpp(content: &str) -> bool {
+    content.contains("class ")
+        || content.contains("namespace ")
+        || content.contains("template")
+        || content.contains("::")
+}
+
+struct CExtractor<'a> {
+    content: &'a str,
+    components: Vec<String>,
+    /// Legacy quirk: a `template<...> class/struct` line reset the regex
+    /// context, so plain data fields inside template records were skipped.
+    suppress_plain_fields: usize,
+}
+
+fn line_start(content: &str, byte: usize) -> usize {
+    content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
+}
+
+impl<'a> CExtractor<'a> {
+    fn slice_clean(&self, from: usize, to: usize) -> String {
+        // legacy parse_c ran on comment-stripped content
+        strip_c_comments(&self.content[from..to])
+            .trim_end()
+            .trim_start_matches('\n')
+            .to_string()
+    }
+
+    fn visit(&mut self, node: Node<'a>) {
+        match node.kind() {
+            "function_definition" => {
+                if let Some(body) = node.child_by_field_name("body") {
+                    self.emit_function_definition(node, body);
+                    self.visit(body);
+                } else if let Some(declarator) = find_function_declarator(node) {
+                    // `= default;` / `= delete;` members have no body node
+                    let start = line_start(self.content, node.start_byte());
+                    if let Some(component) =
+                        self.method_form_component(start, node, declarator.end_byte())
+                    {
+                        self.components.push(component);
+                    }
+                }
+            }
+            "ERROR" => {
+                self.salvage_error_region(node);
+            }
+            "struct_specifier" | "class_specifier" | "union_specifier" => {
+                self.emit_record(node, None);
+            }
+            "enum_specifier" => {
+                self.emit_enum(node);
+            }
+            "type_definition" => {
+                self.emit_type_definition(node);
+            }
+            "declaration" => {
+                self.emit_declaration(node);
+            }
+            "field_declaration" => {
+                self.emit_field(node);
+            }
+            "access_specifier" => {
+                // includes the trailing ':' in the slice
+                let start = line_start(self.content, node.start_byte());
+                let mut end = node.end_byte();
+                if self.content[end..].starts_with(':') {
+                    end += 1;
+                }
+                self.components.push(self.slice_clean(start, end));
+            }
+            "preproc_def" | "preproc_function_def" => {
+                // legacy: `#define` + same-line invocation (greedy to the
+                // last `)` on the first line, excluding `\` continuations)
+                static DEFINE_RE: LazyLock<Regex> =
+                    LazyLock::new(|| Regex::new(r"^#define(\s\w+( ?\w* ?\(.*\))?)?").unwrap());
+                let first_line = self.content[node.byte_range()]
+                    .split('\n')
+                    .next()
+                    .unwrap_or("");
+                if let Some(m) = DEFINE_RE.find(first_line) {
+                    self.components.push(m.as_str().trim_end().to_string());
+                }
+            }
+            "template_declaration" => {
+                self.emit_template(node);
+            }
+            "expression_statement" => {
+                // legacy pybind noise: ` *\w+.def("...` captured to first `;`
+                self.maybe_emit_pybind_def(node);
+                // legacy method-pattern noise: indented `asm(...)` statements
+                if let Some(child) = node.named_child(0) {
+                    if child.kind().contains("asm") {
+                        let start = line_start(self.content, node.start_byte());
+                        let indent = &self.content[start..node.start_byte()];
+                        if !indent.is_empty() && indent.chars().all(char::is_whitespace) {
+                            let text = self.slice_clean(start, node.end_byte());
+                            self.components
+                                .push(text.trim_end_matches(';').trim_end().to_string());
+                        }
+                    }
+                }
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    self.visit(child);
+                }
+            }
+            "while_statement" | "for_statement" | "if_statement" | "switch_statement" => {
+                self.maybe_emit_control_noise(node);
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    self.visit(child);
+                }
+            }
+            _ => {
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    self.visit(child);
+                }
+            }
+        }
+    }
+
+    /// struct/class/union with a body: label + fields (+ optional closer).
+    /// `prefix_start` overrides the slice start (e.g. `typedef `/`static `).
+    fn emit_record(&mut self, node: Node<'a>, prefix_start: Option<usize>) {
+        let Some(body) = node.child_by_field_name("body") else {
+            return; // bare type references are not components
+        };
+        let start = prefix_start.unwrap_or_else(|| line_start(self.content, node.start_byte()));
+        let label = self.slice_clean(start, body.start_byte());
+        if !label.is_empty() {
+            self.components.push(label);
+        }
+        self.visit(body);
+    }
+
+    fn emit_enum(&mut self, node: Node<'a>) {
+        let Some(body) = node.child_by_field_name("body") else {
+            return;
+        };
+        let start = line_start(self.content, node.start_byte());
+        let label = self.slice_clean(start, body.start_byte());
+        if !label.is_empty() {
+            self.components.push(label);
+        }
+        let mut cursor = body.walk();
+        for child in body.named_children(&mut cursor) {
+            if child.kind() == "enumerator" {
+                let estart = line_start(self.content, child.start_byte());
+                let mut eend = child.end_byte();
+                if self.content[eend..].starts_with(',') {
+                    eend += 1;
+                }
+                let text = strip_c_comments(&self.content[estart..eend])
+                    .trim_end()
+                    .trim_start_matches('\n')
+                    .replace('\t', "  ");
+                self.components.push(text);
+            }
+        }
+    }
+
+    /// `typedef struct {...} Name;` -> "typedef struct", fields, "} Name;".
+    fn emit_type_definition(&mut self, node: Node<'a>) {
+        let Some(type_node) = node.child_by_field_name("type") else {
+            return;
+        };
+        match type_node.kind() {
+            "struct_specifier" | "union_specifier" | "class_specifier" => {
+                if type_node.child_by_field_name("body").is_some() {
+                    self.emit_record(type_node, Some(node.start_byte()));
+                    self.emit_closer(type_node, node);
+                }
+            }
+            "enum_specifier" => {
+                if type_node.child_by_field_name("body").is_some() {
+                    self.emit_enum(type_node);
+                    self.emit_closer(type_node, node);
+                }
+            }
+            _ => {}
+        }
+    }
+
+    /// `static struct config {...} config;` and similar declarations.
+    fn emit_declaration(&mut self, node: Node<'a>) {
+        let Some(type_node) = node.child_by_field_name("type") else {
+            return;
+        };
+        // legacy `other_static`: `^static (struct )?TYPE name([])?(?= =)`
+        if type_node.child_by_field_name("body").is_none() {
+            static OTHER_STATIC_RE: LazyLock<Regex> =
+                LazyLock::new(|| Regex::new(r"^static (?:struct )?\w+ \w+(?:\[\])?").unwrap());
+            let text = &self.content[node.byte_range()];
+            if self.content[..node.start_byte()].ends_with('\n') {
+                if let Some(m) = OTHER_STATIC_RE.find(text) {
+                    if text[m.end()..].starts_with(" =") {
+                        self.components.push(m.as_str().to_string());
+                        return;
+                    }
+                }
+            }
+        }
+        match type_node.kind() {
+            "struct_specifier" | "union_specifier" | "class_specifier" => {
+                if type_node.child_by_field_name("body").is_some() {
+                    let start = line_start(self.content, node.start_byte());
+                    self.emit_record(type_node, Some(start));
+                    self.emit_closer(type_node, node);
+                }
+            }
+            "enum_specifier" => {
+                if type_node.child_by_field_name("body").is_some() {
+                    self.emit_enum(type_node);
+                    self.emit_closer(type_node, node);
+                }
+            }
+            _ => {}
+        }
+    }
+
+    /// The legacy `} name;` closing component for named record declarations.
+    fn emit_closer(&mut self, type_node: Node<'a>, decl: Node<'a>) {
+        let Some(body) = type_node.child_by_field_name("body") else {
+            return;
+        };
+        // only when a declarator follows the body on the closing line
+        if decl.end_byte() <= body.end_byte() {
+            return;
+        }
+        let closer_text = self.content[body.end_byte() - 1..decl.end_byte()].trim_end();
+        if closer_text.len() <= 2 {
+            return; // "};" alone -> legacy emitted nothing
+        }
+        let start = line_start(self.content, body.end_byte() - 1);
+        self.components
+            .push(self.slice_clean(start, decl.end_byte()));
+    }
+
+    /// Function definitions must satisfy one of the two legacy forms.
+    fn emit_function_definition(&mut self, node: Node<'a>, body: Node<'a>) {
+        let start = line_start(self.content, node.start_byte());
+        let candidate = self.slice_clean(start, body.start_byte());
+        if candidate.is_empty() {
+            return;
+        }
+        // legacy pybind11 module pattern
+        if candidate.starts_with("PYBIND11_MODULE") {
+            self.components.push(candidate);
+            return;
+        }
+        // function form: the `{` follows the `)` after exactly one whitespace
+        if FUNCTION_FORM_RE.is_match(&candidate)
+            && single_whitespace_before(self.content, body.start_byte())
+        {
+            self.components.push(candidate);
+            return;
+        }
+        // method form: indented members (with const/override/initializers)
+        if let Some(component) = self.method_form_component(start, node, body.start_byte()) {
+            self.components.push(component);
+        }
+    }
+
+    /// Validate the legacy method form; returns the trimmed component.
+    fn method_form_component(&self, start: usize, node: Node<'a>, end: usize) -> Option<String> {
+        // find the parameter list of the declarator
+        let declarator = find_function_declarator(node)?;
+        let params = declarator.child_by_field_name("parameters")?;
+        let prefix = strip_c_comments(&self.content[start..params.start_byte()]);
+        if !METHOD_PREFIX_RE.is_match(&prefix) {
+            return None;
+        }
+        // legacy args `\(.*\)` were single-line
+        if self.content[params.byte_range()].contains('\n') {
+            return None;
+        }
+        let text = self.slice_clean(start, end);
+        Some(text.trim_end_matches(';').trim_end().to_string())
+    }
+
+    fn emit_field(&mut self, node: Node<'a>) {
+        // legacy field patterns require an indented member starting its line
+        let fstart = line_start(self.content, node.start_byte());
+        let indent = &self.content[fstart..node.start_byte()];
+        if indent.is_empty() || !indent.chars().all(char::is_whitespace) {
+            return;
+        }
+        // nested records/enums inside the field (anonymous struct members)
+        if let Some(type_node) = node.child_by_field_name("type") {
+            if matches!(
+                type_node.kind(),
+                "struct_specifier" | "union_specifier" | "enum_specifier" | "class_specifier"
+            ) && type_node.child_by_field_name("body").is_some()
+            {
+                self.visit(type_node);
+                self.emit_closer(type_node, node); // e.g. `} inner;`
+                return;
+            }
+        }
+        if node_has_function_declarator(node) {
+            // member function declarations follow the legacy method form;
+            // trailing `= 0` / `= default` / `= delete` were never captured
+            if let Some(declarator) = find_function_declarator(node) {
+                if let Some(component) =
+                    self.method_form_component(fstart, node, declarator.end_byte())
+                {
+                    self.components.push(component);
+                }
+            }
+            return;
+        }
+        if self.suppress_plain_fields > 0 {
+            return;
+        }
+        let start = fstart;
+        let text = strip_c_comments(&self.content[start..node.end_byte()])
+            .trim_end()
+            .trim_start_matches('\n')
+            .replace('\t', "  ");
+        let trimmed = text.trim_end_matches(' ').to_string();
+        // legacy fields had to terminate with `;` (struct) or `,` (enum)
+        if !(trimmed.ends_with(';') || trimmed.ends_with(',')) {
+            return;
+        }
+        self.components.push(trimmed);
+    }
+
+    /// Partial recovery inside ERROR regions: invalid syntax must still
+    /// yield the obvious components instead of nothing.
+    fn salvage_error_region(&mut self, node: Node<'a>) {
+        static RECORD_HEADER_RE: LazyLock<Regex> =
+            LazyLock::new(|| Regex::new(r"^[ \t]*((?:class|struct) \w+[^\n{;]*)").unwrap());
+        let text = &self.content[node.byte_range()];
+        if let Some(caps) = RECORD_HEADER_RE.captures(text) {
+            if let Some(m) = caps.get(1) {
+                self.components.push(m.as_str().trim_end().to_string());
+            }
+        }
+        // ordered pass: salvage loose function declarators, recurse the rest
+        let mut cursor = node.walk();
+        for child in node.named_children(&mut cursor) {
+            if child.kind() == "function_declarator" {
+                let start = line_start(self.content, child.start_byte());
+                let candidate = self.slice_clean(start, child.end_byte());
+                if FUNCTION_FORM_RE.is_match(&candidate) {
+                    let after = &self.content[child.end_byte()..];
+                    let mut chars = after.chars();
+                    if chars.next().is_some_and(char::is_whitespace) && chars.next() == Some('{') {
+                        self.components.push(candidate);
+                    }
+                }
+            } else {
+                self.visit(child);
+            }
+        }
+    }
+
+    /// Legacy pybind pattern: `^ *\w+\.def\("` ... up to the first `;`.
+    fn maybe_emit_pybind_def(&mut self, stmt: Node<'a>) {
+        let start = line_start(self.content, stmt.start_byte());
+        let region = &self.content[start..];
+        static DEF_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r#"^ *\w+\.def\(""#).unwrap());
+        if !DEF_RE.is_match(region) {
+            return;
+        }
+        let Some(semi) = region.find(';') else { return };
+        let text = strip_c_comments(&region[..semi])
+            .trim_end()
+            .trim_start_matches('\n')
+            .to_string();
+        self.components.push(text);
+    }
+
+    /// Legacy method-pattern noise: an indented `keyword(...)` with no space
+    /// before the paren (e.g. `while((ln = listNext(&li)))`) was emitted.
+    fn maybe_emit_control_noise(&mut self, node: Node<'a>) {
+        let Some(cond) = node
+            .child_by_field_name("condition")
+            .or_else(|| node.child_by_field_name("initializer"))
+        else {
+            return;
+        };
+        let kw = node.start_byte();
+        let start = line_start(self.content, kw);
+        let indent = &self.content[start..kw];
+        if indent.is_empty() || !indent.chars().all(|c| c == ' ' || c == '\t') {
+            return;
+        }
+        // the legacy pattern required no whitespace between keyword and `(`
+        let kw_end = kw + node.kind().split('_').next().unwrap_or("").len();
+        if !self.content[kw_end..].starts_with('(') {
+            return;
+        }
+        if self.content[cond.byte_range()].contains('\n') {
+            return;
+        }
+        // fin: a following `\s{` (compound body) -- the common legacy case
+        let after = &self.content[cond.end_byte()..];
+        if !(after.starts_with(" {") || after.starts_with("\n")) {
+            return;
+        }
+        self.components
+            .push(self.slice_clean(start, cond.end_byte()).replace('\t', "  "));
+    }
+
+    fn emit_template(&mut self, node: Node<'a>) {
+        // slice from `template` through the inner declaration's signature
+        let start = line_start(self.content, node.start_byte());
+        let mut cursor = node.walk();
+        for child in node.named_children(&mut cursor) {
+            match child.kind() {
+                "function_definition" => {
+                    if let Some(body) = child.child_by_field_name("body") {
+                        let text = self.slice_clean(start, body.start_byte());
+                        self.components.push(text);
+                        self.visit(body);
+                    }
+                    return;
+                }
+                "declaration" => {
+                    // prototype: `template <typename T> T cos(T);`
+                    // (`= delete` etc. were never captured by the legacy regex)
+                    let end = find_function_declarator(child)
+                        .map(|d| d.end_byte())
+                        .unwrap_or_else(|| child.end_byte());
+                    let text = self.slice_clean(start, end);
+                    self.components
+                        .push(text.trim_end_matches(';').trim_end().to_string());
+                    return;
+                }
+                "struct_specifier" | "class_specifier" => {
+                    if let Some(body) = child.child_by_field_name("body") {
+                        let text = self.slice_clean(start, body.start_byte());
+                        self.components.push(text);
+                        self.suppress_plain_fields += 1;
+                        self.visit(body);
+                        self.suppress_plain_fields -= 1;
+                    } else {
+                        let text = self.slice_clean(start, child.end_byte());
+                        self.components.push(text);
+                    }
+                    return;
+                }
+                "alias_declaration" | "type_definition" => {
+                    let text = self.slice_clean(start, child.end_byte());
+                    self.components
+                        .push(text.trim_end_matches(';').trim_end().to_string());
+                    return;
+                }
+                _ => {}
+            }
+        }
+    }
+}
+
+fn node_has_function_declarator(node: Node<'_>) -> bool {
+    find_function_declarator(node).is_some()
+}
+
+/// Walk the declarator chain to the `function_declarator`, if any.
+fn find_function_declarator(node: Node<'_>) -> Option<Node<'_>> {
+    let mut current = node.child_by_field_name("declarator");
+    while let Some(d) = current {
+        match d.kind() {
+            "function_declarator" => return Some(d),
+            _ => current = d.child_by_field_name("declarator"),
+        }
+    }
+    None
+}
+
+/// Legacy `(?=\s\{)`: exactly one whitespace char before the brace.
+fn single_whitespace_before(content: &str, brace: usize) -> bool {
+    let before = &content[..brace];
+    before
+        .chars()
+        .next_back()
+        .is_some_and(|c| c.is_whitespace())
+        && !before.ends_with("  ")
+        && !before.ends_with("\n\n")
+        && !before.ends_with(" \n")
+        && !before.ends_with("\n ")
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn c_struct_and_functions() {
+        let content = "struct Point {\n    int x;\n    int y;\n};\n\nstruct Point getOrigin() {\n    struct Point o;\n    return o;\n}\n";
+        assert_eq!(
+            extract(content, ".c").unwrap(),
+            vec![
+                "struct Point",
+                "    int x;",
+                "    int y;",
+                "struct Point getOrigin()"
+            ]
+        );
+    }
+
+    #[test]
+    fn typedef_struct_with_closer() {
+        let content = "typedef struct {\n    char name[50];\n} Person;\n";
+        assert_eq!(
+            extract(content, ".c").unwrap(),
+            vec!["typedef struct", "    char name[50];", "} Person;"]
+        );
+    }
+
+    #[test]
+    fn cpp_class_with_access() {
+        let content = "class Animal {\npublic:\n    Animal(const std::string &name) : name(name) {}\n    virtual void speak() const {}\nprotected:\n    std::string name;\n};\n";
+        assert_eq!(
+            extract(content, ".cpp").unwrap(),
+            vec![
+                "class Animal",
+                "public:",
+                "    Animal(const std::string &name) : name(name)",
+                "    virtual void speak() const",
+                "protected:",
+                "    std::string name;",
+            ]
+        );
+    }
+
+    #[test]
+    fn enums_with_commas() {
+        let content = "enum days {\n  SUN,\n  MON,\n  SAT\n};\n";
+        assert_eq!(
+            extract(content, ".c").unwrap(),
+            vec!["enum days", "  SUN,", "  MON,", "  SAT"]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/treesitter/mod.rs b/crates/tree_plus_core/src/extract/treesitter/mod.rs
new file mode 100644
index 0000000..652ad05
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/treesitter/mod.rs
@@ -0,0 +1,66 @@
+//! Tree-sitter based extractors for the big version-1 languages:
+//! Rust, Python, JavaScript/TypeScript, C and C++.
+//!
+//! Formatters emit source-text signature slices so output matches the legacy
+//! regex-based components; golden parity tests in tests/ enforce this.
+
+pub mod c_cpp;
+pub mod python;
+pub mod rust;
+pub mod typescript;
+
+use std::cell::RefCell;
+
+use tree_sitter::{Language, Node, Parser, Tree};
+
+use super::{ExtractError, ExtractResult};
+
+thread_local! {
+    /// One parser per thread (tree-sitter parser state is mutable).
+    static PARSER: RefCell<Parser> = RefCell::new(Parser::new());
+}
+
+/// Parse `content` with `language`, reusing a thread-local parser.
+pub fn parse(content: &str, language: &Language) -> Result<Tree, ExtractError> {
+    PARSER.with(|cell| {
+        let mut parser = cell.borrow_mut();
+        parser
+            .set_language(language)
+            .map_err(|e| ExtractError::Parse(e.to_string()))?;
+        parser
+            .parse(content, None)
+            .ok_or_else(|| ExtractError::Parse("tree-sitter returned no tree".to_string()))
+    })
+}
+
+/// UTF-8 text of a node (lossy-safe since content is valid UTF-8).
+pub fn node_text<'a>(node: Node<'_>, content: &'a str) -> &'a str {
+    &content[node.byte_range()]
+}
+
+/// Walk all nodes depth-first, calling `visit` on each.
+pub fn walk_tree<'t, F: FnMut(Node<'t>)>(tree: &'t Tree, mut visit: F) {
+    let mut cursor = tree.walk();
+    let mut reached_root = false;
+    while !reached_root {
+        visit(cursor.node());
+        if cursor.goto_first_child() {
+            continue;
+        }
+        loop {
+            if cursor.goto_next_sibling() {
+                break;
+            }
+            if !cursor.goto_parent() {
+                reached_root = true;
+                break;
+            }
+        }
+    }
+}
+
+/// Used by extractors that are still stubs.
+#[allow(dead_code)]
+pub fn unimplemented_language() -> ExtractResult {
+    Ok(Vec::new())
+}
diff --git a/crates/tree_plus_core/src/extract/treesitter/python.rs b/crates/tree_plus_core/src/extract/treesitter/python.rs
new file mode 100644
index 0000000..57c39fe
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/treesitter/python.rs
@@ -0,0 +1,346 @@
+//! Python component extraction (tree-sitter), matching legacy `parse_py`.
+//!
+//! The legacy extractor was a single multi-pattern regex over
+//! comment-and-docstring-stripped source. The formatter here reproduces its
+//! observable quirks deliberately, e.g.:
+//! - `async def` is never emitted (the legacy pattern required `^ *def`);
+//! - decorators are carried until the next matching def/class and dropped
+//!   only when consumed;
+//! - ALL-CAPS assignments ("enum variants") match anywhere, with int, double
+//!   quoted string, or multiline constructor values;
+//! - annotated fields are emitted only while the most recent component was a
+//!   class ("class context"), which persists past the class body.
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+use tree_sitter::Node;
+
+use super::{node_text, parse, ExtractResult};
+
+static COMMENT_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"\s*#.*\n").unwrap());
+static DECORATOR_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"^@\w+(\(.*\))?$").unwrap());
+static RETURN_TYPE_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r#"^[\w"'\[\],. ]+$"#).unwrap());
+static SUPERCLASSES_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^\([\w\[\]\s,=.]*\)$").unwrap());
+static VERSION_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r#"^__version__ = ".*""#).unwrap());
+static TYPEVAR_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^\w+ = TypeVar\([^)]+\)").unwrap());
+static INT_VALUE_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"^\d+").unwrap());
+static STR_VALUE_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r#"^"+[^"]+"+"#).unwrap());
+static STRUCT_CLOSING_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"(?m)^\s*\)$").unwrap());
+static FIELD_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^\s+\w+:\s+[\w\[\]|., ]+").unwrap());
+static ALLCAPS_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r"^[A-Z_\d]+$").unwrap());
+
+/// Remove `# ...` comments the way the legacy extractor did.
+fn strip_comments(slice: &str) -> String {
+    COMMENT_RE.replace_all(slice, "\n").into_owned()
+}
+
+struct PyExtractor<'a> {
+    content: &'a str,
+    components: Vec<String>,
+    pending_decorators: Vec<String>,
+    in_class: bool,
+}
+
+/// Extract Python components: defs, classes, decorators, TypeVars,
+/// `__version__`, enum variants, dataclass fields.
+pub fn extract(content: &str) -> ExtractResult {
+    let tree = parse(content, &tree_sitter_python::LANGUAGE.into())?;
+    let mut extractor = PyExtractor {
+        content,
+        components: Vec::new(),
+        pending_decorators: Vec::new(),
+        in_class: false,
+    };
+    extractor.visit(tree.root_node());
+    Ok(extractor.components)
+}
+
+impl<'a> PyExtractor<'a> {
+    /// Byte offset of the start of the run of spaces before `start`
+    /// (legacy patterns matched ` *` indentation, spaces only).
+    fn space_indent_start(&self, start: usize) -> usize {
+        let bytes = self.content.as_bytes();
+        let mut i = start;
+        while i > 0 && bytes[i - 1] == b' ' {
+            i -= 1;
+        }
+        i
+    }
+
+    /// Byte offset of the start of `[ \t]*` indentation before `start`.
+    fn ws_indent_start(&self, start: usize) -> usize {
+        let bytes = self.content.as_bytes();
+        let mut i = start;
+        while i > 0 && (bytes[i - 1] == b' ' || bytes[i - 1] == b'\t') {
+            i -= 1;
+        }
+        i
+    }
+
+    fn is_line_start(&self, offset: usize) -> bool {
+        offset == 0 || self.content.as_bytes()[offset - 1] == b'\n'
+    }
+
+    fn emit_with_decorators(&mut self, component: String) {
+        if self.pending_decorators.is_empty() {
+            self.components.push(component);
+        } else {
+            let mut joined = self.pending_decorators.join("\n");
+            joined.push('\n');
+            joined.push_str(&component);
+            self.pending_decorators.clear();
+            self.components.push(joined);
+        }
+    }
+
+    fn visit(&mut self, node: Node<'a>) {
+        match node.kind() {
+            "decorated_definition" => {
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    if child.kind() == "decorator" {
+                        self.collect_decorator(child);
+                    } else {
+                        self.visit(child);
+                    }
+                }
+            }
+            "function_definition" => self.handle_function(node),
+            "class_definition" => self.handle_class(node),
+            "expression_statement" => {
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    if child.kind() == "assignment" {
+                        self.handle_assignment(child);
+                    }
+                }
+            }
+            // expressions cannot contain statements; recursion elsewhere is safe
+            _ => {
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    self.visit(child);
+                }
+            }
+        }
+    }
+
+    fn collect_decorator(&mut self, node: Node<'a>) {
+        // legacy: ` *@\w+(\(.*\))?\n` -- single line, simple name, no dots
+        let text = node_text(node, self.content);
+        if node.start_position().row != node.end_position().row {
+            return;
+        }
+        // the decorator must end its line (only whitespace/comment was removed)
+        let after = &self.content[node.end_byte()..];
+        let line_rest = after.split('\n').next().unwrap_or("");
+        if !line_rest.trim_start_matches([' ', '\t']).is_empty()
+            && !line_rest.trim_start().starts_with('#')
+        {
+            return;
+        }
+        if !DECORATOR_RE.is_match(text) {
+            return;
+        }
+        let indent_start = self.space_indent_start(node.start_byte());
+        self.pending_decorators
+            .push(self.content[indent_start..node.end_byte()].to_string());
+    }
+
+    fn handle_function(&mut self, node: Node<'a>) {
+        let body = node.child_by_field_name("body");
+        let is_async = node.child(0).map(|c| c.kind() == "async").unwrap_or(false);
+        if !is_async {
+            if let Some(component) = self.build_function_signature(node) {
+                self.emit_with_decorators(component);
+                self.in_class = false;
+            }
+        }
+        if let Some(body) = body {
+            self.visit(body);
+        }
+    }
+
+    fn build_function_signature(&self, node: Node<'a>) -> Option<String> {
+        let params = node.child_by_field_name("parameters")?;
+        // legacy params pattern allowed nesting depth <= 2 inside the parens
+        let params_text = node_text(params, self.content);
+        let mut depth: i32 = 0;
+        let mut max_inner: i32 = 0;
+        for ch in params_text.chars() {
+            match ch {
+                '(' => {
+                    depth += 1;
+                    max_inner = max_inner.max(depth - 1);
+                }
+                ')' => depth -= 1,
+                _ => {}
+            }
+        }
+        if max_inner > 2 {
+            return None;
+        }
+        let end = match node.child_by_field_name("return_type") {
+            Some(ret) => {
+                let ret_text = node_text(ret, self.content);
+                if !RETURN_TYPE_RE.is_match(&strip_comments(ret_text)) {
+                    return None; // legacy pattern would fail entirely
+                }
+                ret.end_byte()
+            }
+            None => params.end_byte(),
+        };
+        // find the `def` keyword start (node start may equal it)
+        let def_kw = node
+            .children(&mut node.walk())
+            .find(|c| c.kind() == "def")
+            .map(|c| c.start_byte())
+            .unwrap_or(node.start_byte());
+        let start = self.space_indent_start(def_kw);
+        let slice = &self.content[start..end];
+        let cleaned = strip_comments(slice);
+        Some(cleaned)
+    }
+
+    fn handle_class(&mut self, node: Node<'a>) {
+        let body = node.child_by_field_name("body");
+        if let Some(component) = self.build_class_signature(node) {
+            self.emit_with_decorators(component);
+            self.in_class = true;
+        }
+        if let Some(body) = body {
+            self.visit(body);
+        }
+    }
+
+    fn build_class_signature(&self, node: Node<'a>) -> Option<String> {
+        let name = node.child_by_field_name("name")?;
+        let mut end = name.end_byte();
+        if let Some(tp) = node.child_by_field_name("type_parameters") {
+            end = tp.end_byte();
+        }
+        if let Some(supers) = node.child_by_field_name("superclasses") {
+            let text = node_text(supers, self.content);
+            if !SUPERCLASSES_RE.is_match(text) {
+                return None; // legacy charset would fail to match
+            }
+            end = supers.end_byte();
+        }
+        let start = self.space_indent_start(node.start_byte());
+        Some(self.content[start..end].to_string())
+    }
+
+    fn handle_assignment(&mut self, node: Node<'a>) {
+        let Some(left) = node.child_by_field_name("left") else {
+            return;
+        };
+        if left.kind() != "identifier" {
+            return;
+        }
+        let left_text = node_text(left, self.content);
+        let stmt_start = left.start_byte();
+        let indent_start = self.ws_indent_start(stmt_start);
+        let at_col0 = indent_start == stmt_start && self.is_line_start(stmt_start);
+        let line_region = &self.content[stmt_start..];
+
+        if at_col0 {
+            // legacy version/typevar patterns require a preceding newline;
+            // a statement at byte 0 of the file can never match
+            if stmt_start == 0 {
+                return;
+            }
+            if left_text == "__version__" {
+                if let Some(m) = VERSION_RE.find(line_region.split('\n').next().unwrap_or("")) {
+                    self.components.push(m.as_str().to_string());
+                    self.in_class = false;
+                }
+            } else if let Some(m) = TYPEVAR_RE.find(line_region) {
+                self.components.push(m.as_str().to_string());
+                self.in_class = false;
+            }
+            return;
+        }
+
+        // indented assignments: enum variants first (legacy alternation order)
+        let type_annotation = node.child_by_field_name("type");
+        let right = node.child_by_field_name("right");
+        if ALLCAPS_RE.is_match(left_text) && type_annotation.is_none() {
+            if let Some(right) = right {
+                let value_start = right.start_byte();
+                let value_text = &self.content[value_start..];
+                // only `NAME = value` with whitespace-separated `=` matches
+                let between = &self.content[left.end_byte()..value_start];
+                let eq_ok = {
+                    let trimmed = between.trim();
+                    trimmed == "="
+                        && between.starts_with(char::is_whitespace)
+                        && between.ends_with(char::is_whitespace)
+                };
+                if eq_ok {
+                    if let Some(m) = INT_VALUE_RE.find(value_text) {
+                        let end = value_start + m.end();
+                        self.components
+                            .push(self.content[indent_start..end].to_string());
+                        return;
+                    }
+                    if let Some(m) = STR_VALUE_RE.find(value_text) {
+                        let end = value_start + m.end();
+                        self.components
+                            .push(self.content[indent_start..end].to_string());
+                        return;
+                    }
+                    // multiline constructor value: [A-Z]\w*( ... ^\s+\)$
+                    if right.kind() == "call" {
+                        let func_ok = right
+                            .child_by_field_name("function")
+                            .map(|f| {
+                                f.kind() == "identifier"
+                                    && node_text(f, self.content)
+                                        .starts_with(|c: char| c.is_ascii_uppercase())
+                            })
+                            .unwrap_or(false);
+                        if func_ok {
+                            let call_region = &self.content[value_start..right.end_byte()];
+                            if let Some(m) = STRUCT_CLOSING_RE.find(call_region) {
+                                let end = value_start + m.end();
+                                self.components
+                                    .push(self.content[indent_start..end].to_string());
+                                return;
+                            }
+                        }
+                    }
+                }
+            }
+        }
+
+        // dataclass fields only in class context, with a type annotation
+        if self.in_class && type_annotation.is_some() {
+            let line_start = self.ws_indent_start(stmt_start);
+            if !self.is_line_start(line_start) {
+                return;
+            }
+            let region = &self.content[line_start..];
+            let with_indent = format!(
+                "{}{}",
+                &self.content[line_start..stmt_start],
+                region[stmt_start - line_start..]
+                    .split('\n')
+                    .next()
+                    .unwrap_or("")
+            );
+            // legacy: `^\s+\w+:\s+[\w\[\]\|\., ]+`
+            let padded = format!(" {with_indent}"); // \s+ needs >= 1 char
+            if let Some(m) = FIELD_RE.find(&padded) {
+                let captured = &padded[1..m.end()];
+                let component = captured.trim_start_matches('\n').trim_end_matches(' ');
+                self.components.push(component.to_string());
+            }
+        }
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/treesitter/rust.rs b/crates/tree_plus_core/src/extract/treesitter/rust.rs
new file mode 100644
index 0000000..3befc1d
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/treesitter/rust.rs
@@ -0,0 +1,312 @@
+//! Rust component extraction (tree-sitter), matching legacy `parse_rs`.
+//!
+//! Tree-sitter locates items (functions, structs, impls, enums, traits,
+//! mods, macros); each item is then formatted by a direct port of the legacy
+//! per-construct regex applied to the comment-stripped item slice, so the
+//! emitted labels keep every legacy quirk (e.g. indented `impl` blocks are
+//! skipped, enums require trailing commas on all variants, `pub(crate)`
+//! visibility never matches).
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+use tree_sitter::Node;
+
+use super::{parse, ExtractResult};
+
+/// Port of legacy `remove_c_comments` (note: `/* */` only on one line).
+static C_COMMENT_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"(?m)(\s)*//.*(\s*,\s*|\s*\))?$|\s*/\*.*?\*/").unwrap());
+
+pub fn strip_c_comments(content: &str) -> String {
+    C_COMMENT_RE.replace_all(content, "").into_owned()
+}
+
+/// Prefix of the legacy function pattern, up to the opening paren. The
+/// argument scan (with its `\)(?!:)` lookahead) is emulated manually in
+/// `format_fn` because the regex crate has no look-around.
+static FN_PREFIX_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"^\s*(pub\s+?)?((?:async|const)\s+)?fn\s+\w+(<[^>]*?>)?\(").unwrap()
+});
+
+static FN_END_RE: LazyLock<Regex> = LazyLock::new(|| Regex::new(r";\s|\{").unwrap());
+
+/// Legacy fn-arguments character set: `[&\w,.':\[\]()<>${}/\s]`.
+fn fn_args_char_ok(c: char) -> bool {
+    c.is_alphanumeric()
+        || c == '_'
+        || c.is_whitespace()
+        || matches!(
+            c,
+            '&' | ','
+                | '.'
+                | '\''
+                | ':'
+                | '['
+                | ']'
+                | '('
+                | ')'
+                | '<'
+                | '>'
+                | '$'
+                | '{'
+                | '}'
+                | '/'
+        )
+}
+
+static STRUCT_IMPL_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"\n(?P<struct_impl>(?: *((?:pub\s+)?struct)|impl)[^\{;]*?) ?[\{;]").unwrap()
+});
+
+static ENUM_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(
+        r"(?m)^\s*(?P<enum>(?P<visibility>pub\s+?)?enum (?P<enum_name>\w+)(?P<maybe_generics><[^>]*?>)? \{(?P<variant>(?P<maybe_decorator>\s+#\[.*\]+?)?\s+(?P<variant_name>\w+)(?P<maybe_data>(?P<tuple_variant>\([\s\S]+?\))|(?P<struct_variant> \{[^}]+\}))?,)*\s\})\s",
+    )
+    .unwrap()
+});
+
+static TRAIT_MOD_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"\n(?P<trait_mod> *(?:pub\s+)?(trait|mod)\s+\w*(<[^\{]*>)?)").unwrap()
+});
+
+static MACRO_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"\n(?P<macro>(#\[macro_export\]\n)?macro_rules!\s+[a-z_][a-z_0-9]*)").unwrap()
+});
+
+/// Port of rich's `escape` (legacy applied it to enum components containing
+/// attribute "decorators" when syntax highlighting was off).
+static ESCAPE_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"(\\*)(\[[a-z#/@][^\[]*?\])").unwrap());
+
+pub fn rich_escape(markup: &str) -> String {
+    let escaped = ESCAPE_RE.replace_all(markup, |caps: &regex::Captures<'_>| {
+        format!("{0}{0}\\{1}", &caps[1], &caps[2])
+    });
+    let escaped = escaped.into_owned();
+    if escaped.ends_with('\\') && !escaped.ends_with("\\\\") {
+        format!("{escaped}\\")
+    } else {
+        escaped
+    }
+}
+
+/// Extract Rust components: fns, structs, impls, enums (with variants),
+/// traits, mods, macro_rules.
+pub fn extract(content: &str, syntax: bool) -> ExtractResult {
+    let tree = parse(content, &tree_sitter_rust::LANGUAGE.into())?;
+    let mut components: Vec<String> = Vec::new();
+    visit(tree.root_node(), content, syntax, &mut components);
+    Ok(components)
+}
+
+fn visit(node: Node<'_>, content: &str, syntax: bool, out: &mut Vec<String>) {
+    let mut cursor = node.walk();
+    for child in node.named_children(&mut cursor) {
+        match child.kind() {
+            "function_item" | "function_signature_item" => {
+                if let Some(c) = format_fn(child, content) {
+                    out.push(c);
+                }
+                // nested items inside fn bodies still match
+                visit(child, content, syntax, out);
+            }
+            "struct_item" => {
+                if let Some(c) = format_struct_impl(child, content, false) {
+                    out.push(c);
+                }
+            }
+            "impl_item" => {
+                if let Some(c) = format_struct_impl(child, content, true) {
+                    out.push(c);
+                }
+                visit(child, content, syntax, out);
+            }
+            "enum_item" => {
+                if let Some(c) = format_enum(child, content, syntax) {
+                    out.push(c);
+                }
+            }
+            "trait_item" => {
+                if let Some(c) = format_trait_mod(child, content) {
+                    out.push(c);
+                }
+                visit(child, content, syntax, out);
+            }
+            "mod_item" => {
+                if let Some(c) = format_trait_mod(child, content) {
+                    out.push(c);
+                }
+                visit(child, content, syntax, out);
+            }
+            "macro_definition" => {
+                if let Some(c) = format_macro(child, content) {
+                    out.push(c);
+                }
+            }
+            _ => visit(child, content, syntax, out),
+        }
+    }
+}
+
+/// Start of the line containing `byte` (offset just after the previous `\n`).
+fn line_start(content: &str, byte: usize) -> usize {
+    content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
+}
+
+fn format_fn(node: Node<'_>, content: &str) -> Option<String> {
+    node.child_by_field_name("parameters")?;
+    let start = line_start(content, node.start_byte());
+    // include one byte past the node so `;\s` can see the newline
+    let end = (node.end_byte() + 1).min(content.len());
+    let mut slice = strip_c_comments(&content[start..end]);
+    if !slice.ends_with(['\n', ' ', '{']) {
+        slice.push('\n'); // give `;\s` a whitespace to match at EOF
+    }
+    let open = FN_PREFIX_RE.find(&slice)?.end(); // offset just past `(`
+                                                 // emulate `(?P<arguments>charset+?)?\)(?!:)`: accept the first `)` whose
+                                                 // preceding argument text fits the charset and which `:` does not follow
+    let chars: Vec<(usize, char)> = slice[open..].char_indices().collect();
+    let mut args_ok = true;
+    let mut close: Option<usize> = None; // byte offset in `slice` of `)`
+    for (i, (off, c)) in chars.iter().enumerate() {
+        if *c == ')' {
+            let next = chars.get(i + 1).map(|(_, n)| *n);
+            if next != Some(':') {
+                close = Some(open + off);
+                break;
+            }
+            // `)` followed by `:` -> keep scanning, `)` stays in arguments
+        }
+        if !fn_args_char_ok(*c) {
+            args_ok = false;
+            break;
+        }
+    }
+    if !args_ok {
+        return None;
+    }
+    let close = close?;
+    let end_match = FN_END_RE.find(&slice[close + 1..])?;
+    let component = &slice[..close + 1 + end_match.start()];
+    let component = component
+        .trim_end()
+        .trim_end_matches(',')
+        .trim_end_matches('\n')
+        .trim_end_matches(';');
+    Some(component.trim_start_matches('\n').to_string())
+}
+
+fn format_struct_impl(node: Node<'_>, content: &str, is_impl: bool) -> Option<String> {
+    let start = line_start(content, node.start_byte());
+    if start == 0 {
+        return None; // legacy pattern required a preceding newline
+    }
+    if is_impl && node.start_byte() != start {
+        return None; // legacy: impl only matches at column 0
+    }
+    let end = node.end_byte().min(content.len());
+    // prepend the newline the legacy pattern consumed
+    let slice = format!("\n{}", strip_c_comments(&content[start..end]));
+    let caps = STRUCT_IMPL_RE.captures(&slice)?;
+    if caps.get(0)?.start() != 0 {
+        return None;
+    }
+    let component = caps.name("struct_impl")?.as_str().trim_end();
+    Some(component.trim_start_matches('\n').to_string())
+}
+
+fn format_enum(node: Node<'_>, content: &str, syntax: bool) -> Option<String> {
+    // attributes (e.g. #[derive]) come before the enum keyword inside the
+    // node? No: attribute_item is a sibling. The enum slice starts at the
+    // `pub`/`enum` keyword.
+    let start = line_start(content, node.start_byte());
+    let end = (node.end_byte() + 1).min(content.len());
+    let mut slice = strip_c_comments(&content[start..end]);
+    if !slice.ends_with(char::is_whitespace) {
+        slice.push('\n'); // trailing `\s` in the legacy pattern
+    }
+    let caps = ENUM_RE.captures(&slice)?;
+    if caps.get(0)?.start() != 0 {
+        return None;
+    }
+    let component = caps.name("enum")?.as_str().trim_start_matches('\n');
+    let has_decorator = caps.name("maybe_decorator").is_some();
+    if !syntax && has_decorator {
+        Some(rich_escape(component))
+    } else {
+        Some(component.to_string())
+    }
+}
+
+fn format_trait_mod(node: Node<'_>, content: &str) -> Option<String> {
+    let start = line_start(content, node.start_byte());
+    if start == 0 {
+        return None;
+    }
+    let end = node.end_byte().min(content.len());
+    let slice = format!("\n{}", strip_c_comments(&content[start..end]));
+    let caps = TRAIT_MOD_RE.captures(&slice)?;
+    if caps.get(0)?.start() != 0 {
+        return None;
+    }
+    Some(
+        caps.name("trait_mod")?
+            .as_str()
+            .trim_start_matches('\n')
+            .to_string(),
+    )
+}
+
+fn format_macro(node: Node<'_>, content: &str) -> Option<String> {
+    // include a possible #[macro_export] attribute line directly above
+    let mut start = line_start(content, node.start_byte());
+    let prefix = "#[macro_export]\n";
+    if start >= prefix.len() && content[..start].ends_with(prefix) {
+        start -= prefix.len();
+    }
+    if start == 0 {
+        return None;
+    }
+    let end = node.end_byte().min(content.len());
+    let slice = format!("\n{}", strip_c_comments(&content[start..end]));
+    let caps = MACRO_RE.captures(&slice)?;
+    if caps.get(0)?.start() != 0 {
+        return None;
+    }
+    Some(
+        caps.name("macro")?
+            .as_str()
+            .trim_start_matches('\n')
+            .to_string(),
+    )
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn basic_fn() {
+        let content = "fn add(x1: i64, x2: i64) -> i64 {\n    x1 + x2\n}\n";
+        assert_eq!(
+            extract(content, false).unwrap(),
+            vec!["fn add(x1: i64, x2: i64) -> i64"]
+        );
+    }
+
+    #[test]
+    fn impl_and_method() {
+        let content = "struct Point;\nimpl Point {\n    fn get_origin() -> Point {\n        Point\n    }\n}\n";
+        assert_eq!(
+            extract(content, false).unwrap(),
+            vec!["impl Point", "    fn get_origin() -> Point"]
+        );
+    }
+
+    #[test]
+    fn escape_matches_rich() {
+        assert_eq!(rich_escape("#[default]"), "#\\[default]");
+        assert_eq!(rich_escape("[Topic]"), "[Topic]"); // uppercase: untouched
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/treesitter/typescript.rs b/crates/tree_plus_core/src/extract/treesitter/typescript.rs
new file mode 100644
index 0000000..55dd887
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/treesitter/typescript.rs
@@ -0,0 +1,420 @@
+//! JavaScript/TypeScript component extraction (tree-sitter).
+//!
+//! Emits source-text signature slices in legacy `parse_ts` style:
+//! - classes/interfaces with their heritage clauses;
+//! - methods and constructors (signature only, original indentation);
+//! - function declarations and expressions (signature only);
+//! - arrow functions bound to variables or object members, ending at `=>`;
+//! - `const name = {` "object scope" lines when the object holds functions;
+//! - `type Name` aliases.
+//!
+//! Known intentional differences from the legacy regex (documented in
+//! docs/rust-port-differences.md): regex noise like bare call statements
+//! (`super(...)`, `innerFunction("inner")`) and string literals containing
+//! the word "function" are no longer emitted.
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+use tree_sitter::Node;
+
+use super::rust::strip_c_comments;
+use super::{parse, ExtractResult};
+
+/// Legacy `parse_ts` function preamble: what may precede the `function`
+/// keyword on its line for the match to fire.
+static FN_PREAMBLE_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(
+        r"^ *(export (default )?)?(\(|(const|var|let) \w+ = |return |\w+: )?(async )? ?(\w+\()?$",
+    )
+    .unwrap()
+});
+
+/// Legacy jsdoc line patterns (applied to a `/** ... */` comment preceding a
+/// component; tagged lines keep matching, untagged only before any tag).
+static JSDOC_TAGGED_RE: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r"(?m)^ +\* +@(category|typedefn|sig|param|returns?)( \{\{?[\s\S]*?\}?\})?.*")
+        .unwrap()
+});
+static JSDOC_UNTAGGED_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"(?m)^ \* \w+.*").unwrap());
+
+/// Extract JS/TS components.
+pub fn extract(content: &str, tsx: bool) -> ExtractResult {
+    let language = if tsx {
+        tree_sitter_typescript::LANGUAGE_TSX
+    } else {
+        tree_sitter_typescript::LANGUAGE_TYPESCRIPT
+    };
+    let tree = parse(content, &language.into())?;
+    let mut extractor = TsExtractor {
+        content,
+        components: Vec::new(),
+    };
+    extractor.visit(tree.root_node());
+    Ok(extractor.components)
+}
+
+struct TsExtractor<'a> {
+    content: &'a str,
+    components: Vec<String>,
+}
+
+fn line_start(content: &str, byte: usize) -> usize {
+    content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
+}
+
+impl<'a> TsExtractor<'a> {
+    /// Legacy jsdoc carry: a `/** ... */` comment directly above a component
+    /// contributes its tagged (and leading untagged) lines.
+    fn jsdoc_prefix(&self, slice_start: usize) -> Option<String> {
+        let before = self.content[..slice_start].trim_end();
+        if !before.ends_with("*/") {
+            return None;
+        }
+        let open = before.rfind("/**")?;
+        let comment = &before[open..];
+        let mut lines: Vec<&str> = Vec::new();
+        let mut seen_tagged = false;
+        let mut pos = 0;
+        while pos < comment.len() {
+            let tagged = JSDOC_TAGGED_RE.find_at(comment, pos);
+            let untagged = JSDOC_UNTAGGED_RE.find_at(comment, pos);
+            let pick = match (tagged, untagged) {
+                (Some(t), Some(u)) => {
+                    if t.start() <= u.start() {
+                        (t, true)
+                    } else {
+                        (u, false)
+                    }
+                }
+                (Some(t), None) => (t, true),
+                (None, Some(u)) => (u, false),
+                (None, None) => break,
+            };
+            let (m, is_tagged) = pick;
+            if is_tagged {
+                seen_tagged = true;
+                lines.push(m.as_str());
+            } else if !seen_tagged {
+                lines.push(m.as_str());
+            }
+            pos = m.end().max(pos + 1);
+        }
+        if lines.is_empty() {
+            return None;
+        }
+        Some(format!("/**\n{}\n */\n", lines.join("\n")))
+    }
+
+    fn push_with_jsdoc(&mut self, slice_start: usize, component: String) {
+        match self.jsdoc_prefix(slice_start) {
+            Some(prefix) => self.components.push(format!("{prefix}{component}")),
+            None => self.components.push(component),
+        }
+    }
+
+    fn slice(&self, from: usize, to: usize) -> String {
+        // legacy parse_ts ran on comment-stripped content
+        strip_c_comments(&self.content[from..to])
+            .trim_end()
+            .to_string()
+    }
+
+    fn visit(&mut self, node: Node<'a>) {
+        match node.kind() {
+            "class_declaration" | "abstract_class_declaration" => {
+                self.emit_class_like(node);
+                if let Some(body) = node.child_by_field_name("body") {
+                    self.visit(body);
+                }
+            }
+            "interface_declaration" => {
+                self.emit_class_like(node);
+            }
+            "type_alias_declaration" => {
+                if let Some(name) = node.child_by_field_name("name") {
+                    let mut label = format!("type {}", &self.content[name.byte_range()]);
+                    if let Some(tp) = node.child_by_field_name("type_parameters") {
+                        label.push_str(&self.content[tp.byte_range()]);
+                    }
+                    self.components.push(label);
+                }
+            }
+            "function_declaration" | "generator_function_declaration" => {
+                self.emit_function(node);
+                if let Some(body) = node.child_by_field_name("body") {
+                    self.visit(body);
+                }
+            }
+            "function_expression" | "generator_function" => {
+                if self.function_expression_context(node) {
+                    self.emit_function(node);
+                }
+                if let Some(body) = node.child_by_field_name("body") {
+                    self.visit(body);
+                }
+            }
+            "method_definition" | "method_signature" | "abstract_method_signature" => {
+                self.emit_signature_to_params_or_return(node);
+                if let Some(body) = node.child_by_field_name("body") {
+                    self.visit(body);
+                }
+            }
+            "arrow_function" => {
+                if self.arrow_context(node) {
+                    self.emit_arrow(node);
+                }
+                if let Some(body) = node.child_by_field_name("body") {
+                    self.visit(body);
+                }
+            }
+            "expression_statement" => {
+                // legacy "method" noise: indented bare calls like `super(x)`
+                if let Some(call) = node
+                    .named_child(0)
+                    .filter(|c| matches!(c.kind(), "call_expression"))
+                {
+                    self.maybe_emit_bare_call(node, call);
+                }
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    self.visit(child);
+                }
+            }
+            "variable_declarator" => {
+                // object scope: `const name = {` when the object holds functions
+                if let Some(value) = node.child_by_field_name("value") {
+                    if value.kind() == "object" && object_holds_functions(value) {
+                        let start = line_start(self.content, node.start_byte());
+                        // slice through the object's opening brace
+                        let brace = value.start_byte();
+                        self.components
+                            .push(self.content[start..=brace].trim_end().to_string());
+                    }
+                    self.visit(value);
+                }
+            }
+            _ => {
+                let mut cursor = node.walk();
+                for child in node.named_children(&mut cursor) {
+                    self.visit(child);
+                }
+            }
+        }
+    }
+
+    /// class/interface: line start through heritage (extends/implements).
+    fn emit_class_like(&mut self, node: Node<'a>) {
+        let Some(name) = node.child_by_field_name("name") else {
+            return;
+        };
+        let mut end = name.end_byte();
+        if let Some(tp) = node.child_by_field_name("type_parameters") {
+            end = end.max(tp.end_byte());
+        }
+        let mut cursor = node.walk();
+        for child in node.children(&mut cursor) {
+            if matches!(
+                child.kind(),
+                "class_heritage" | "extends_clause" | "implements_clause" | "extends_type_clause"
+            ) {
+                end = end.max(child.end_byte());
+            }
+        }
+        let start = line_start(self.content, outer_start(node));
+        self.components.push(self.slice(start, end));
+    }
+
+    /// Functions: line start through return type or params; legacy stripped
+    /// leading `(` from IIFE matches.
+    fn emit_function(&mut self, node: Node<'a>) {
+        let Some(params) = node.child_by_field_name("parameters") else {
+            return;
+        };
+        let end = node
+            .child_by_field_name("return_type")
+            .map(|r| {
+                // legacy return-type capture was lazy and stopped before " {"
+                let text = &self.content[r.byte_range()];
+                match text.find(" {") {
+                    Some(i) => r.start_byte() + i,
+                    None => r.end_byte(),
+                }
+            })
+            .unwrap_or_else(|| params.end_byte());
+        let start = line_start(self.content, outer_start(node));
+        let text = self.slice(start, end);
+        self.push_with_jsdoc(start, text.trim_start_matches('(').to_string());
+    }
+
+    /// Methods/constructors: indentation + signature through return type.
+    fn emit_signature_to_params_or_return(&mut self, node: Node<'a>) {
+        let Some(params) = node.child_by_field_name("parameters") else {
+            return;
+        };
+        let end = node
+            .child_by_field_name("return_type")
+            .map(|r| r.end_byte())
+            .unwrap_or_else(|| params.end_byte());
+        let start = line_start(self.content, node.start_byte());
+        self.components.push(self.slice(start, end));
+    }
+
+    /// Legacy method-pattern noise: an indented statement `name(args);` (or
+    /// `super(args);`) matched the method signature regex and was emitted.
+    fn maybe_emit_bare_call(&mut self, stmt: Node<'a>, call: Node<'a>) {
+        let Some(callee) = call.child_by_field_name("function") else {
+            return;
+        };
+        if !matches!(callee.kind(), "identifier" | "super") {
+            return;
+        }
+        let Some(args) = call.child_by_field_name("arguments") else {
+            return;
+        };
+        // `\w+\(...` -- no gap between callee and args, single-line args
+        if args.start_byte() != callee.end_byte() {
+            return;
+        }
+        if self.content[args.byte_range()].contains('\n') {
+            return;
+        }
+        // must start an indented line
+        let start = line_start(self.content, stmt.start_byte());
+        let indent = &self.content[start..callee.start_byte()];
+        if indent.is_empty() || !indent.chars().all(|c| c == ' ' || c == '\t') {
+            return;
+        }
+        // fin: `;` or ` {` after the call (legacy pattern requirement)
+        let after = &self.content[call.end_byte()..];
+        if !(after.starts_with(';') || after.starts_with(" {")) {
+            return;
+        }
+        self.components.push(self.slice(start, call.end_byte()));
+    }
+
+    /// Arrows: line start through the `=>` token.
+    fn emit_arrow(&mut self, node: Node<'a>) {
+        let mut arrow_end = None;
+        let mut cursor = node.walk();
+        for child in node.children(&mut cursor) {
+            if child.kind() == "=>" {
+                arrow_end = Some(child.end_byte());
+                break;
+            }
+        }
+        let Some(end) = arrow_end else { return };
+        let start = line_start(self.content, outer_start(node));
+        let text = self.slice(start, end);
+        self.push_with_jsdoc(start, text);
+    }
+
+    /// Legacy gate for function expressions: the text before the `function`
+    /// keyword on its line had to match the legacy preamble pattern
+    /// (optional export/assignment/return/key, optional `name(` wrapper).
+    fn function_expression_context(&self, node: Node<'a>) -> bool {
+        let fn_kw = node.start_byte();
+        let start = line_start(self.content, fn_kw);
+        let prefix = &self.content[start..fn_kw];
+        FN_PREAMBLE_RE.is_match(prefix)
+    }
+
+    fn arrow_context(&self, node: Node<'a>) -> bool {
+        let Some(parent) = node.parent() else {
+            return false;
+        };
+        matches!(
+            parent.kind(),
+            "variable_declarator" | "pair" | "assignment_expression"
+        )
+    }
+}
+
+/// Start byte including an `export ...` wrapper, but excluding decorators
+/// (the legacy extractor never captured them).
+fn outer_start(node: Node<'_>) -> usize {
+    // skip the node's own leading decorators
+    let mut start = node.start_byte();
+    let mut cursor = node.walk();
+    for child in node.children(&mut cursor) {
+        if child.kind() != "decorator" {
+            start = child.start_byte();
+            break;
+        }
+    }
+    if let Some(parent) = node.parent() {
+        if parent.kind() == "export_statement" {
+            // anchor on the `export` keyword itself, past any decorators
+            let mut pcursor = parent.walk();
+            for child in parent.children(&mut pcursor) {
+                if child.kind() == "export" {
+                    return child.start_byte();
+                }
+            }
+        }
+    }
+    start
+}
+
+/// Whether an object literal directly contains function/arrow members.
+fn object_holds_functions(object: Node<'_>) -> bool {
+    let mut cursor = object.walk();
+    for pair in object.named_children(&mut cursor) {
+        if pair.kind() == "pair" {
+            if let Some(value) = pair.child_by_field_name("value") {
+                if matches!(
+                    value.kind(),
+                    "function_expression" | "arrow_function" | "generator_function"
+                ) {
+                    return true;
+                }
+            }
+        }
+        if pair.kind() == "method_definition" {
+            return true;
+        }
+    }
+    false
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn class_with_methods() {
+        let content = "export class AppComponent {\n  constructor(private http: HttpClient) {}\n  checkSession() {}\n  async goToEvent(event_id: string) {}\n}\n";
+        assert_eq!(
+            extract(content, false).unwrap(),
+            vec![
+                "export class AppComponent",
+                "  constructor(private http: HttpClient)",
+                "  checkSession()",
+                "  async goToEvent(event_id: string)",
+            ]
+        );
+    }
+
+    #[test]
+    fn arrows_and_functions() {
+        let content = "const arrow = (a: String, b: Number) => {\n  return a;\n};\nfunction tsFunction() {\n}\n";
+        assert_eq!(
+            extract(content, false).unwrap(),
+            vec![
+                "const arrow = (a: String, b: Number) =>",
+                "function tsFunction()"
+            ]
+        );
+    }
+
+    #[test]
+    fn object_scope() {
+        let content =
+            "const myObject = {\n  myMethod: function (stuff) {\n    return stuff;\n  },\n};\n";
+        assert_eq!(
+            extract(content, false).unwrap(),
+            vec!["const myObject = {", "  myMethod: function (stuff)"]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/ignore.rs b/crates/tree_plus_core/src/ignore.rs
new file mode 100644
index 0000000..d102eba
--- /dev/null
+++ b/crates/tree_plus_core/src/ignore.rs
@@ -0,0 +1,307 @@
+//! Ignore and glob handling, matching legacy `tree_plus_src/ignore.py`.
+//!
+//! Legacy semantics:
+//! - `should_ignore(path, ignore, globs)` returns true when any *component*
+//!   of the normalized path matches any ignore pattern via `fnmatch.fnmatch`,
+//!   or when amortized globs are present and the path is not in the match set.
+//! - `DEFAULT_IGNORE` is a fixed pattern set; user patterns are unioned with
+//!   it unless `override` is set.
+
+use std::collections::HashSet;
+use std::path::{Path, PathBuf};
+
+use regex::Regex;
+
+/// Default ignore patterns (legacy `DEFAULT_IGNORE_FROZENSET`).
+pub const DEFAULT_IGNORE: &[&str] = &[
+    "__init__.py",
+    "__pycache__",
+    "._*",
+    ".angular",
+    ".cache",
+    ".coverage",
+    ".DS_Store",
+    ".flake8",
+    ".git",
+    ".hypothesis",
+    ".idea",
+    ".ipynb_checkpoints",
+    ".pytest_cache",
+    ".rustc_info.json",
+    ".vscode",
+    "*_memmap",
+    "*:Zone.Identifier",
+    "*.a",
+    "*.ai",
+    "*.bak",
+    "*.bin",
+    "*.bz2",
+    "*.chk",
+    "*.class",
+    "*.d",
+    "*.dat",
+    "*.dll",
+    "*.dylib",
+    "*.ear",
+    "*.egg-info",
+    "*.eot",
+    "*.eps",
+    "*.flac",
+    "*.flv",
+    "*.framework",
+    "*.img",
+    "*.ipa",
+    "*.iso",
+    "*.jar",
+    "*.lib",
+    "*.lock",
+    "*.log",
+    "*.nib",
+    "*.node",
+    "*.o",
+    "*.obj",
+    "*.odg",
+    "*.pack",
+    "*.psd",
+    "*.pyc",
+    "*.pyd",
+    "*.pyo",
+    "*.rar",
+    "*.rlib",
+    "*.rmeta",
+    "*.so",
+    "*.storyboardc",
+    "*.swo",
+    "*.swp",
+    "*.tar",
+    "*.tml",
+    "*.ttf",
+    "*.war",
+    "*.woff",
+    "*.xcarchive",
+    "*.zip",
+    "*.zst",
+    "**/target/debug/**",
+    "**/tmp/",
+    "*~",
+    "babel-webpack",
+    "build",
+    "CACHEDIR.TAG",
+    "Cargo.lock",
+    "detritus",
+    "dist",
+    "env",
+    "node_modules",
+    "target",
+    "venv",
+];
+
+/// Translate a Python `fnmatch` pattern to an anchored regex string.
+///
+/// Supports `*`, `?`, `[seq]`, `[!seq]`; everything else is escaped, matching
+/// `fnmatch.translate` semantics closely enough for the ignore pattern set.
+pub fn fnmatch_translate(pattern: &str) -> String {
+    let chars: Vec<char> = pattern.chars().collect();
+    let mut out = String::from("(?s)^");
+    let mut i = 0;
+    while i < chars.len() {
+        let c = chars[i];
+        i += 1;
+        match c {
+            '*' => out.push_str(".*"),
+            '?' => out.push('.'),
+            '[' => {
+                let mut j = i;
+                if j < chars.len() && (chars[j] == '!' || chars[j] == ']') {
+                    j += 1;
+                }
+                while j < chars.len() && chars[j] != ']' {
+                    j += 1;
+                }
+                if j >= chars.len() {
+                    out.push_str("\\[");
+                } else {
+                    let inner: String = chars[i..j].iter().collect();
+                    let inner = inner.replace('\\', "\\\\");
+                    out.push('[');
+                    if let Some(rest) = inner.strip_prefix('!') {
+                        out.push('^');
+                        out.push_str(rest);
+                    } else if inner.starts_with('^') {
+                        out.push('\\');
+                        out.push_str(&inner);
+                    } else {
+                        out.push_str(&inner);
+                    }
+                    out.push(']');
+                    i = j + 1;
+                }
+            }
+            other => {
+                if "\\.+()|{}^$".contains(other) {
+                    out.push('\\');
+                }
+                out.push(other);
+            }
+        }
+    }
+    out.push('$');
+    out
+}
+
+/// A compiled set of ignore patterns.
+#[derive(Debug, Clone)]
+pub struct IgnorePatterns {
+    regexes: Vec<Regex>,
+}
+
+impl IgnorePatterns {
+    pub fn new<I: IntoIterator<Item = S>, S: AsRef<str>>(patterns: I) -> Self {
+        let regexes = patterns
+            .into_iter()
+            .filter_map(|p| Regex::new(&fnmatch_translate(p.as_ref())).ok())
+            .collect();
+        IgnorePatterns { regexes }
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.regexes.is_empty()
+    }
+
+    /// Whether any path component matches any pattern (legacy `should_ignore`).
+    pub fn matches_path(&self, path: &Path) -> bool {
+        if self.regexes.is_empty() {
+            return false;
+        }
+        for part in path.iter() {
+            let Some(part) = part.to_str() else { continue };
+            if self.regexes.iter().any(|r| r.is_match(part)) {
+                return true;
+            }
+        }
+        false
+    }
+}
+
+/// Build the effective ignore pattern list (legacy `parse_ignore`).
+pub fn parse_ignore(user_patterns: &[String], override_ignore: bool) -> Vec<String> {
+    let mut patterns: Vec<String> = user_patterns.to_vec();
+    if !override_ignore {
+        patterns.extend(DEFAULT_IGNORE.iter().map(|s| s.to_string()));
+    }
+    patterns.sort();
+    patterns.dedup();
+    patterns
+}
+
+/// Pre-resolved glob matches (legacy `AmortizedGlobs`): the set of every
+/// matching path plus all ancestor directories.
+#[derive(Debug, Clone)]
+pub struct AmortizedGlobs {
+    pub matches: HashSet<PathBuf>,
+}
+
+impl AmortizedGlobs {
+    pub fn contains(&self, path: &Path) -> bool {
+        self.matches.contains(path)
+    }
+}
+
+/// Recursively find glob matches under `paths` (legacy `amortize_globs`).
+///
+/// Patterns are name globs (like `Path.rglob`), matched against each entry's
+/// file name at any depth. Returns `None` when nothing matches.
+pub fn amortize_globs(paths: &[PathBuf], globs: &[String]) -> Option<AmortizedGlobs> {
+    if paths.is_empty() || globs.is_empty() {
+        return None;
+    }
+    let regexes: Vec<Regex> = globs
+        .iter()
+        .filter_map(|g| Regex::new(&fnmatch_translate(g)).ok())
+        .collect();
+    let mut matches: HashSet<PathBuf> = HashSet::new();
+    fn walk(dir: &Path, regexes: &[Regex], matches: &mut HashSet<PathBuf>) {
+        let Ok(entries) = std::fs::read_dir(dir) else {
+            return;
+        };
+        for entry in entries.flatten() {
+            let p = entry.path();
+            let name_matches = p
+                .file_name()
+                .and_then(|n| n.to_str())
+                .map(|n| regexes.iter().any(|r| r.is_match(n)))
+                .unwrap_or(false);
+            if name_matches && !matches.contains(&p) {
+                matches.insert(p.clone());
+                for parent in p.ancestors().skip(1) {
+                    matches.insert(parent.to_path_buf());
+                }
+            }
+            if p.is_dir() {
+                walk(&p, regexes, matches);
+            }
+        }
+    }
+    for path in paths {
+        walk(path, &regexes, &mut matches);
+    }
+    if matches.is_empty() {
+        return None;
+    }
+    Some(AmortizedGlobs { matches })
+}
+
+/// Combined skip decision (legacy `should_ignore`).
+pub fn should_ignore(path: &Path, ignore: &IgnorePatterns, globs: Option<&AmortizedGlobs>) -> bool {
+    if let Some(globs) = globs {
+        if !globs.contains(path) {
+            return true;
+        }
+    }
+    ignore.matches_path(path)
+}
+
+/// Whether a string looks like a glob (legacy `is_glob`).
+pub fn is_glob(x: &str) -> bool {
+    static GLOB_RE: std::sync::LazyLock<Regex> =
+        std::sync::LazyLock::new(|| Regex::new(r"(\*)?\*|\?|\[(.\-.)+\]").unwrap());
+    GLOB_RE.is_match(x)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn fnmatch_star_and_class() {
+        let p = IgnorePatterns::new(["*.pyc"]);
+        assert!(p.matches_path(Path::new("a/b/c.pyc")));
+        assert!(!p.matches_path(Path::new("a/b/c.py")));
+        let dot = IgnorePatterns::new(["._*"]);
+        assert!(dot.matches_path(Path::new("x/._hidden")));
+        assert!(!dot.matches_path(Path::new("x/h._idden_not_component")));
+    }
+
+    #[test]
+    fn component_matching() {
+        let p = IgnorePatterns::new(["target"]);
+        assert!(p.matches_path(Path::new("crates/x/target/debug/foo")));
+        assert!(!p.matches_path(Path::new("crates/x/targeted/foo")));
+    }
+
+    #[test]
+    fn is_glob_examples() {
+        assert!(is_glob("*.rs"));
+        assert!(is_glob("file?.py"));
+        assert!(!is_glob("plain.txt"));
+    }
+
+    #[test]
+    fn parse_ignore_union() {
+        let with_default = parse_ignore(&["*.java".to_string()], false);
+        assert!(with_default.iter().any(|p| p == "*.java"));
+        assert!(with_default.iter().any(|p| p == ".git"));
+        let only_user = parse_ignore(&["*.java".to_string()], true);
+        assert_eq!(only_user, vec!["*.java".to_string()]);
+    }
+}
diff --git a/crates/tree_plus_core/src/lib.rs b/crates/tree_plus_core/src/lib.rs
new file mode 100644
index 0000000..6b021b9
--- /dev/null
+++ b/crates/tree_plus_core/src/lib.rs
@@ -0,0 +1,29 @@
+//! tree_plus_core: a `tree` util enhanced with tokens, lines, and components.
+//!
+//! Rust port of the Python `tree_plus` package (version-1 scope: local
+//! filesystem mode). See docs/architecture.md for the design and
+//! docs/rust-port-differences.md for intentional differences.
+
+pub mod config;
+pub mod count;
+pub mod extract;
+pub mod ignore;
+pub mod model;
+pub mod render;
+pub mod sort;
+pub mod walk;
+
+pub use config::TreePlusConfig;
+pub use count::{count_tokens_lines, TokenLineCount, TokenizerName};
+pub use extract::extract_components;
+pub use ignore::DEFAULT_IGNORE;
+pub use model::{Category, TreePlus};
+pub use render::{render_to_string, DEFAULT_WIDTH};
+pub use walk::from_seeds;
+
+impl TreePlus {
+    /// Render this tree exactly like the legacy `TreePlus.into_str()`.
+    pub fn into_str(&self) -> String {
+        render::render_to_string(self, render::DEFAULT_WIDTH)
+    }
+}
diff --git a/crates/tree_plus_core/src/model.rs b/crates/tree_plus_core/src/model.rs
new file mode 100644
index 0000000..2e4c10c
--- /dev/null
+++ b/crates/tree_plus_core/src/model.rs
@@ -0,0 +1,125 @@
+//! Core data model: `TreePlus` and `Category`.
+//!
+//! Mirrors the Python `tree_plus_src.engine.TreePlus` dataclass for the
+//! version-1 scope (local filesystem mode: ROOT, GLOB, FOLDER, FILE,
+//! COMPONENT categories; URL/TAG are deferred web-mode categories).
+
+/// Category of a `TreePlus` node.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum Category {
+    Root,
+    Glob,
+    Folder,
+    File,
+    Component,
+}
+
+/// A node in the tree_plus tree.
+///
+/// `components` holds the extracted display labels for FILE nodes
+/// (the legacy Python stored these in `subtrees` as `List[str]`).
+#[derive(Debug, Clone)]
+pub struct TreePlus {
+    pub category: Category,
+    pub name: String,
+    pub line_count: u64,
+    pub token_count: u64,
+    /// Child trees (folders/files/globs under folders or root).
+    pub subtrees: Vec<TreePlus>,
+    /// Extracted component labels (only for FILE nodes).
+    pub components: Vec<String>,
+}
+
+impl TreePlus {
+    pub fn new(category: Category, name: impl Into<String>) -> Self {
+        TreePlus {
+            category,
+            name: name.into(),
+            line_count: 0,
+            token_count: 0,
+            subtrees: Vec::new(),
+            components: Vec::new(),
+        }
+    }
+
+    pub fn is_folder(&self) -> bool {
+        self.category == Category::Folder
+    }
+
+    pub fn is_file(&self) -> bool {
+        self.category == Category::File
+    }
+
+    /// Total folder count including self (legacy `n_folders`).
+    pub fn n_folders(&self) -> u64 {
+        let own = u64::from(self.is_folder());
+        self.subtrees.iter().map(TreePlus::n_folders).sum::<u64>() + own
+    }
+
+    /// Total file count (legacy `n_files`).
+    pub fn n_files(&self) -> u64 {
+        let own = u64::from(self.is_file());
+        self.subtrees.iter().map(TreePlus::n_files).sum::<u64>() + own
+    }
+
+    /// Total line count (legacy `n_lines`).
+    pub fn n_lines(&self) -> u64 {
+        self.line_count + self.subtrees.iter().map(TreePlus::n_lines).sum::<u64>()
+    }
+
+    /// Total token count (legacy `n_tokens`).
+    pub fn n_tokens(&self) -> u64 {
+        self.token_count + self.subtrees.iter().map(TreePlus::n_tokens).sum::<u64>()
+    }
+
+    /// Legacy `stats()` string, e.g. `1 folder(s), 6 file(s), 1,234 line(s), 5,678 token(s)`.
+    pub fn stats(&self) -> String {
+        format!(
+            "{} folder(s), {} file(s), {} line(s), {} token(s)",
+            commafy(self.n_folders()),
+            commafy(self.n_files()),
+            commafy(self.n_lines()),
+            commafy(self.n_tokens()),
+        )
+    }
+}
+
+/// Format an integer with thousands separators like Python's `{:,}`.
+pub fn commafy(n: u64) -> String {
+    let digits = n.to_string();
+    let mut out = String::with_capacity(digits.len() + digits.len() / 3);
+    let offset = digits.len() % 3;
+    for (i, ch) in digits.chars().enumerate() {
+        if i != 0 && (i + 3 - offset).is_multiple_of(3) {
+            out.push(',');
+        }
+        out.push(ch);
+    }
+    out
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn commafy_matches_python_format() {
+        assert_eq!(commafy(0), "0");
+        assert_eq!(commafy(999), "999");
+        assert_eq!(commafy(1000), "1,000");
+        assert_eq!(commafy(1234567), "1,234,567");
+    }
+
+    #[test]
+    fn counts_roll_up() {
+        let mut folder = TreePlus::new(Category::Folder, "f");
+        let mut file = TreePlus::new(Category::File, "a.py");
+        file.line_count = 10;
+        file.token_count = 100;
+        folder.subtrees.push(file);
+        assert_eq!(folder.n_folders(), 1);
+        assert_eq!(folder.n_files(), 1);
+        assert_eq!(folder.n_lines(), 10);
+        assert_eq!(folder.n_tokens(), 100);
+    }
+}
diff --git a/crates/tree_plus_core/src/render.rs b/crates/tree_plus_core/src/render.rs
new file mode 100644
index 0000000..438276b
--- /dev/null
+++ b/crates/tree_plus_core/src/render.rs
@@ -0,0 +1,347 @@
+//! Deterministic tree rendering byte-compatible with the legacy renderer
+//! (rich.tree.Tree printed via a no-color Console at a fixed width).
+//!
+//! Reimplements the relevant parts of rich:
+//! - tree guides: `"    "`, `"│   "`, `"├── "`, `"└── "`;
+//! - word wrapping (`rich._wrap.divide_line` with fold=True) measured in
+//!   terminal cells (emoji are 2 cells wide);
+//! - trailing-space removal (legacy `remove_trailing_space`).
+
+use unicode_width::{UnicodeWidthChar, UnicodeWidthStr};
+
+use crate::model::{commafy, Category, TreePlus};
+
+pub const DEFAULT_WIDTH: usize = 80;
+
+const FILE_CHAR: &str = "📄";
+const FOLDER_CHAR: &str = "📁";
+const ROOT_CHAR: &str = "🌵";
+const GLOB_CHAR: &str = "🌀";
+
+fn plural(n: u64) -> &'static str {
+    if n == 1 {
+        ""
+    } else {
+        "s"
+    }
+}
+
+/// Build the display label for a node (legacy `_into_rich_tree`).
+pub fn node_label(node: &TreePlus) -> String {
+    match node.category {
+        Category::File => {
+            let tokens = node.n_tokens();
+            let lines = node.n_lines();
+            format!(
+                "{FILE_CHAR} {} ({} token{}, {} line{})",
+                node.name,
+                commafy(tokens),
+                plural(tokens),
+                commafy(lines),
+                plural(lines),
+            )
+        }
+        Category::Folder => {
+            let folders = node.n_folders();
+            let files = node.n_files();
+            format!(
+                "{FOLDER_CHAR} {} ({} folder{}, {} file{})",
+                node.name,
+                commafy(folders),
+                plural(folders),
+                commafy(files),
+                plural(files),
+            )
+        }
+        Category::Root => {
+            let mut label = format!("{ROOT_CHAR} {}", node.name);
+            if node.n_tokens() > 0 && node.n_lines() > 0 {
+                let folders = node.n_folders();
+                let files = node.n_files();
+                label.push_str(&format!(
+                    " ({} folder{}, {} file{})",
+                    commafy(folders),
+                    plural(folders),
+                    commafy(files),
+                    plural(files),
+                ));
+            }
+            label
+        }
+        Category::Glob => {
+            let n = node.subtrees.len() as u64;
+            format!(
+                "{GLOB_CHAR} {} ({} match{})",
+                node.name,
+                commafy(n),
+                if n == 1 { "" } else { "es" },
+            )
+        }
+        Category::Component => node.name.clone(),
+    }
+}
+
+/// Iterate "words" like rich's `re.compile(r"\s*\S+\s*")`.
+fn words(text: &str) -> Vec<(usize, String)> {
+    let chars: Vec<char> = text.chars().collect();
+    let mut out = Vec::new();
+    let mut pos = 0;
+    while pos < chars.len() {
+        let start = pos;
+        let mut i = pos;
+        while i < chars.len() && chars[i].is_whitespace() {
+            i += 1;
+        }
+        if i >= chars.len() {
+            break; // trailing whitespace w/o a word: no match
+        }
+        while i < chars.len() && !chars[i].is_whitespace() {
+            i += 1;
+        }
+        while i < chars.len() && chars[i].is_whitespace() {
+            i += 1;
+        }
+        out.push((start, chars[start..i].iter().collect()));
+        pos = i;
+    }
+    out
+}
+
+fn cell_len(s: &str) -> usize {
+    UnicodeWidthStr::width(s)
+}
+
+/// Split text into chunks of at most `width` cells (rich `chop_cells`).
+fn chop_cells(text: &str, width: usize) -> Vec<String> {
+    let mut lines: Vec<String> = vec![String::new()];
+    let mut total_width = 0;
+    for ch in text.chars() {
+        let w = UnicodeWidthChar::width(ch).unwrap_or(0);
+        if total_width + w > width {
+            lines.push(ch.to_string());
+            total_width = w;
+        } else {
+            lines.last_mut().unwrap().push(ch);
+            total_width += w;
+        }
+    }
+    lines
+}
+
+/// Port of `rich._wrap.divide_line` (fold=True). Returns char offsets.
+fn divide_line(text: &str, width: usize) -> Vec<usize> {
+    let mut break_positions: Vec<usize> = Vec::new();
+    let mut cell_offset = 0usize;
+    for (start, word) in words(text) {
+        let word_length = cell_len(word.trim_end());
+        let remaining_space = width.saturating_sub(cell_offset);
+        if remaining_space >= word_length {
+            cell_offset += cell_len(&word);
+        } else if word_length > width {
+            // fold the word across multiple lines
+            let folded = chop_cells(&word, width);
+            let mut start = start;
+            let n = folded.len();
+            for (i, line) in folded.iter().enumerate() {
+                if start > 0 {
+                    break_positions.push(start);
+                }
+                if i + 1 == n {
+                    cell_offset = cell_len(line);
+                } else {
+                    start += line.chars().count();
+                }
+            }
+        } else if cell_offset > 0 && start > 0 {
+            break_positions.push(start);
+            cell_offset = cell_len(&word);
+        }
+    }
+    break_positions
+}
+
+/// Wrap one logical line into physical lines at `width` cells.
+fn wrap_line(line: &str, width: usize) -> Vec<String> {
+    let width = width.max(1);
+    let breaks = divide_line(line, width);
+    if breaks.is_empty() {
+        return vec![line.to_string()];
+    }
+    let chars: Vec<char> = line.chars().collect();
+    let mut out = Vec::new();
+    let mut prev = 0;
+    for &b in &breaks {
+        out.push(chars[prev..b].iter().collect());
+        prev = b;
+    }
+    out.push(chars[prev..].iter().collect());
+    out
+}
+
+/// Expand tabs to 8-cell stops like rich's `Text.expand_tabs`.
+fn expand_tabs(line: &str, tab_size: usize) -> String {
+    if !line.contains('\t') {
+        return line.to_string();
+    }
+    let mut out = String::with_capacity(line.len());
+    let mut col = 0;
+    for ch in line.chars() {
+        if ch == '\t' {
+            let pad = tab_size - (col % tab_size);
+            out.extend(std::iter::repeat_n(' ', pad));
+            col += pad;
+        } else {
+            out.push(ch);
+            col += UnicodeWidthChar::width(ch).unwrap_or(0);
+        }
+    }
+    out
+}
+
+/// Wrap a (possibly multiline) label into physical lines.
+fn wrap_label(label: &str, width: usize) -> Vec<String> {
+    let mut out = Vec::new();
+    for logical in label.split('\n') {
+        out.extend(wrap_line(&expand_tabs(logical, 8), width));
+    }
+    out
+}
+
+struct Renderer {
+    width: usize,
+    out: String,
+}
+
+impl Renderer {
+    fn emit(&mut self, prefix_first: &str, prefix_rest: &str, label: &str) {
+        let avail = self.width.saturating_sub(cell_len(prefix_first)).max(1);
+        for (i, line) in wrap_label(label, avail).into_iter().enumerate() {
+            let prefix = if i == 0 { prefix_first } else { prefix_rest };
+            let mut rendered = format!("{prefix}{line}");
+            while rendered.ends_with(' ') {
+                rendered.pop();
+            }
+            self.out.push_str(&rendered);
+            self.out.push('\n');
+        }
+    }
+
+    fn render_node(&mut self, node: &TreePlus, ancestors: &str, is_root: bool, is_last: bool) {
+        let (first, rest) = if is_root {
+            (String::new(), String::new())
+        } else {
+            let fork = if is_last { "└── " } else { "├── " };
+            let cont = if is_last { "    " } else { "│   " };
+            (format!("{ancestors}{fork}"), format!("{ancestors}{cont}"))
+        };
+        self.emit(&first, &rest, &node_label(node));
+
+        let child_ancestors = if is_root {
+            String::new()
+        } else if is_last {
+            format!("{ancestors}    ")
+        } else {
+            format!("{ancestors}│   ")
+        };
+
+        if node.category == Category::File {
+            let n = node.components.len();
+            for (i, component) in node.components.iter().enumerate() {
+                let last = i + 1 == n;
+                let fork = if last { "└── " } else { "├── " };
+                let cont = if last { "    " } else { "│   " };
+                self.emit(
+                    &format!("{child_ancestors}{fork}"),
+                    &format!("{child_ancestors}{cont}"),
+                    component,
+                );
+            }
+        } else {
+            let n = node.subtrees.len();
+            for (i, sub) in node.subtrees.iter().enumerate() {
+                self.render_node(sub, &child_ancestors, false, i + 1 == n);
+            }
+        }
+    }
+}
+
+/// Render a tree to a string identical to legacy `TreePlus.into_str()`.
+pub fn render_to_string(root: &TreePlus, width: usize) -> String {
+    let mut r = Renderer {
+        width,
+        out: String::new(),
+    };
+    r.render_node(root, "", true, true);
+    r.out
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::model::{Category, TreePlus};
+
+    #[test]
+    fn single_file_render() {
+        let mut file = TreePlus::new(Category::File, "file.py");
+        file.token_count = 19;
+        file.line_count = 3;
+        file.components.push("def hello_world()".to_string());
+        let s = render_to_string(&file, DEFAULT_WIDTH);
+        assert_eq!(
+            s,
+            "📄 file.py (19 tokens, 3 lines)\n└── def hello_world()\n"
+        );
+    }
+
+    #[test]
+    fn wrap_matches_rich_probe() {
+        // mirrors a probe of the legacy renderer
+        let mut root = TreePlus::new(Category::Component, "root");
+        root.category = Category::Folder; // unused; emit manually below
+        let long = "function julia_is_awesome(prob::DiffEqBase.AbstractDAEProblem{uType, duType, tType, isinplace};";
+        let mut r = Renderer {
+            width: 80,
+            out: String::new(),
+        };
+        r.emit("├── ", "│   ", long);
+        assert_eq!(
+            r.out,
+            "├── function julia_is_awesome(prob::DiffEqBase.AbstractDAEProblem{uType, duType,\n│   tType, isinplace};\n"
+        );
+    }
+
+    #[test]
+    fn fold_long_word() {
+        let mut r = Renderer {
+            width: 80,
+            out: String::new(),
+        };
+        let label = format!("word {} tail", "x".repeat(100));
+        r.emit("├── ", "│   ", &label);
+        let expected = format!(
+            "├── word\n│   {}\n│   {} tail\n",
+            "x".repeat(76),
+            "x".repeat(24)
+        );
+        assert_eq!(r.out, expected);
+    }
+
+    #[test]
+    fn multiline_label_continuation() {
+        let mut folder = TreePlus::new(Category::Folder, "f");
+        let mut file = TreePlus::new(Category::File, "a.py");
+        file.components
+            .push("    @staticmethod\n    def m()".to_string());
+        file.components.push("class X".to_string());
+        folder.subtrees.push(file);
+        let s = render_to_string(&folder, DEFAULT_WIDTH);
+        let expected = "\
+📁 f (1 folder, 1 file)
+└── 📄 a.py (0 tokens, 0 lines)
+    ├──     @staticmethod
+    │       def m()
+    └── class X
+";
+        assert_eq!(s, expected);
+    }
+}
diff --git a/crates/tree_plus_core/src/sort.rs b/crates/tree_plus_core/src/sort.rs
new file mode 100644
index 0000000..b150c96
--- /dev/null
+++ b/crates/tree_plus_core/src/sort.rs
@@ -0,0 +1,214 @@
+//! Natural path sorting compatible with `natsort.os_sorted` (POSIX, no PyICU).
+//!
+//! Empirically verified key structure (natsort 8.4.0):
+//! - a path is split into components (by `/`);
+//! - the final component is split into (base, suffix...) where at most the
+//!   last TWO suffixes are split off, a suffix must be `.`-prefixed with
+//!   total length <= 5 (e.g. `.json` yes, `.jsonl` no), and a purely numeric
+//!   suffix (e.g. `.12345`) stops suffix splitting entirely;
+//! - every piece is chunked into alternating (string, number, string, ...)
+//!   starting with a (possibly empty) string chunk; digit runs become
+//!   unsigned integers; text chunks compare case-insensitively (casefold).
+//!
+//! Ties (identical keys, e.g. names differing only by case) are broken by the
+//! raw string to keep the order deterministic; the Python implementation
+//! falls back to filesystem enumeration order here (documented difference).
+
+use std::cmp::Ordering;
+
+/// One chunk of a natural-sort key piece.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub enum Chunk {
+    Text(String),
+    Num(u128),
+}
+
+impl Chunk {
+    fn cmp_chunk(&self, other: &Chunk) -> Ordering {
+        match (self, other) {
+            (Chunk::Text(a), Chunk::Text(b)) => a.cmp(b),
+            (Chunk::Num(a), Chunk::Num(b)) => a.cmp(b),
+            // alternation makes mixed comparisons rare; numbers first matches
+            // natsort's empty-string-prefix convention
+            (Chunk::Num(_), Chunk::Text(_)) => Ordering::Less,
+            (Chunk::Text(_), Chunk::Num(_)) => Ordering::Greater,
+        }
+    }
+}
+
+/// Natural sort key for one path-like string.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct OsSortKey {
+    pieces: Vec<Vec<Chunk>>,
+    raw: String,
+}
+
+impl PartialOrd for OsSortKey {
+    fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
+        Some(self.cmp(other))
+    }
+}
+
+impl Ord for OsSortKey {
+    fn cmp(&self, other: &Self) -> Ordering {
+        let n = self.pieces.len().min(other.pieces.len());
+        for i in 0..n {
+            let a = &self.pieces[i];
+            let b = &other.pieces[i];
+            let m = a.len().min(b.len());
+            for j in 0..m {
+                match a[j].cmp_chunk(&b[j]) {
+                    Ordering::Equal => {}
+                    non_eq => return non_eq,
+                }
+            }
+            match a.len().cmp(&b.len()) {
+                Ordering::Equal => {}
+                non_eq => return non_eq,
+            }
+        }
+        match self.pieces.len().cmp(&other.pieces.len()) {
+            Ordering::Equal => self.raw.cmp(&other.raw),
+            non_eq => non_eq,
+        }
+    }
+}
+
+/// Chunk a string into alternating text/number chunks, starting with text.
+fn chunk(s: &str) -> Vec<Chunk> {
+    let mut chunks = Vec::new();
+    let mut text = String::new();
+    let mut num: Option<u128> = None;
+    for ch in s.chars() {
+        if ch.is_ascii_digit() {
+            if num.is_none() {
+                // flush text chunk (natsort keys always start with a string)
+                chunks.push(Chunk::Text(std::mem::take(&mut text).to_lowercase()));
+                num = Some(0);
+            }
+            let d = u128::from(ch as u8 - b'0');
+            num = Some(num.unwrap().saturating_mul(10).saturating_add(d));
+        } else {
+            if let Some(n) = num.take() {
+                chunks.push(Chunk::Num(n));
+            }
+            text.push(ch);
+        }
+    }
+    if let Some(n) = num {
+        chunks.push(Chunk::Num(n));
+    } else {
+        chunks.push(Chunk::Text(text.to_lowercase()));
+    }
+    chunks
+}
+
+/// Split the last path component into base + up to two short suffixes.
+fn split_suffixes(name: &str) -> Vec<String> {
+    // collect candidate suffixes from the right
+    let mut parts: Vec<String> = Vec::new();
+    let mut base = name.to_string();
+    for _ in 0..2 {
+        // a hidden-file leading dot is not a suffix boundary
+        let search_region = &base[1.min(base.len())..];
+        let Some(dot_rel) = search_region.rfind('.') else {
+            break;
+        };
+        let dot = dot_rel + 1;
+        let suffix = &base[dot..];
+        // suffix includes the dot; must be <= 5 chars and not purely numeric
+        if suffix.len() > 5 || suffix.len() < 2 {
+            break;
+        }
+        if suffix[1..].chars().all(|c| c.is_ascii_digit()) {
+            break;
+        }
+        parts.push(suffix.to_string());
+        base.truncate(dot);
+        if base.ends_with('.') {
+            base.pop();
+        }
+    }
+    let mut out = vec![base];
+    out.extend(parts.into_iter().rev());
+    out
+}
+
+/// Build the natural sort key for a path-like string.
+pub fn os_sort_key(path: &str) -> OsSortKey {
+    let mut pieces: Vec<Vec<Chunk>> = Vec::new();
+    let components: Vec<&str> = path.split('/').filter(|c| !c.is_empty()).collect();
+    for (i, comp) in components.iter().enumerate() {
+        if i + 1 == components.len() {
+            for piece in split_suffixes(comp) {
+                pieces.push(chunk(&piece));
+            }
+        } else {
+            pieces.push(chunk(comp));
+        }
+    }
+    if pieces.is_empty() {
+        pieces.push(chunk(path));
+    }
+    OsSortKey {
+        pieces,
+        raw: path.to_string(),
+    }
+}
+
+/// Sort strings like `natsort.os_sorted`.
+pub fn os_sorted<T: AsRef<str>>(items: &mut [T]) {
+    items.sort_by_key(|a| os_sort_key(a.as_ref()));
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    fn sorted_strs(mut v: Vec<&str>) -> Vec<&str> {
+        v.sort_by_key(|a| os_sort_key(a));
+        v
+    }
+
+    #[test]
+    fn numeric_runs_sort_naturally() {
+        assert_eq!(
+            sorted_strs(vec!["file10.py", "file2.py"]),
+            vec!["file2.py", "file10.py"]
+        );
+    }
+
+    #[test]
+    fn case_insensitive() {
+        assert_eq!(
+            sorted_strs(vec!["CUSTOMER-INVOICE.CBL", "addamt.cobol"]),
+            vec!["addamt.cobol", "CUSTOMER-INVOICE.CBL"]
+        );
+        assert_eq!(
+            sorted_strs(vec!["LuaTest.lua", "lesson.cbl", "KotlinTest.kt"]),
+            vec!["KotlinTest.kt", "lesson.cbl", "LuaTest.lua"]
+        );
+    }
+
+    #[test]
+    fn suffix_split_rules() {
+        // ".py" suffix split means "file.py" < "file2.py" (prefix tuple wins)
+        assert_eq!(
+            sorted_strs(vec!["file2.py", "file.py"]),
+            vec!["file.py", "file2.py"]
+        );
+        // long suffixes are not split: a.json < a.jsonl
+        assert_eq!(
+            sorted_strs(vec!["a.jsonl", "a.json"]),
+            vec!["a.json", "a.jsonl"]
+        );
+    }
+
+    #[test]
+    fn dotfiles_sort_before_letters() {
+        assert_eq!(
+            sorted_strs(vec!["claude.md", ".github", ".env.test"]),
+            vec![".env.test", ".github", "claude.md"]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/walk.rs b/crates/tree_plus_core/src/walk.rs
new file mode 100644
index 0000000..c519409
--- /dev/null
+++ b/crates/tree_plus_core/src/walk.rs
@@ -0,0 +1,260 @@
+//! Tree construction: the Rust analog of `tree_plus_src/engine.py`
+//! `from_seeds` / `_map_seeds` / `_from_folder` / `_from_file` / `_from_glob`.
+
+use std::path::{Path, PathBuf};
+
+use rayon::prelude::*;
+
+use crate::config::TreePlusConfig;
+use crate::count::count_tokens_lines;
+use crate::extract::extract_components;
+use crate::ignore::{
+    amortize_globs, is_glob, parse_ignore, should_ignore, AmortizedGlobs, IgnorePatterns,
+};
+use crate::model::{Category, TreePlus};
+use crate::sort::os_sort_key;
+
+/// Build a `TreePlus` from seed path/glob strings (legacy `from_seeds`).
+pub fn from_seeds(seeds: &[String], config: &TreePlusConfig) -> TreePlus {
+    let mut seeds: Vec<String> = if seeds.is_empty() {
+        vec![std::env::current_dir()
+            .map(|p| p.to_string_lossy().into_owned())
+            .unwrap_or_else(|_| ".".to_string())]
+    } else {
+        seeds.to_vec()
+    };
+    // legacy dedupes via set(); order is later fixed by sorting
+    seeds.sort();
+    seeds.dedup();
+
+    let ignore = IgnorePatterns::new(parse_ignore(&config.ignore, config.override_ignore));
+
+    // categorize seeds (legacy `_map_seeds`)
+    let mut folder_paths: Vec<PathBuf> = Vec::new();
+    let mut file_paths: Vec<PathBuf> = Vec::new();
+    let mut glob_seed_patterns: Vec<String> = Vec::new();
+    let mut filter_globs: Vec<String> = config.globs.clone();
+    for seed in &seeds {
+        let path = Path::new(seed);
+        if path.is_file() {
+            file_paths.push(path.to_path_buf());
+        } else if path.is_dir() {
+            folder_paths.push(path.to_path_buf());
+        } else if is_glob(seed) {
+            if seed.starts_with("*.") {
+                // legacy: bare `*.ext` seeds become filter globs
+                filter_globs.push(seed.clone());
+            } else {
+                glob_seed_patterns.push(seed.clone());
+            }
+        }
+        // non-existent non-glob seeds are skipped with a warning in legacy
+    }
+
+    // amortize filter globs (legacy)
+    let globs: Option<AmortizedGlobs> = if !filter_globs.is_empty() {
+        let glob_roots: Vec<PathBuf> = if folder_paths.is_empty() {
+            vec![std::env::current_dir().unwrap_or_else(|_| PathBuf::from("."))]
+        } else {
+            folder_paths.clone()
+        };
+        let mut amortized = amortize_globs(&glob_roots, &filter_globs);
+        if let Some(a) = amortized.as_mut() {
+            // directly-provided files dodge the glob filter
+            for f in &file_paths {
+                a.matches.insert(f.clone());
+            }
+        }
+        amortized
+    } else {
+        None
+    };
+    if !filter_globs.is_empty() && folder_paths.is_empty() && glob_seed_patterns.is_empty() {
+        // legacy: globs with no folder seeds scan the cwd
+        folder_paths.push(std::env::current_dir().unwrap_or_else(|_| PathBuf::from(".")));
+    }
+
+    // sort all seeds together by natural path order
+    enum Seed {
+        Folder(PathBuf),
+        File(PathBuf),
+        GlobPattern(String),
+    }
+    let mut parsed: Vec<(String, Seed)> = Vec::new();
+    for p in folder_paths {
+        parsed.push((p.to_string_lossy().into_owned(), Seed::Folder(p)));
+    }
+    for p in file_paths {
+        parsed.push((p.to_string_lossy().into_owned(), Seed::File(p)));
+    }
+    for g in glob_seed_patterns {
+        parsed.push((g.clone(), Seed::GlobPattern(g)));
+    }
+    parsed.sort_by_key(|a| os_sort_key(&a.0));
+
+    let forest: Vec<TreePlus> = parsed
+        .into_iter()
+        .map(|(_, seed)| match seed {
+            Seed::Folder(p) => from_folder(&p, &ignore, globs.as_ref(), config),
+            Seed::File(p) => from_file(&p, config),
+            Seed::GlobPattern(g) => from_glob_pattern(&g, &ignore, config),
+        })
+        .collect();
+
+    match forest.len() {
+        0 => TreePlus::new(Category::Root, "No match"),
+        1 => forest.into_iter().next().unwrap(),
+        _ => {
+            let mut root = TreePlus::new(Category::Root, "Root");
+            root.subtrees = forest;
+            root
+        }
+    }
+}
+
+/// Build a tree from a folder (legacy `_from_folder`).
+pub fn from_folder(
+    folder_path: &Path,
+    ignore: &IgnorePatterns,
+    globs: Option<&AmortizedGlobs>,
+    config: &TreePlusConfig,
+) -> TreePlus {
+    // resolve() fixes the `Path(".").name == ""` case
+    let name = folder_path
+        .canonicalize()
+        .ok()
+        .and_then(|p| p.file_name().map(|n| n.to_string_lossy().into_owned()))
+        .unwrap_or_else(|| folder_path.to_string_lossy().into_owned());
+    let mut node = TreePlus::new(Category::Folder, name);
+
+    let mut entries: Vec<PathBuf> = std::fs::read_dir(folder_path)
+        .map(|rd| rd.flatten().map(|e| e.path()).collect())
+        .unwrap_or_default();
+    entries.sort_by_key(|a| os_sort_key(&a.to_string_lossy()));
+
+    // partition first so files can be processed in parallel deterministically
+    let kept: Vec<(PathBuf, bool)> = entries
+        .into_iter()
+        .filter(|p| !should_ignore(p, ignore, globs))
+        .filter_map(|p| {
+            let is_dir = p.is_dir();
+            let is_file = p.is_file();
+            if is_dir || is_file {
+                Some((p, is_dir))
+            } else {
+                None
+            }
+        })
+        .collect();
+
+    let subtrees: Vec<TreePlus> = kept
+        .into_par_iter()
+        .map(|(p, is_dir)| {
+            if is_dir {
+                from_folder(&p, ignore, globs, config)
+            } else {
+                from_file(&p, config)
+            }
+        })
+        .collect();
+    node.subtrees = subtrees;
+    node
+}
+
+/// Build a tree from a file (legacy `_from_file`).
+pub fn from_file(file_path: &Path, config: &TreePlusConfig) -> TreePlus {
+    let counts = count_tokens_lines(file_path).unwrap_or_default();
+    let name = file_path
+        .file_name()
+        .map(|n| n.to_string_lossy().into_owned())
+        .unwrap_or_else(|| file_path.to_string_lossy().into_owned());
+    let mut node = TreePlus::new(Category::File, name);
+    node.token_count = counts.n_tokens;
+    node.line_count = counts.n_lines;
+    if !config.concise && counts.n_tokens <= config.max_tokens {
+        node.components = extract_components(file_path, config.syntax);
+    }
+    node
+}
+
+/// Build a tree from a glob-pattern seed like `tree_plus_src/*.py`
+/// (legacy `_from_glob`, which uses non-recursive `Path().glob(pattern)`).
+pub fn from_glob_pattern(
+    pattern: &str,
+    ignore: &IgnorePatterns,
+    config: &TreePlusConfig,
+) -> TreePlus {
+    let mut node = TreePlus::new(Category::Glob, pattern);
+    let mut matches = glob_relative(Path::new(""), pattern);
+    matches.sort_by_key(|a| os_sort_key(&a.to_string_lossy()));
+    for m in matches {
+        if m.is_dir() {
+            node.subtrees.push(from_folder(&m, ignore, None, config));
+        } else if m.is_file() {
+            node.subtrees.push(from_file(&m, config));
+        }
+    }
+    node
+}
+
+/// Minimal `pathlib.Path.glob` for relative patterns: segments are fnmatch
+/// globs, `**` matches any number of directories.
+fn glob_relative(base: &Path, pattern: &str) -> Vec<PathBuf> {
+    use crate::ignore::fnmatch_translate;
+    use regex::Regex;
+
+    let segments: Vec<&str> = pattern.split('/').filter(|s| !s.is_empty()).collect();
+    let mut current: Vec<PathBuf> = vec![if base.as_os_str().is_empty() {
+        PathBuf::from(".")
+    } else {
+        base.to_path_buf()
+    }];
+    for (i, segment) in segments.iter().enumerate() {
+        let last = i + 1 == segments.len();
+        let mut next: Vec<PathBuf> = Vec::new();
+        if *segment == "**" {
+            for dir in &current {
+                next.push(dir.clone());
+                collect_dirs_recursive(dir, &mut next);
+            }
+        } else {
+            let Ok(re) = Regex::new(&fnmatch_translate(segment)) else {
+                continue;
+            };
+            for dir in &current {
+                let Ok(entries) = std::fs::read_dir(dir) else {
+                    continue;
+                };
+                for entry in entries.flatten() {
+                    let name = entry.file_name();
+                    let Some(name) = name.to_str() else { continue };
+                    if re.is_match(name) {
+                        let p = entry.path();
+                        if last || p.is_dir() {
+                            next.push(p);
+                        }
+                    }
+                }
+            }
+        }
+        current = next;
+    }
+    // strip the leading "./" to mirror Path().glob output
+    current
+        .into_iter()
+        .map(|p| p.strip_prefix("./").map(Path::to_path_buf).unwrap_or(p))
+        .collect()
+}
+
+fn collect_dirs_recursive(dir: &Path, out: &mut Vec<PathBuf>) {
+    let Ok(entries) = std::fs::read_dir(dir) else {
+        return;
+    };
+    for entry in entries.flatten() {
+        let p = entry.path();
+        if p.is_dir() {
+            out.push(p.clone());
+            collect_dirs_recursive(&p, out);
+        }
+    }
+}
diff --git a/crates/tree_plus_core/tests/golden_parity.rs b/crates/tree_plus_core/tests/golden_parity.rs
new file mode 100644
index 0000000..468aac0
--- /dev/null
+++ b/crates/tree_plus_core/tests/golden_parity.rs
@@ -0,0 +1,243 @@
+//! Golden parity tests against the legacy Python implementation.
+//!
+//! Goldens are generated by `python tests/golden/generate_legacy_goldens.py`
+//! at the repository root and stored under tests/golden/legacy/.
+
+use std::collections::BTreeMap;
+use std::path::{Path, PathBuf};
+
+use tree_plus_core::{extract_components, from_seeds, TreePlusConfig};
+
+fn repo_root() -> PathBuf {
+    Path::new(env!("CARGO_MANIFEST_DIR"))
+        .join("../..")
+        .canonicalize()
+        .expect("repo root")
+}
+
+/// Extensions and file names covered by the version-1 scope.
+fn in_v1_scope(path: &str) -> bool {
+    let name = path.rsplit('/').next().unwrap_or(path);
+    let lower = name.to_lowercase();
+    if name == "Makefile" || name == "Justfile" || name.starts_with(".env") {
+        return true;
+    }
+    if name.ends_with("Cargo.toml") || name.ends_with("pyproject.toml") {
+        return true;
+    }
+    const V1_EXTS: &[&str] = &[
+        ".py",
+        ".pyi",
+        ".rs",
+        ".js",
+        ".jsx",
+        ".ts",
+        ".tsx",
+        ".c",
+        ".h",
+        ".cc",
+        ".cpp",
+        ".cu",
+        ".cuh",
+        ".hpp",
+        ".md",
+        ".markdown",
+        ".mdx",
+        ".mdc",
+        ".json",
+        ".jsonl",
+        ".yml",
+        ".yaml",
+        ".csv",
+        ".txt",
+        ".db",
+        ".sqlite",
+    ];
+    V1_EXTS.iter().any(|e| lower.ends_with(e))
+}
+
+/// Intentional differences from the legacy output, documented in
+/// docs/rust-port-differences.md. The expectations are adjusted precisely so
+/// any other drift still fails.
+fn apply_intentional_differences(rel: &str, expected: Vec<String>) -> Vec<String> {
+    match rel {
+        // The TensorFlow flag special case was deliberately dropped; the
+        // remaining C components must still match exactly.
+        "tests/more_languages/group6/tensorflow_flags.h" => expected
+            .into_iter()
+            .filter(|c| {
+                !(c.starts_with("Flag('")
+                    || c.starts_with("TF_DECLARE_FLAG('")
+                    || c.starts_with("TF_PY_DECLARE_FLAG('"))
+            })
+            .collect(),
+        // The legacy regex mistook a string literal containing the word
+        // "function" for a function definition; tree-sitter does not.
+        "tests/more_languages/group1/test.ts" => expected
+            .into_iter()
+            .filter(|c| c != "  return(\"Standalone function with parameters\")")
+            .collect(),
+        _ => expected,
+    }
+}
+
+fn parse_golden(path: &Path) -> (String, Vec<String>) {
+    let raw = std::fs::read_to_string(path).expect("read golden");
+    let value: serde_json::Value = serde_json::from_str(&raw).expect("parse golden json");
+    let rel = value["path"].as_str().expect("path").to_string();
+    let components = value["components"]
+        .as_array()
+        .expect("components")
+        .iter()
+        .map(|v| v.as_str().expect("component str").to_string())
+        .collect();
+    (rel, components)
+}
+
+#[test]
+fn components_match_legacy_goldens() {
+    let root = repo_root();
+    let goldens_dir = root.join("tests/golden/legacy/components");
+    let mut failures: BTreeMap<String, (Vec<String>, Vec<String>)> = BTreeMap::new();
+    let mut checked = 0;
+    let mut entries: Vec<PathBuf> = std::fs::read_dir(&goldens_dir)
+        .expect("goldens generated? run python tests/golden/generate_legacy_goldens.py")
+        .flatten()
+        .map(|e| e.path())
+        .collect();
+    entries.sort();
+    for golden_path in entries {
+        let (rel, expected) = parse_golden(&golden_path);
+        if !in_v1_scope(&rel) {
+            continue;
+        }
+        let expected = apply_intentional_differences(&rel, expected);
+        let actual = extract_components(&root.join(&rel), false);
+        checked += 1;
+        if actual != expected {
+            failures.insert(rel, (expected, actual));
+        }
+    }
+    assert!(
+        checked > 50,
+        "expected to check many fixtures, got {checked}"
+    );
+    if !failures.is_empty() {
+        let mut report = format!(
+            "{} of {} v1-scope fixtures differ from legacy goldens:\n",
+            failures.len(),
+            checked
+        );
+        for (rel, (expected, actual)) in &failures {
+            report.push_str(&format!(
+                "--- {rel}\n  expected ({}): {:?}\n  actual   ({}): {:?}\n",
+                expected.len(),
+                expected,
+                actual.len(),
+                actual
+            ));
+        }
+        panic!("{report}");
+    }
+}
+
+/// chdir is process-global; serialize the tree tests.
+static CWD_LOCK: std::sync::Mutex<()> = std::sync::Mutex::new(());
+
+fn tree_parity(name: &str, seeds: &[&str]) {
+    let root = repo_root();
+    // trees_v1: legacy renders with non-v1 extractors stubbed (deferred
+    // languages keep their counts and markers but lose components)
+    let golden =
+        std::fs::read_to_string(root.join(format!("tests/golden/legacy/trees_v1/{name}.txt")))
+            .expect("tree golden");
+    // legacy goldens were produced with cwd at the repo root
+    let _guard = CWD_LOCK
+        .lock()
+        .unwrap_or_else(std::sync::PoisonError::into_inner);
+    let original = std::env::current_dir().expect("cwd");
+    std::env::set_current_dir(&root).expect("chdir to repo root");
+    let seeds: Vec<String> = seeds.iter().map(|s| s.to_string()).collect();
+    let config = TreePlusConfig::default();
+    let tree = from_seeds(&seeds, &config);
+    let rendered = tree.into_str();
+    std::env::set_current_dir(original).expect("chdir back");
+    assert_eq!(
+        rendered, golden,
+        "tree render for {name} differs from legacy golden"
+    );
+}
+
+#[test]
+fn tree_path_to_test_matches() {
+    tree_parity("path_to_test", &["tests/path_to_test"]);
+}
+
+#[test]
+fn tree_dot_dot_matches() {
+    tree_parity("dot_dot", &["tests/dot_dot"]);
+}
+
+#[test]
+fn tree_multi_seed_matches() {
+    tree_parity(
+        "multi_seed",
+        &["tests/path_to_test", "tests/more_languages/group1"],
+    );
+}
+
+#[test]
+fn tree_group4_matches() {
+    tree_parity("more_languages_group4", &["tests/more_languages/group4"]);
+}
+
+#[test]
+fn tree_group6_matches() {
+    tree_parity("more_languages_group6", &["tests/more_languages/group6"]);
+}
+
+#[test]
+fn tree_more_languages_matches() {
+    tree_parity("more_languages", &["tests/more_languages"]);
+}
+
+#[test]
+fn tree_group1_matches() {
+    tree_parity("more_languages_group1", &["tests/more_languages/group1"]);
+}
+
+#[test]
+fn tree_group2_matches() {
+    tree_parity("more_languages_group2", &["tests/more_languages/group2"]);
+}
+
+#[test]
+fn tree_group3_matches() {
+    tree_parity("more_languages_group3", &["tests/more_languages/group3"]);
+}
+
+#[test]
+fn tree_group5_matches() {
+    tree_parity("more_languages_group5", &["tests/more_languages/group5"]);
+}
+
+#[test]
+fn tree_group7_matches() {
+    tree_parity("more_languages_group7", &["tests/more_languages/group7"]);
+}
+
+#[test]
+fn tree_group_todo_matches() {
+    tree_parity(
+        "more_languages_group_todo",
+        &["tests/more_languages/group_todo"],
+    );
+}
+
+#[test]
+fn tree_group_lisp_matches() {
+    tree_parity(
+        "more_languages_group_lisp",
+        &["tests/more_languages/group_lisp"],
+    );
+}
diff --git a/crates/tree_plus_core/tests/robustness.rs b/crates/tree_plus_core/tests/robustness.rs
new file mode 100644
index 0000000..79bb957
--- /dev/null
+++ b/crates/tree_plus_core/tests/robustness.rs
@@ -0,0 +1,86 @@
+//! Robustness: arbitrary bytes must never panic, hang, or render
+//! nondeterministically. (Deterministic pseudo-random corpus; no fuzz deps.)
+
+use std::io::Write;
+use std::path::PathBuf;
+
+use tree_plus_core::extract_components;
+
+/// xorshift64* — deterministic byte stream without external crates.
+struct Rng(u64);
+impl Rng {
+    fn next_u64(&mut self) -> u64 {
+        let mut x = self.0;
+        x ^= x >> 12;
+        x ^= x << 25;
+        x ^= x >> 27;
+        self.0 = x;
+        x.wrapping_mul(0x2545F4914F6CDD1D)
+    }
+    fn fill(&mut self, buf: &mut [u8]) {
+        for chunk in buf.chunks_mut(8) {
+            let bytes = self.next_u64().to_le_bytes();
+            let n = chunk.len();
+            chunk.copy_from_slice(&bytes[..n]);
+        }
+    }
+}
+
+const EXTENSIONS: &[&str] = &[
+    "py", "rs", "ts", "tsx", "js", "c", "cpp", "h", "md", "json", "jsonl", "yml", "toml", "csv",
+    "txt", "env", "rst",
+];
+
+fn write_temp(name: &str, bytes: &[u8]) -> PathBuf {
+    let dir = std::env::temp_dir().join("tree_plus_robustness");
+    std::fs::create_dir_all(&dir).unwrap();
+    let path = dir.join(name);
+    let mut f = std::fs::File::create(&path).unwrap();
+    f.write_all(bytes).unwrap();
+    path
+}
+
+#[test]
+fn arbitrary_bytes_never_panic_and_are_deterministic() {
+    let mut rng = Rng(0x5EED_CAFE_F00D_BEEF);
+    for round in 0..40 {
+        let size = (rng.next_u64() % 4096) as usize + 1;
+        let mut bytes = vec![0u8; size];
+        rng.fill(&mut bytes);
+        for ext in EXTENSIONS {
+            let path = write_temp(&format!("garbage_{round}.{ext}"), &bytes);
+            let first = extract_components(&path, false);
+            let second = extract_components(&path, false);
+            assert_eq!(first, second, "nondeterministic output for .{ext}");
+        }
+    }
+}
+
+#[test]
+fn truncated_real_sources_never_panic() {
+    let root = PathBuf::from(env!("CARGO_MANIFEST_DIR")).join("../..");
+    let sources = [
+        "tests/more_languages/group4/rust_test.rs",
+        "tests/path_to_test/class_method_type.py",
+        "tests/more_languages/group1/test.ts",
+        "tests/more_languages/group3/cpp_test.cpp",
+        "tests/more_languages/group2/test.csv",
+    ];
+    for rel in sources {
+        let bytes = std::fs::read(root.join(rel)).unwrap();
+        let ext = rel.rsplit('.').next().unwrap();
+        // cut at byte boundaries, including mid-UTF-8 and mid-token
+        for cut in [1, 7, bytes.len() / 3, bytes.len() / 2, bytes.len() - 1] {
+            let cut = cut.min(bytes.len());
+            let path = write_temp(&format!("truncated_{cut}.{ext}"), &bytes[..cut]);
+            let _ = extract_components(&path, false); // must not panic
+        }
+    }
+}
+
+#[test]
+fn invalid_utf8_yields_no_components() {
+    // legacy read_file used strict UTF-8 and returned "" on decode errors
+    let path = write_temp("invalid.py", &[0xC3, 0x28, b'\n', b'd', b'e', b'f', b' ']);
+    assert!(extract_components(&path, false).is_empty());
+}

From a26c45d3281b5b3200d3a4f97d93af8961c379be Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 15:48:03 -0400
Subject: [PATCH 4/8] docs: architecture, port differences, language roadmap,
 performance

Plus a Justfile with install (tprs), test, lint, goldens, and bench
recipes.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 Justfile                      |  23 +++++++
 docs/architecture.md          | 113 ++++++++++++++++++++++++++++++++++
 docs/language-roadmap.md      |  64 +++++++++++++++++++
 docs/performance.md           |  64 +++++++++++++++++++
 docs/rust-port-differences.md |  78 +++++++++++++++++++++++
 5 files changed, 342 insertions(+)
 create mode 100644 Justfile
 create mode 100644 docs/architecture.md
 create mode 100644 docs/language-roadmap.md
 create mode 100644 docs/performance.md
 create mode 100644 docs/rust-port-differences.md

diff --git a/Justfile b/Justfile
new file mode 100644
index 0000000..797b3fe
--- /dev/null
+++ b/Justfile
@@ -0,0 +1,23 @@
+# tree_plus Rust port
+
+# install the `tprs` alias binary (avoids PATH collision with Python tree_plus)
+install:
+	cargo install --path crates/tree_plus_cli --bin tprs
+
+# install both binaries: `tree_plus` and `tprs`
+install-all:
+	cargo install --path crates/tree_plus_cli
+
+test:
+	cargo test --workspace --all-features
+
+lint:
+	cargo fmt --all -- --check
+	cargo clippy --all-targets --all-features -- -D warnings
+
+# regenerate legacy goldens from the Python implementation
+goldens:
+	python tests/golden/generate_legacy_goldens.py
+
+bench:
+	cargo bench -p tree_plus_core
diff --git a/docs/architecture.md b/docs/architecture.md
new file mode 100644
index 0000000..00f5351
--- /dev/null
+++ b/docs/architecture.md
@@ -0,0 +1,113 @@
+# tree_plus Rust Port — Architecture
+
+## Layout
+
+```
+crates/tree_plus_core      # library
+  src/model.rs             # TreePlus, Category, counts
+  src/config.rs            # TreePlusConfig (CLI flags map here)
+  src/walk.rs              # traversal + tree construction (legacy engine.py)
+  src/ignore.rs            # fnmatch-compatible ignore + amortized globs
+  src/sort.rs              # natural path sort (natsort.os_sorted parity)
+  src/count.rs             # wc-style token/line counting
+  src/render.rs            # rich-Tree-compatible deterministic renderer
+  src/extract/mod.rs       # dispatch (the legacy parse_file, renamed)
+  src/extract/markers.rs   # TODO/BUG/NOTE markers
+  src/extract/markdown.rs  # Markdown, RST, txt checkboxes
+  src/extract/simple.rs    # .env, requirements, Makefile, Angular helpers
+  src/extract/data.rs      # JSON family, JSONL, YAML, TOML, CSV, SQLite
+  src/extract/treesitter/  # Rust, Python, JS/TS, C/C++ extractors
+  tests/golden_parity.rs   # byte-level parity vs legacy Python goldens
+  benches/tree_plus_bench.rs
+crates/tree_plus_cli       # clap binary named `tree_plus`
+tests/golden/              # golden generator + legacy goldens
+```
+
+## Naming correction
+
+The legacy `parse_file` name was misleading: it never parsed files
+generically; it extracted displayable component labels. The Rust analog is
+`extract::extract_components(path, syntax)`.
+
+## Why Tree-sitter for the major languages
+
+Rust, Python, JavaScript/TypeScript, C and C++ use tree-sitter grammars:
+
+- real parsing handles strings, comments, and nesting that regexes cannot;
+- invalid syntax degrades to partial components (ERROR-node salvage) instead
+  of catastrophic behavior;
+- parsers are reused per thread (`thread_local` parser, immutable grammars).
+
+The formatters do **not** dump AST text. Each language has a formatter that
+emits source-text signature slices matching the legacy tree_plus style, with
+the legacy regex quirks reproduced deliberately where tests depend on them
+(e.g. Python `async def` is skipped, Rust `pub(crate)` items are skipped,
+C++ enum variants keep their trailing commas). Where the legacy regex
+produced demonstrably wrong output, the difference is documented in
+docs/rust-port-differences.md rather than replicated.
+
+## Why regex stays for the simple extractors
+
+Markers, Markdown/RST headings, .env, requirements.txt, Makefile targets and
+similar line-oriented formats are matched with the `regex` crate, compiled
+once via `std::sync::LazyLock`. These patterns are anchored, bounded, and
+have no look-around, so the linear-time `regex` engine is sufficient and
+safe (no catastrophic backtracking by construction).
+
+## Why fancy-regex is not used
+
+An early draft ported the legacy TypeScript/C regexes verbatim with
+fancy-regex (the originals need look-around). It was removed in favor of
+tree-sitter formatters: backtracking regexes in the hot path are exactly the
+failure mode the rewrite exists to remove (see `catastrophic.c`, a fixture
+that exists because the legacy regex approach nearly exploded on it).
+
+## Why nom/pest are not used
+
+No mainstream language gets a homegrown grammar. Tree-sitter covers the big
+languages; everything else in scope is line-oriented (regex) or structured
+data with a native parser (serde_json, toml, serde_yaml, rusqlite).
+
+## Data and config parsers
+
+- JSON / package.json / JSON-RPC / OpenRPC / JSON Schema: `serde_json`
+  (with `preserve_order` so output order matches the file).
+- TOML (Cargo.toml, pyproject.toml): `toml` (with `preserve_order`).
+- YAML (k8s, Ansible, GitHub workflows, OpenAPI): `serde_yaml` multi-doc.
+- SQLite (.db/.sqlite): `rusqlite` behind the default-on `sqlite` feature.
+- CSV: header line split, mirroring the legacy first-3-lines behavior
+  (the csv crate would change behavior; the legacy reads only the header).
+
+## Traversal, determinism, parallelism
+
+- `walk.rs` recursion mirrors the legacy `Path.iterdir()` walk; entries are
+  sorted with `sort::os_sort_key` before any parallel work, then files are
+  processed with rayon (`into_par_iter`) and collected in order, so output
+  is deterministic regardless of thread scheduling.
+- Ignores match any path component via fnmatch translation; glob filters are
+  amortized exactly like the legacy `AmortizedGlobs` (match set + ancestors).
+- Binary detection reads at most 1 KiB; counting skips the legacy extension
+  list; files above 1M tokens skip extraction.
+
+## Rendering
+
+`render.rs` reimplements the small slice of `rich` the legacy output
+depends on: tree guides, multiline label continuation, cell-width-aware
+word wrapping (`divide_line`/`chop_cells` ports, fold semantics), 8-cell tab
+expansion, and trailing-space removal. Width is 80 when piped (the legacy
+capture width), the terminal width when interactive, and 128 under
+`TREE_PLUS_UPDATE_README=YES`.
+
+## How to add a new language
+
+1. Add the extension to the dispatch in `extract/mod.rs` (order matters and
+   is tested; mirror the legacy dispatch position).
+2. Tree-sitter route: add the grammar crate, write a formatter in
+   `extract/treesitter/<lang>.rs` that emits signature slices, and recurse
+   into bodies for nested items.
+3. Regenerate goldens: `python tests/golden/generate_legacy_goldens.py`
+   (remove the language from `DEFERRED_PARSERS` in the generator so the v1
+   tree goldens include its components).
+4. The fixture parity test picks the extension up via `in_v1_scope` in
+   `crates/tree_plus_core/tests/golden_parity.rs`; add it there.
+5. Add unit tests for the formatter and at least one pathological input.
diff --git a/docs/language-roadmap.md b/docs/language-roadmap.md
new file mode 100644
index 0000000..5f2cb7a
--- /dev/null
+++ b/docs/language-roadmap.md
@@ -0,0 +1,64 @@
+# Language Roadmap (Rust Port)
+
+Version-1 implements: Rust, Python, JavaScript, TypeScript, C, C++, Markdown
+(+ RST), JSON (package.json / schema / RPC / OpenRPC), JSONL, YAML, TOML
+(Cargo/pyproject), CSV, Makefile/Justfile, .env, requirements.txt, SQLite,
+and TODO/BUG/NOTE markers everywhere except `.md`/`.txt` (legacy rule).
+
+Everything below is recognized by the legacy Python implementation and
+**deferred**: files keep counts and markers, but emit no components. Legacy
+goldens for every deferred language are checked in under
+`tests/golden/legacy/components/` and serve as the acceptance contract.
+For each: legacy extension(s) → legacy extractor → tree-sitter grammar
+availability → suggested path → missing tests.
+
+| Language | Extensions | Legacy extractor | TS grammar? | Suggested path | Missing tests |
+|---|---|---|---|---|---|
+| Go | .go | parse_go | yes (tree-sitter-go) | tree-sitter formatter | port golden `go_test.go` |
+| Java | .java | parse_java | yes | tree-sitter formatter | port golden `JavaTest.java` |
+| Kotlin | .kt | parse_kt | yes (community) | tree-sitter formatter | port golden `KotlinTest.kt` |
+| Swift | .swift | parse_swift | yes (community) | tree-sitter formatter | port golden `swift_test.swift` |
+| C# | .cs | parse_cs | yes | tree-sitter formatter | port golden `csharp_test.cs` |
+| PHP | .php | parse_php | yes | tree-sitter formatter | port golden `php_test.php` |
+| Ruby | .rb | parse_rb | yes | tree-sitter formatter | port golden `ruby_test.rb` |
+| Scala | .scala | parse_scala | yes (community) | tree-sitter formatter | port golden `scala_test.scala` |
+| Julia | .jl | parse_jl | yes (community) | tree-sitter formatter | port golden `JuliaTest.jl` |
+| Lua | .lua | parse_lua | yes | tree-sitter formatter | port golden `LuaTest.lua` |
+| Haskell | .hs | parse_hs | yes (community) | tree-sitter formatter | port golden `haskell_test.hs` |
+| OCaml | .ml | parse_ocaml | yes | tree-sitter formatter | port golden `OcamlTest.ml` |
+| F# | .fs | parse_fsharp | partial | regex port (line-oriented) | port golden `fsharp_test.fs` |
+| Erlang | .erl .hrl | parse_erl | yes (community) | regex port (attribute lines) | port golden `erl_test.erl` |
+| Lisp family | .lisp .clj .scm .el .rkt | parse_lisp | per-dialect | regex port (defun-ish lines) | port group_lisp goldens |
+| Perl | .pl | parse_perl | yes (community) | regex port | port golden `perl_test.pl` |
+| PowerShell | .ps1 | parse_ps1 | no maintained | regex port | port golden `powershell_test.ps1` |
+| Bash | .sh | parse_bash | yes | tree-sitter formatter | port golden `bash_test.sh` |
+| Zig | .zig | parse_zig | yes (community) | tree-sitter formatter | port golden `zig_test.zig` |
+| R | .r .R | parse_r | yes (community) | regex port | port golden `r_test.r` |
+| MATLAB | .m .matlab | parse_matlab | no maintained | regex port + content sniff | port golden `matlab_test.m` |
+| Objective-C | .m | parse_objective_c | yes (community) | tree-sitter + content sniff | port golden `objective_c_test.m` |
+| Mathematica | .nb .wl | parse_mathematica | no | regex port | port golden `mathematica_test.nb` |
+| COBOL | .cbl .cobol | parse_cbl | no maintained | regex port (division/para lines) | port group1 goldens |
+| Fortran | .f .f90 ... | parse_fortran | yes (community) | regex port | port golden `fortran_test.f90` |
+| APL | .apl | parse_apl | no | regex port | port golden `apl_test.apl` |
+| SQL | .sql | parse_sql | yes (community) | regex port (CREATE TABLE) | port golden `sql_test.sql` |
+| GraphQL | .graphql | parse_graphql | yes (community) | line port (trivial) | port golden `graphql_test.graphql` |
+| Protobuf | .proto | parse_grpc | yes (community) | line/regex port | port golden `proto_test.proto` |
+| Cap'n Proto | .capnp | parse_capnp | no | regex port | port golden `capnp_test.capnp` |
+| LaTeX | .tex | parse_tex | yes (community) | regex port (sections) | port golden `tex_test.tex` |
+| Lean | .lean | parse_lean | community | regex port | port golden `lean_test.lean` |
+| Isabelle | .thy | parse_isabelle | no | regex port + symbol table | port golden `isabelle_test.thy` |
+| Terraform | .tf | parse_tf | yes (hcl) | regex port | port golden `terraform_test.tf` |
+| TCL | .tcl | parse_tcl | community | regex port (proc lines) | port golden `tcl_test.tcl` |
+| Metal | .metal | parse_metal | no | regex port | port golden `metal_test.metal` |
+| WGSL | .wgsl | parse_wgsl | community | regex port | port golden `wgsl_test.wgsl` |
+| HTML | .html | parse_html (returns []) | yes | decide scope first (legacy emitted nothing) | n/a |
+| TensorFlow flags | flags-in-tensorflow .cc/.h | parse_tensorflow_flags | n/a | intentionally removed (see differences doc) | n/a |
+
+Also deferred (not languages): web mode (`URL seeds`, `--yc`), syntax
+highlighting, tiktoken tokenizers — see docs/rust-port-differences.md.
+
+## Suggested order of attack
+
+1. Go, Java, Kotlin, C#, Ruby, Bash (mature grammars, heavily used).
+2. SQL/GraphQL/Protobuf/requirements-style line formats (cheap regex ports).
+3. The long tail, prioritized by user demand.
diff --git a/docs/performance.md b/docs/performance.md
new file mode 100644
index 0000000..b273e28
--- /dev/null
+++ b/docs/performance.md
@@ -0,0 +1,64 @@
+# Performance Report (Rust Port)
+
+Measured 2026-06-09 on Linux (WSL2), release build, default features.
+Comparisons via hyperfine (warmup 1) against the legacy Python CLI
+(Python 3.10, tree_plus 1.0.79). "Full" = components extracted; "concise"
+= traversal + counting only.
+
+## End-to-end CLI, this repository (473 files, ~50.7k lines, ~540k tokens)
+
+| Invocation | Rust | Python | Speedup |
+|---|---|---|---|
+| `tree_plus -c .` (traversal+counting) | 11.8 ms ± 1.3 | 782.8 ms ± 3.5 | **66×** |
+| `tree_plus .` (full extraction) | 28.0 ms ± 1.9 | 2.385 s ± 0.022 | **85×** |
+
+Throughput at the full-extraction figure: ~17k files/s; ~75 MB/s of
+repository text (2.1 MB of counted text; 7.5 MB on disk incl. binaries).
+
+## Medium repositories (full extraction)
+
+| Repository | Kind | Stats (Rust footer) | Rust wall | Speedup vs Python |
+|---|---|---|---|---|
+| Gymnasium | Python | 684 files, 63k lines, 704k tokens | ~20 ms | **84×** |
+| diamond-types | Rust | 290 files, 75k lines, 1.27M tokens | ~40 ms | **70×** |
+| ramda | JavaScript | 700 files, 46k lines, 389k tokens | ~30 ms | **55×** |
+
+Reproduce with any local tree:
+
+```
+cargo build --release
+hyperfine './target/release/tree_plus /path/to/repo > /dev/null' \
+          'python tree_plus_cli.py /path/to/repo > /dev/null'
+```
+
+A Linux-kernel-sized tree was not available locally; the command above is
+the reproducible benchmark for one (also:
+`TREE_PLUS_BENCH_PATH=/path/to/linux cargo bench -p tree_plus_core`).
+
+## Single-file extraction latency (criterion, mean)
+
+| Fixture | Extractor | Mean |
+|---|---|---|
+| class_method_type.py (101 lines) | tree-sitter Python | 217 µs |
+| test.ts (165 lines) | tree-sitter TypeScript | 386 µs |
+| rust_test.rs (~190 lines) | tree-sitter Rust | 670 µs |
+| catastrophic.c (754 lines, pathological) | tree-sitter C | 1.81 ms |
+
+Criterion reports p50≈mean here with tight bounds (±1%); the pathological
+C file that motivated the rewrite extracts in under 2 ms (the legacy regex
+needed a timeout guard for files like it).
+
+Run: `cargo bench -p tree_plus_core` (whole-repo benches accept
+`TREE_PLUS_BENCH_PATH`).
+
+## Where the time goes / tuning applied
+
+- Files are processed in parallel with rayon after a deterministic sort, so
+  output order never depends on scheduling; no locks in the hot path.
+- One tree-sitter parser per thread (`thread_local`), grammars shared.
+- All regexes compiled once (`LazyLock`); only the linear-time `regex`
+  crate is used (no backtracking engine anywhere).
+- Binary sniffing is capped at 1 KiB; CSV/JSONL read only the first lines;
+  files over 1M tokens skip extraction (legacy behavior).
+- Peak memory (full run on this repo): 43 MB RSS (`/usr/bin/time -v`),
+  dominated by per-file contents held during parallel extraction.
diff --git a/docs/rust-port-differences.md b/docs/rust-port-differences.md
new file mode 100644
index 0000000..eb7e7f9
--- /dev/null
+++ b/docs/rust-port-differences.md
@@ -0,0 +1,78 @@
+# Intentional Differences from the Python Implementation
+
+Everything not listed here is byte-identical for the version-1 scope; the
+golden parity suite (`cargo test -p tree_plus_core --test golden_parity`)
+enforces it against outputs captured from the legacy implementation.
+
+## Output differences (version-1 languages)
+
+1. **TensorFlow flag special case removed.** Legacy `parse_file` special-cased
+   `.cc`/`.h` files whose path contained "tensorflow" and whose name contained
+   "flags", emitting `Flag('...')` components. Deliberately dropped (decision:
+   too project-specific for a general tool). `tensorflow_flags.h` still gets
+   full C/C++ extraction. The parity test pins the exact remaining components.
+
+2. **String literals are no longer mistaken for code.** The legacy TS regex
+   emitted `  return("Standalone function with parameters")` because the
+   string contains the word "function". Tree-sitter does not parse string
+   contents, so this artifact is gone.
+
+3. **A few regex-noise lines are no longer emitted** for TS/JS and C/C++:
+   the legacy "method" pattern matched some bare call statements. The common
+   ones (`super(...)`, simple `name(args);` calls, `while((...))`, indented
+   `asm(...)`, pybind `.def("...` chains) are still emitted for parity, but
+   awaited calls and calls with complex callees that the legacy regex happened
+   to match in other codebases may differ. Fixture-covered cases all match.
+
+4. **Type predicates are no longer truncated mid-token in new code paths.**
+   Where the legacy lazily stopped a TS return type at the first ` {`
+   (e.g. `): ticket is`), the port reproduces the truncation for parity; this
+   is noted as a known legacy wart to revisit in version 2.
+
+## Binary name / migration
+
+The Rust CLI builds two identical binaries: `tree_plus` (legacy-compatible
+name) and `tprs` (alias that avoids PATH collisions with the Python
+`tree_plus` entry point, e.g. from a conda env). Install one or both:
+
+```
+cargo install --path crates/tree_plus_cli --bin tprs        # alias only
+cargo install --path crates/tree_plus_cli                   # both binaries
+```
+
+## Behavioral differences
+
+5. **Tokenizers.** Only the default word-count tokenizer (`wc`: tokens =
+   characters / 4) is implemented. `-t` / `-T gpt4o` produce an explicit
+   error instead of importing tiktoken. (`-T wc` is accepted.)
+
+6. **Web modes are deferred.** URL seeds, `--yc`/`--hn`, `-n`, `-m`, and
+   `-l` are not implemented; the CLI does not accept the HN flags. Local
+   filesystem mode is complete.
+
+7. **Syntax highlighting is deferred.** Output is plain text (identical to
+   the legacy `into_str()` capture). `-s` is accepted for compatibility.
+   Note the legacy enum markup escaping (e.g. `#\[default]`) is preserved so
+   plain-text output matches the legacy renderer byte-for-byte.
+
+8. **`--timeout` is accepted and ignored.** The Rust extractors are
+   linear-time; there is no regex timeout to configure. The legacy behavior
+   of returning zero components when extraction fails is preserved.
+
+9. **Sort tie-breaking.** Names that differ only by case (identical natural
+   sort keys) fall back to raw byte order in Rust; Python fell back to
+   filesystem enumeration order (nondeterministic). Deterministic on purpose.
+
+10. **`COLUMNS`/TTY width.** When piped, output wraps at width 80 exactly like
+    the legacy `into_str()`. Interactive terminals use the real width, like
+    the legacy CLI did; `TREE_PLUS_UPDATE_README=YES` forces 128. The legacy
+    interactive renderer also applied `tab_size=2` and rich markup; the Rust
+    CLI always renders the deterministic plain-text form (tab size 8).
+
+11. **`read_file` caching.** The legacy cached file contents process-wide
+    (`lru_cache`); the port reads files once per run with no global cache.
+
+12. **Deferred-language files** (see docs/language-roadmap.md) still appear
+    in trees with correct counts and TODO/BUG/NOTE markers, but no language
+    components. The legacy goldens for those languages are kept as the
+    contract for future ports (`tests/golden/legacy/components/`).

From 840145c37c356057389a2e34d3f77be25b2d43d7 Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 15:51:16 -0400
Subject: [PATCH 5/8] feat: Go component extraction (tree-sitter)

type struct/interface headers and column-0 func signatures, matching
legacy parse_go exactly (golden-parity tested, including the multiline
method signature fixture). Legacy quirks preserved: `func f() {}` and
generic type headers are skipped, raw tabs kept in signatures.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 Cargo.lock                                    |  11 ++
 Cargo.toml                                    |   1 +
 crates/tree_plus_core/Cargo.toml              |   1 +
 crates/tree_plus_core/src/extract/mod.rs      |   3 +-
 .../src/extract/treesitter/go.rs              | 152 ++++++++++++++++++
 .../src/extract/treesitter/mod.rs             |   1 +
 crates/tree_plus_core/tests/golden_parity.rs  |   1 +
 docs/language-roadmap.md                      |  12 +-
 tests/golden/generate_legacy_goldens.py       |   2 +-
 tests/golden/legacy/trees/repo_concise.txt    |  58 +++----
 .../golden/legacy/trees_v1/more_languages.txt |  13 ++
 .../legacy/trees_v1/more_languages_group2.txt |  13 ++
 12 files changed, 233 insertions(+), 35 deletions(-)
 create mode 100644 crates/tree_plus_core/src/extract/treesitter/go.rs

diff --git a/Cargo.lock b/Cargo.lock
index 1510d6f..700cc68 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -857,6 +857,16 @@ dependencies = [
  "tree-sitter-language",
 ]
 
+[[package]]
+name = "tree-sitter-go"
+version = "0.25.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c8560a4d2f835cc0d4d2c2e03cbd0dde2f6114b43bc491164238d333e28b16ea"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
+
 [[package]]
 name = "tree-sitter-javascript"
 version = "0.25.0"
@@ -928,6 +938,7 @@ dependencies = [
  "tree-sitter",
  "tree-sitter-c",
  "tree-sitter-cpp",
+ "tree-sitter-go",
  "tree-sitter-javascript",
  "tree-sitter-python",
  "tree-sitter-rust",
diff --git a/Cargo.toml b/Cargo.toml
index afff7ec..528074f 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -25,5 +25,6 @@ tree-sitter-javascript = "0.25"
 tree-sitter-typescript = "0.23"
 tree-sitter-c = "0.24"
 tree-sitter-cpp = "0.23"
+tree-sitter-go = "0.25"
 rusqlite = { version = "0.32", features = ["bundled"] }
 criterion = "0.5"
diff --git a/crates/tree_plus_core/Cargo.toml b/crates/tree_plus_core/Cargo.toml
index 89fead3..58bf87b 100644
--- a/crates/tree_plus_core/Cargo.toml
+++ b/crates/tree_plus_core/Cargo.toml
@@ -27,6 +27,7 @@ tree-sitter-typescript.workspace = true
 tree-sitter-c.workspace = true
 tree-sitter-cpp.workspace = true
 rusqlite = { workspace = true, optional = true }
+tree-sitter-go.workspace = true
 
 [dev-dependencies]
 criterion.workspace = true
diff --git a/crates/tree_plus_core/src/extract/mod.rs b/crates/tree_plus_core/src/extract/mod.rs
index c3572ff..5e28bd4 100644
--- a/crates/tree_plus_core/src/extract/mod.rs
+++ b/crates/tree_plus_core/src/extract/mod.rs
@@ -109,7 +109,7 @@ const MARKDOWN_EXTENSIONS: &[&str] = &[".md", ".markdown", ".mdx", ".mdc"];
 /// Rust port version 1. Files still get TODO/BUG/NOTE markers; component
 /// extraction is tracked in docs/language-roadmap.md.
 const DEFERRED_EXTENSIONS: &[&str] = &[
-    ".php", ".kt", ".swift", ".go", ".sh", ".ps1", ".zig", ".rb", ".sql", ".graphql", ".cs", ".jl",
+    ".php", ".kt", ".swift", ".sh", ".ps1", ".zig", ".rb", ".sql", ".graphql", ".cs", ".jl",
     ".scala", ".java", ".pl", ".hs", ".fs", ".lisp", ".clj", ".scm", ".el", ".rkt", ".erl", ".hrl",
     ".capnp", ".proto", ".tex", ".lean", ".f", ".for", ".f77", ".f90", ".f95", ".f03", ".f08",
     ".tf", ".thy", ".lua", ".tcl", ".m", ".r", ".nb", ".wl", ".matlab", ".ml", ".cbl", ".cobol",
@@ -206,6 +206,7 @@ fn try_extract_components(path: &Path, syntax: bool) -> ExtractResult {
         ".yml" | ".yaml" => data::extract_yml(&content)?,
         e if C_EXTENSIONS.contains(&e) => treesitter::c_cpp::extract(&content, e)?,
         ".rs" => treesitter::rust::extract(&content, syntax)?,
+        ".go" => treesitter::go::extract(&content)?,
         ".jsonl" => data::extract_jsonl(&content)?,
         ".env" => simple::dot_env(&content),
         ".txt" => {
diff --git a/crates/tree_plus_core/src/extract/treesitter/go.rs b/crates/tree_plus_core/src/extract/treesitter/go.rs
new file mode 100644
index 0000000..9efde45
--- /dev/null
+++ b/crates/tree_plus_core/src/extract/treesitter/go.rs
@@ -0,0 +1,152 @@
+//! Go component extraction (tree-sitter), matching legacy `parse_go`.
+//!
+//! Legacy semantics, reproduced deliberately:
+//! - `type Name struct` / `type Name interface` headers (no generics: the
+//!   legacy pattern required exactly `type \w+ (struct|interface)`);
+//! - column-0 `func ...` signatures (declarations and methods), sliced up
+//!   to the body brace, which must be surrounded by whitespace (the legacy
+//!   lookahead was `(?=\s\{\s)`, so `func f() {}` never matched);
+//! - signatures keep raw source text (tabs included; the legacy extractor
+//!   did not strip comments for Go).
+
+use std::sync::LazyLock;
+
+use regex::Regex;
+use tree_sitter::Node;
+
+use super::{parse, ExtractResult};
+
+static TYPE_HEADER_RE: LazyLock<Regex> =
+    LazyLock::new(|| Regex::new(r"^ *type \w+ (struct|interface)$").unwrap());
+
+/// Extract Go components: type struct/interface headers and func signatures.
+pub fn extract(content: &str) -> ExtractResult {
+    let tree = parse(content, &tree_sitter_go::LANGUAGE.into())?;
+    let mut components: Vec<String> = Vec::new();
+    visit(tree.root_node(), content, &mut components);
+    Ok(components)
+}
+
+fn line_start(content: &str, byte: usize) -> usize {
+    content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
+}
+
+fn visit(node: Node<'_>, content: &str, out: &mut Vec<String>) {
+    let mut cursor = node.walk();
+    for child in node.named_children(&mut cursor) {
+        match child.kind() {
+            "type_declaration" => {
+                let mut tcursor = child.walk();
+                for spec in child.named_children(&mut tcursor) {
+                    if spec.kind() == "type_spec" {
+                        emit_type_spec(spec, content, out);
+                    }
+                }
+            }
+            "function_declaration" | "method_declaration" => {
+                emit_func(child, content, out);
+                if let Some(body) = child.child_by_field_name("body") {
+                    visit(body, content, out);
+                }
+            }
+            _ => visit(child, content, out),
+        }
+    }
+}
+
+/// `type Name struct` / `type Name interface` when the body brace follows.
+fn emit_type_spec(spec: Node<'_>, content: &str, out: &mut Vec<String>) {
+    let Some(type_node) = spec.child_by_field_name("type") else {
+        return;
+    };
+    let keyword_len = match type_node.kind() {
+        "struct_type" => "struct".len(),
+        "interface_type" => "interface".len(),
+        _ => return,
+    };
+    // slice from the `type` keyword line through the struct/interface keyword
+    let decl_start = spec
+        .parent()
+        .map(|p| p.start_byte())
+        .unwrap_or_else(|| spec.start_byte());
+    let start = line_start(content, decl_start);
+    let end = type_node.start_byte() + keyword_len;
+    let header = &content[start..end];
+    if !TYPE_HEADER_RE.is_match(header) {
+        return; // legacy pattern: single-space, no generics, no aliases
+    }
+    // legacy lookahead `(?=\s*\{)`
+    if !content[end..].trim_start().starts_with('{') {
+        return;
+    }
+    out.push(header.to_string());
+}
+
+/// Column-0 `func` signature up to (not including) the body brace.
+fn emit_func(node: Node<'_>, content: &str, out: &mut Vec<String>) {
+    let start = node.start_byte();
+    // legacy `^func`: column 0 only
+    if line_start(content, start) != start {
+        return;
+    }
+    let Some(body) = node.child_by_field_name("body") else {
+        return;
+    };
+    // legacy lookahead `(?=\s\{\s)`: whitespace on both sides of the brace
+    let brace = body.start_byte();
+    let before_ok = content[..brace]
+        .chars()
+        .next_back()
+        .is_some_and(char::is_whitespace);
+    let after_ok = content[brace + 1..]
+        .chars()
+        .next()
+        .is_some_and(char::is_whitespace);
+    if !before_ok || !after_ok {
+        return;
+    }
+    let component = content[start..brace].trim_end();
+    out.push(component.to_string());
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn types_and_funcs() {
+        let content = "type Greeting struct {\n\tmessage string\n}\n\nfunc (g Greeting) sayHello() {\n\tfmt.Println(g.message)\n}\n\nfunc createGreeting(m string) Greeting {\n\treturn Greeting{message: m}\n}\n";
+        assert_eq!(
+            extract(content).unwrap(),
+            vec![
+                "type Greeting struct",
+                "func (g Greeting) sayHello()",
+                "func createGreeting(m string) Greeting",
+            ]
+        );
+    }
+
+    #[test]
+    fn multiline_signature_keeps_tabs() {
+        let content = "func WithAReasonableName(\n\tctx context.Context,\n\tparam1 string,\n) (resultType, error) {\n\treturn resultType{}, nil\n}\n";
+        assert_eq!(
+            extract(content).unwrap(),
+            vec!["func WithAReasonableName(\n\tctx context.Context,\n\tparam1 string,\n) (resultType, error)"]
+        );
+    }
+
+    #[test]
+    fn legacy_quirks_preserved() {
+        // empty body `{}`: the brace is not whitespace-surrounded -> skipped
+        assert!(extract("func f() {}\n").unwrap().is_empty());
+        // generic types don't match the legacy header pattern
+        assert!(extract("type Box[T any] struct {\n\tv T\n}\n")
+            .unwrap()
+            .is_empty());
+        // interfaces are headers too
+        assert_eq!(
+            extract("type Animal interface {\n\tSpeak() string\n}\n").unwrap(),
+            vec!["type Animal interface"]
+        );
+    }
+}
diff --git a/crates/tree_plus_core/src/extract/treesitter/mod.rs b/crates/tree_plus_core/src/extract/treesitter/mod.rs
index 652ad05..feaefe7 100644
--- a/crates/tree_plus_core/src/extract/treesitter/mod.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/mod.rs
@@ -5,6 +5,7 @@
 //! regex-based components; golden parity tests in tests/ enforce this.
 
 pub mod c_cpp;
+pub mod go;
 pub mod python;
 pub mod rust;
 pub mod typescript;
diff --git a/crates/tree_plus_core/tests/golden_parity.rs b/crates/tree_plus_core/tests/golden_parity.rs
index 468aac0..8ec8f33 100644
--- a/crates/tree_plus_core/tests/golden_parity.rs
+++ b/crates/tree_plus_core/tests/golden_parity.rs
@@ -29,6 +29,7 @@ fn in_v1_scope(path: &str) -> bool {
         ".py",
         ".pyi",
         ".rs",
+        ".go",
         ".js",
         ".jsx",
         ".ts",
diff --git a/docs/language-roadmap.md b/docs/language-roadmap.md
index 5f2cb7a..a571141 100644
--- a/docs/language-roadmap.md
+++ b/docs/language-roadmap.md
@@ -1,9 +1,10 @@
 # Language Roadmap (Rust Port)
 
-Version-1 implements: Rust, Python, JavaScript, TypeScript, C, C++, Markdown
-(+ RST), JSON (package.json / schema / RPC / OpenRPC), JSONL, YAML, TOML
-(Cargo/pyproject), CSV, Makefile/Justfile, .env, requirements.txt, SQLite,
-and TODO/BUG/NOTE markers everywhere except `.md`/`.txt` (legacy rule).
+Version-1 implements: Rust, Python, JavaScript, TypeScript, C, C++, Go,
+Markdown (+ RST), JSON (package.json / schema / RPC / OpenRPC), JSONL, YAML,
+TOML (Cargo/pyproject), CSV, Makefile/Justfile, .env, requirements.txt,
+SQLite, and TODO/BUG/NOTE markers everywhere except `.md`/`.txt` (legacy
+rule).
 
 Everything below is recognized by the legacy Python implementation and
 **deferred**: files keep counts and markers, but emit no components. Legacy
@@ -14,7 +15,6 @@ availability → suggested path → missing tests.
 
 | Language | Extensions | Legacy extractor | TS grammar? | Suggested path | Missing tests |
 |---|---|---|---|---|---|
-| Go | .go | parse_go | yes (tree-sitter-go) | tree-sitter formatter | port golden `go_test.go` |
 | Java | .java | parse_java | yes | tree-sitter formatter | port golden `JavaTest.java` |
 | Kotlin | .kt | parse_kt | yes (community) | tree-sitter formatter | port golden `KotlinTest.kt` |
 | Swift | .swift | parse_swift | yes (community) | tree-sitter formatter | port golden `swift_test.swift` |
@@ -59,6 +59,6 @@ highlighting, tiktoken tokenizers — see docs/rust-port-differences.md.
 
 ## Suggested order of attack
 
-1. Go, Java, Kotlin, C#, Ruby, Bash (mature grammars, heavily used).
+1. Java, Kotlin, C#, Ruby, Bash (mature grammars, heavily used).
 2. SQL/GraphQL/Protobuf/requirements-style line formats (cheap regex ports).
 3. The long tail, prioritized by user demand.
diff --git a/tests/golden/generate_legacy_goldens.py b/tests/golden/generate_legacy_goldens.py
index d50bd79..6976136 100644
--- a/tests/golden/generate_legacy_goldens.py
+++ b/tests/golden/generate_legacy_goldens.py
@@ -50,7 +50,7 @@ def sanitize(p: Path) -> str:
 
 
 DEFERRED_PARSERS = [
-    "parse_php", "parse_kt", "parse_swift", "parse_go", "parse_bash",
+    "parse_php", "parse_kt", "parse_swift", "parse_bash",
     "parse_ps1", "parse_zig", "parse_rb", "parse_sql", "parse_graphql",
     "parse_cs", "parse_jl", "parse_scala", "parse_java", "parse_perl",
     "parse_hs", "parse_fsharp", "parse_lisp", "parse_erl", "parse_capnp",
diff --git a/tests/golden/legacy/trees/repo_concise.txt b/tests/golden/legacy/trees/repo_concise.txt
index 17182cb..e2786e5 100644
--- a/tests/golden/legacy/trees/repo_concise.txt
+++ b/tests/golden/legacy/trees/repo_concise.txt
@@ -1,42 +1,44 @@
-📁 tree_plus (44 folders, 439 files)
+📁 tree_plus (44 folders, 443 files)
 ├── 📄 .env.test (4 tokens, 0 lines)
-├── 📁 .github (2 folders, 3 files)
+├── 📁 .github (2 folders, 4 files)
 │   ├── 📄 dependabot.yml (128 tokens, 11 lines)
-│   └── 📁 workflows (1 folder, 2 files)
-│       ├── 📄 microsoft.yml (284 tokens, 40 lines)
-│       └── 📄 unix.yml (713 tokens, 92 lines)
-├── 📄 .gitignore (219 tokens, 57 lines)
+│   └── 📁 workflows (1 folder, 3 files)
+│       ├── 📄 microsoft.yml (361 tokens, 54 lines)
+│       ├── 📄 rust.yml (292 tokens, 44 lines)
+│       └── 📄 unix.yml (790 tokens, 106 lines)
+├── 📄 .gitignore (226 tokens, 60 lines)
 ├── 📄 .mcp_server.pid (2 tokens, 1 line)
-├── 📄 Cargo.toml (206 tokens, 29 lines)
+├── 📄 Cargo.toml (212 tokens, 30 lines)
 ├── 📄 claude-fable-5-rust-rewrite-goal.md (3,394 tokens, 434 lines)
 ├── 📁 coverage (1 folder, 1 file)
 │   └── 📄 lcov.info (17,359 tokens, 2,180 lines)
-├── 📁 crates (11 folders, 27 files)
+├── 📁 crates (11 folders, 28 files)
 │   ├── 📁 tree_plus_cli (3 folders, 3 files)
-│   │   ├── 📄 Cargo.toml (86 tokens, 15 lines)
+│   │   ├── 📄 Cargo.toml (120 tokens, 21 lines)
 │   │   ├── 📁 src (1 folder, 1 file)
-│   │   │   └── 📄 main.rs (1,332 tokens, 174 lines)
+│   │   │   └── 📄 main.rs (1,339 tokens, 173 lines)
 │   │   └── 📁 tests (1 folder, 1 file)
 │   │       └── 📄 cli.rs (701 tokens, 92 lines)
-│   └── 📁 tree_plus_core (7 folders, 24 files)
+│   └── 📁 tree_plus_core (7 folders, 25 files)
 │       ├── 📁 benches (1 folder, 1 file)
 │       │   └── 📄 tree_plus_bench.rs (608 tokens, 78 lines)
-│       ├── 📄 Cargo.toml (228 tokens, 36 lines)
+│       ├── 📄 Cargo.toml (236 tokens, 37 lines)
 │       ├── 📁 examples (1 folder, 2 files)
 │       │   ├── 📄 dump_ast.rs (516 tokens, 55 lines)
 │       │   └── 📄 extract.rs (129 tokens, 16 lines)
-│       ├── 📁 src (3 folders, 18 files)
+│       ├── 📁 src (3 folders, 19 files)
 │       │   ├── 📄 config.rs (304 tokens, 39 lines)
 │       │   ├── 📄 count.rs (1,346 tokens, 203 lines)
-│       │   ├── 📁 extract (2 folders, 10 files)
+│       │   ├── 📁 extract (2 folders, 11 files)
 │       │   │   ├── 📄 data.rs (5,115 tokens, 582 lines)
 │       │   │   ├── 📄 markdown.rs (1,531 tokens, 180 lines)
 │       │   │   ├── 📄 markers.rs (438 tokens, 60 lines)
-│       │   │   ├── 📄 mod.rs (2,520 tokens, 277 lines)
+│       │   │   ├── 📄 mod.rs (2,532 tokens, 278 lines)
 │       │   │   ├── 📄 simple.rs (1,629 tokens, 216 lines)
-│       │   │   └── 📁 treesitter (1 folder, 5 files)
+│       │   │   └── 📁 treesitter (1 folder, 6 files)
 │       │   │       ├── 📄 c_cpp.rs (5,979 tokens, 591 lines)
-│       │   │       ├── 📄 mod.rs (488 tokens, 66 lines)
+│       │   │       ├── 📄 go.rs (1,364 tokens, 152 lines)
+│       │   │       ├── 📄 mod.rs (491 tokens, 67 lines)
 │       │   │       ├── 📄 python.rs (3,487 tokens, 346 lines)
 │       │   │       ├── 📄 rust.rs (2,785 tokens, 312 lines)
 │       │   │       └── 📄 typescript.rs (3,897 tokens, 420 lines)
@@ -47,20 +49,21 @@
 │       │   ├── 📄 sort.rs (1,693 tokens, 214 lines)
 │       │   └── 📄 walk.rs (2,245 tokens, 260 lines)
 │       └── 📁 tests (1 folder, 2 files)
-│           ├── 📄 golden_parity.rs (1,809 tokens, 243 lines)
+│           ├── 📄 golden_parity.rs (1,812 tokens, 244 lines)
 │           └── 📄 robustness.rs (743 tokens, 86 lines)
 ├── 📁 docs (1 folder, 4 files)
 │   ├── 📄 architecture.md (1,392 tokens, 113 lines)
 │   ├── 📄 language-roadmap.md (1,262 tokens, 64 lines)
 │   ├── 📄 performance.md (690 tokens, 64 lines)
-│   └── 📄 rust-port-differences.md (922 tokens, 67 lines)
+│   └── 📄 rust-port-differences.md (1,022 tokens, 78 lines)
+├── 📄 Justfile (141 tokens, 23 lines)
 ├── 📄 LICENSE (2,744 tokens, 81 lines)
 ├── 📄 Makefile (801 tokens, 121 lines)
 ├── 📄 nodemon.json (112 tokens, 24 lines)
 ├── 📄 pyproject.toml (366 tokens, 51 lines)
 ├── 📄 pytest.ini (20 tokens, 4 lines)
 ├── 📄 README.md (99,851 tokens, 3,708 lines)
-├── 📁 tests (25 folders, 378 files)
+├── 📁 tests (25 folders, 379 files)
 │   ├── 📄 .env.test (4 tokens, 0 lines)
 │   ├── 📄 build_absurdly_huge_jsonl.py (506 tokens, 65 lines)
 │   ├── 📁 dot_dot (2 folders, 4 files)
@@ -73,10 +76,10 @@
 │   │   └── 📁 is_empty (1 folder, 0 files)
 │   ├── 📁 folder_with_evil_logging (1 folder, 1 file)
 │   │   └── 📄 logging.py (11 tokens, 0 lines)
-│   ├── 📁 golden (6 folders, 246 files)
+│   ├── 📁 golden (6 folders, 247 files)
 │   │   ├── 📄 diff_components.py (417 tokens, 59 lines)
-│   │   ├── 📄 generate_legacy_goldens.py (1,499 tokens, 156 lines)
-│   │   └── 📁 legacy (5 folders, 244 files)
+│   │   ├── 📄 generate_legacy_goldens.py (1,496 tokens, 156 lines)
+│   │   └── 📁 legacy (5 folders, 245 files)
 │   │       ├── 📁 components (1 folder, 109 files)
 │   │       │   ├── 📄 tests__dot_dot__my_test_file.py.json (21 tokens, 5 lines)
 │   │       │   ├── 📄 tests__dot_dot__nested_dir__.env.test.json (22 tokens, 5
@@ -513,7 +516,7 @@
 │   │       │   ├── 📄 tests__path_to_test__file.py.json (19 tokens, 0 lines)
 │   │       │   ├── 📄 tests__path_to_test__file.txt.json (20 tokens, 0 lines)
 │   │       │   └── 📄 tests__path_to_test__version.py.json (20 tokens, 0 lines)
-│   │       ├── 📁 trees (1 folder, 13 files)
+│   │       ├── 📁 trees (1 folder, 14 files)
 │   │       │   ├── 📄 dot_dot.txt (99 tokens, 10 lines)
 │   │       │   ├── 📄 more_languages.txt (22,977 tokens, 2,225 lines)
 │   │       │   ├── 📄 more_languages_group1.txt (2,909 tokens, 374 lines)
@@ -526,12 +529,13 @@
 │   │       │   ├── 📄 more_languages_group_lisp.txt (155 tokens, 22 lines)
 │   │       │   ├── 📄 more_languages_group_todo.txt (1,491 tokens, 111 lines)
 │   │       │   ├── 📄 multi_seed.txt (3,789 tokens, 427 lines)
-│   │       │   └── 📄 path_to_test.txt (445 tokens, 51 lines)
+│   │       │   ├── 📄 path_to_test.txt (445 tokens, 51 lines)
+│   │       │   └── 📄 repo_concise.txt (9,776 tokens, 699 lines)
 │   │       └── 📁 trees_v1 (1 folder, 13 files)
 │   │           ├── 📄 dot_dot.txt (99 tokens, 10 lines)
-│   │           ├── 📄 more_languages.txt (10,782 tokens, 1,122 lines)
+│   │           ├── 📄 more_languages.txt (10,912 tokens, 1,135 lines)
 │   │           ├── 📄 more_languages_group1.txt (1,130 tokens, 146 lines)
-│   │           ├── 📄 more_languages_group2.txt (415 tokens, 54 lines)
+│   │           ├── 📄 more_languages_group2.txt (531 tokens, 67 lines)
 │   │           ├── 📄 more_languages_group3.txt (1,240 tokens, 149 lines)
 │   │           ├── 📄 more_languages_group4.txt (1,020 tokens, 123 lines)
 │   │           ├── 📄 more_languages_group5.txt (2,099 tokens, 235 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages.txt b/tests/golden/legacy/trees_v1/more_languages.txt
index a1de0a9..897f9fb 100644
--- a/tests/golden/legacy/trees_v1/more_languages.txt
+++ b/tests/golden/legacy/trees_v1/more_languages.txt
@@ -190,6 +190,19 @@
 │   │   ├──     struct cliSSLconfig sslconfig;
 │   │   └── } config;
 │   ├── 📄 go_test.go (179 tokens, 46 lines)
+│   │   ├── type Greeting struct
+│   │   ├── func (g Greeting) sayHello()
+│   │   ├── func createGreeting(m string) Greeting
+│   │   ├── type SomethingLong struct
+│   │   ├── func (s *SomethingLong) WithAReasonableName(
+│   │   │           ctx context.Context,
+│   │   │           param1 string,
+│   │   │           param2 int,
+│   │   │           param3 map[string]interface{},
+│   │   │           callback func(int) error,
+│   │   │   ) (resultType, error)
+│   │   ├── type resultType struct
+│   │   └── func main()
 │   ├── 📄 PerlTest.pl (63 tokens, 20 lines)
 │   ├── 📄 PhpTest.php (70 tokens, 19 lines)
 │   ├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)
diff --git a/tests/golden/legacy/trees_v1/more_languages_group2.txt b/tests/golden/legacy/trees_v1/more_languages_group2.txt
index 854ea5d..ffbf591 100644
--- a/tests/golden/legacy/trees_v1/more_languages_group2.txt
+++ b/tests/golden/legacy/trees_v1/more_languages_group2.txt
@@ -42,6 +42,19 @@
 │   ├──     struct cliSSLconfig sslconfig;
 │   └── } config;
 ├── 📄 go_test.go (179 tokens, 46 lines)
+│   ├── type Greeting struct
+│   ├── func (g Greeting) sayHello()
+│   ├── func createGreeting(m string) Greeting
+│   ├── type SomethingLong struct
+│   ├── func (s *SomethingLong) WithAReasonableName(
+│   │           ctx context.Context,
+│   │           param1 string,
+│   │           param2 int,
+│   │           param3 map[string]interface{},
+│   │           callback func(int) error,
+│   │   ) (resultType, error)
+│   ├── type resultType struct
+│   └── func main()
 ├── 📄 PerlTest.pl (63 tokens, 20 lines)
 ├── 📄 PhpTest.php (70 tokens, 19 lines)
 ├── 📄 PowershellTest.ps1 (459 tokens, 89 lines)

From 0cc4d789f76b217f3a32e9bea692b2409ddae8b9 Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 15:55:38 -0400
Subject: [PATCH 6/8] fix: collapse match guards for stable clippy; park goal
 prompt in detritus

CI runs stable clippy, which flags collapsible_match patterns the local
nightly toolchain accepted.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 .../src/extract/treesitter/c_cpp.rs           | 38 +++++++++----------
 1 file changed, 17 insertions(+), 21 deletions(-)

diff --git a/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs b/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
index 58a4839..e7ea179 100644
--- a/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
@@ -220,17 +220,15 @@ impl<'a> CExtractor<'a> {
             return;
         };
         match type_node.kind() {
-            "struct_specifier" | "union_specifier" | "class_specifier" => {
-                if type_node.child_by_field_name("body").is_some() {
-                    self.emit_record(type_node, Some(node.start_byte()));
-                    self.emit_closer(type_node, node);
-                }
+            "struct_specifier" | "union_specifier" | "class_specifier"
+                if type_node.child_by_field_name("body").is_some() =>
+            {
+                self.emit_record(type_node, Some(node.start_byte()));
+                self.emit_closer(type_node, node);
             }
-            "enum_specifier" => {
-                if type_node.child_by_field_name("body").is_some() {
-                    self.emit_enum(type_node);
-                    self.emit_closer(type_node, node);
-                }
+            "enum_specifier" if type_node.child_by_field_name("body").is_some() => {
+                self.emit_enum(type_node);
+                self.emit_closer(type_node, node);
             }
             _ => {}
         }
@@ -256,18 +254,16 @@ impl<'a> CExtractor<'a> {
             }
         }
         match type_node.kind() {
-            "struct_specifier" | "union_specifier" | "class_specifier" => {
-                if type_node.child_by_field_name("body").is_some() {
-                    let start = line_start(self.content, node.start_byte());
-                    self.emit_record(type_node, Some(start));
-                    self.emit_closer(type_node, node);
-                }
+            "struct_specifier" | "union_specifier" | "class_specifier"
+                if type_node.child_by_field_name("body").is_some() =>
+            {
+                let start = line_start(self.content, node.start_byte());
+                self.emit_record(type_node, Some(start));
+                self.emit_closer(type_node, node);
             }
-            "enum_specifier" => {
-                if type_node.child_by_field_name("body").is_some() {
-                    self.emit_enum(type_node);
-                    self.emit_closer(type_node, node);
-                }
+            "enum_specifier" if type_node.child_by_field_name("body").is_some() => {
+                self.emit_enum(type_node);
+                self.emit_closer(type_node, node);
             }
             _ => {}
         }

From 6d99a73135b75b4b0bb1434b718002d0a55edf30 Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 18:15:06 -0400
Subject: [PATCH 7/8] fix: stack overflows on huge trees (torvalds/linux)

Two distinct overflows surfaced by running tprs on the Linux kernel:

1. Tree-sitter extractor visitors (C/C++, Rust, Python, TypeScript, Go)
   recursed on AST depth. arch/x86/kernel/cpu/microcode/intel-ucode-defs.h
   is a headerless initializer-list fragment that parses as a deeply
   nested ERROR tree and blew rayon's 2 MiB worker stacks. All visitors
   now traverse with explicit heap stacks; emission order is unchanged
   (golden parity suite still passes byte-for-byte). The C/C++ extractor
   threads deferred work (record closers, template field suppression,
   ERROR salvage) through a Work queue to preserve legacy ordering.

2. Rayon work-stealing nests from_folder frames on a worker while it
   waits in join, leaving too little headroom for tree-sitter's C frames
   on big trees (drivers/ segfaulted in ts_subtree_retain). from_seeds
   now runs in a dedicated pool with 16 MiB worker stacks.

Adds a regression test extracting deeply nested inputs for all six
tree-sitter languages on a 512 KiB thread, and records Linux-kernel
benchmarks in docs/performance.md (101k files: concise 0.50 s,
full 12.4 s; linux/kernel subdir ~56x faster than the Python CLI).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 .../src/extract/treesitter/c_cpp.rs           | 148 ++++++++++++------
 .../src/extract/treesitter/go.rs              |  26 +--
 .../src/extract/treesitter/python.rs          |  63 ++++----
 .../src/extract/treesitter/rust.rs            |  45 +++---
 .../src/extract/treesitter/typescript.rs      |  43 +++--
 crates/tree_plus_core/src/walk.rs             |  18 +++
 crates/tree_plus_core/tests/robustness.rs     |  57 +++++++
 docs/performance.md                           |  32 +++-
 8 files changed, 304 insertions(+), 128 deletions(-)

diff --git a/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs b/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
index e7ea179..f042c33 100644
--- a/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/c_cpp.rs
@@ -46,7 +46,7 @@ pub fn extract(content: &str, ext: &str) -> ExtractResult {
         components: Vec::new(),
         suppress_plain_fields: 0,
     };
-    extractor.visit(tree.root_node());
+    extractor.run(tree.root_node());
     Ok(extractor.components)
 }
 
@@ -70,6 +70,25 @@ fn line_start(content: &str, byte: usize) -> usize {
     content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
 }
 
+/// Deferred traversal work. AST depth is input-controlled (deeply nested
+/// ERROR trees from headerless initializer fragments overflow the small
+/// rayon worker stacks), so descents go through an explicit heap stack
+/// instead of recursion. Ordering is LIFO: push follow-up work (closers,
+/// suppression ends) before the subtree visit that must precede it.
+enum Work<'a> {
+    Visit(Node<'a>),
+    Closer { type_node: Node<'a>, decl: Node<'a> },
+    SalvageFnDecl(Node<'a>),
+    EndSuppress,
+}
+
+/// Push `node`'s named children so they pop in source order.
+fn push_children_rev<'a>(node: Node<'a>, stack: &mut Vec<Work<'a>>) {
+    let mut cursor = node.walk();
+    let children: Vec<Node<'a>> = node.named_children(&mut cursor).collect();
+    stack.extend(children.into_iter().rev().map(Work::Visit));
+}
+
 impl<'a> CExtractor<'a> {
     fn slice_clean(&self, from: usize, to: usize) -> String {
         // legacy parse_c ran on comment-stripped content
@@ -79,12 +98,24 @@ impl<'a> CExtractor<'a> {
             .to_string()
     }
 
-    fn visit(&mut self, node: Node<'a>) {
+    fn run(&mut self, root: Node<'a>) {
+        let mut stack = vec![Work::Visit(root)];
+        while let Some(work) = stack.pop() {
+            match work {
+                Work::Visit(node) => self.visit(node, &mut stack),
+                Work::Closer { type_node, decl } => self.emit_closer(type_node, decl),
+                Work::SalvageFnDecl(child) => self.salvage_fn_declarator(child),
+                Work::EndSuppress => self.suppress_plain_fields -= 1,
+            }
+        }
+    }
+
+    fn visit(&mut self, node: Node<'a>, stack: &mut Vec<Work<'a>>) {
         match node.kind() {
             "function_definition" => {
                 if let Some(body) = node.child_by_field_name("body") {
                     self.emit_function_definition(node, body);
-                    self.visit(body);
+                    stack.push(Work::Visit(body));
                 } else if let Some(declarator) = find_function_declarator(node) {
                     // `= default;` / `= delete;` members have no body node
                     let start = line_start(self.content, node.start_byte());
@@ -96,22 +127,22 @@ impl<'a> CExtractor<'a> {
                 }
             }
             "ERROR" => {
-                self.salvage_error_region(node);
+                self.salvage_error_region(node, stack);
             }
             "struct_specifier" | "class_specifier" | "union_specifier" => {
-                self.emit_record(node, None);
+                self.emit_record(node, None, stack);
             }
             "enum_specifier" => {
                 self.emit_enum(node);
             }
             "type_definition" => {
-                self.emit_type_definition(node);
+                self.emit_type_definition(node, stack);
             }
             "declaration" => {
-                self.emit_declaration(node);
+                self.emit_declaration(node, stack);
             }
             "field_declaration" => {
-                self.emit_field(node);
+                self.emit_field(node, stack);
             }
             "access_specifier" => {
                 // includes the trailing ':' in the slice
@@ -136,7 +167,7 @@ impl<'a> CExtractor<'a> {
                 }
             }
             "template_declaration" => {
-                self.emit_template(node);
+                self.emit_template(node, stack);
             }
             "expression_statement" => {
                 // legacy pybind noise: ` *\w+.def("...` captured to first `;`
@@ -153,30 +184,26 @@ impl<'a> CExtractor<'a> {
                         }
                     }
                 }
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    self.visit(child);
-                }
+                push_children_rev(node, stack);
             }
             "while_statement" | "for_statement" | "if_statement" | "switch_statement" => {
                 self.maybe_emit_control_noise(node);
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    self.visit(child);
-                }
+                push_children_rev(node, stack);
             }
             _ => {
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    self.visit(child);
-                }
+                push_children_rev(node, stack);
             }
         }
     }
 
     /// struct/class/union with a body: label + fields (+ optional closer).
     /// `prefix_start` overrides the slice start (e.g. `typedef `/`static `).
-    fn emit_record(&mut self, node: Node<'a>, prefix_start: Option<usize>) {
+    fn emit_record(
+        &mut self,
+        node: Node<'a>,
+        prefix_start: Option<usize>,
+        stack: &mut Vec<Work<'a>>,
+    ) {
         let Some(body) = node.child_by_field_name("body") else {
             return; // bare type references are not components
         };
@@ -185,7 +212,7 @@ impl<'a> CExtractor<'a> {
         if !label.is_empty() {
             self.components.push(label);
         }
-        self.visit(body);
+        stack.push(Work::Visit(body));
     }
 
     fn emit_enum(&mut self, node: Node<'a>) {
@@ -215,7 +242,7 @@ impl<'a> CExtractor<'a> {
     }
 
     /// `typedef struct {...} Name;` -> "typedef struct", fields, "} Name;".
-    fn emit_type_definition(&mut self, node: Node<'a>) {
+    fn emit_type_definition(&mut self, node: Node<'a>, stack: &mut Vec<Work<'a>>) {
         let Some(type_node) = node.child_by_field_name("type") else {
             return;
         };
@@ -223,8 +250,12 @@ impl<'a> CExtractor<'a> {
             "struct_specifier" | "union_specifier" | "class_specifier"
                 if type_node.child_by_field_name("body").is_some() =>
             {
-                self.emit_record(type_node, Some(node.start_byte()));
-                self.emit_closer(type_node, node);
+                // closer pops after the record's deferred body visit
+                stack.push(Work::Closer {
+                    type_node,
+                    decl: node,
+                });
+                self.emit_record(type_node, Some(node.start_byte()), stack);
             }
             "enum_specifier" if type_node.child_by_field_name("body").is_some() => {
                 self.emit_enum(type_node);
@@ -235,7 +266,7 @@ impl<'a> CExtractor<'a> {
     }
 
     /// `static struct config {...} config;` and similar declarations.
-    fn emit_declaration(&mut self, node: Node<'a>) {
+    fn emit_declaration(&mut self, node: Node<'a>, stack: &mut Vec<Work<'a>>) {
         let Some(type_node) = node.child_by_field_name("type") else {
             return;
         };
@@ -258,8 +289,11 @@ impl<'a> CExtractor<'a> {
                 if type_node.child_by_field_name("body").is_some() =>
             {
                 let start = line_start(self.content, node.start_byte());
-                self.emit_record(type_node, Some(start));
-                self.emit_closer(type_node, node);
+                stack.push(Work::Closer {
+                    type_node,
+                    decl: node,
+                });
+                self.emit_record(type_node, Some(start), stack);
             }
             "enum_specifier" if type_node.child_by_field_name("body").is_some() => {
                 self.emit_enum(type_node);
@@ -329,7 +363,7 @@ impl<'a> CExtractor<'a> {
         Some(text.trim_end_matches(';').trim_end().to_string())
     }
 
-    fn emit_field(&mut self, node: Node<'a>) {
+    fn emit_field(&mut self, node: Node<'a>, stack: &mut Vec<Work<'a>>) {
         // legacy field patterns require an indented member starting its line
         let fstart = line_start(self.content, node.start_byte());
         let indent = &self.content[fstart..node.start_byte()];
@@ -343,8 +377,12 @@ impl<'a> CExtractor<'a> {
                 "struct_specifier" | "union_specifier" | "enum_specifier" | "class_specifier"
             ) && type_node.child_by_field_name("body").is_some()
             {
-                self.visit(type_node);
-                self.emit_closer(type_node, node); // e.g. `} inner;`
+                // closer (e.g. `} inner;`) pops after the nested record
+                stack.push(Work::Closer {
+                    type_node,
+                    decl: node,
+                });
+                stack.push(Work::Visit(type_node));
                 return;
             }
         }
@@ -378,7 +416,7 @@ impl<'a> CExtractor<'a> {
 
     /// Partial recovery inside ERROR regions: invalid syntax must still
     /// yield the obvious components instead of nothing.
-    fn salvage_error_region(&mut self, node: Node<'a>) {
+    fn salvage_error_region(&mut self, node: Node<'a>, stack: &mut Vec<Work<'a>>) {
         static RECORD_HEADER_RE: LazyLock<Regex> =
             LazyLock::new(|| Regex::new(r"^[ \t]*((?:class|struct) \w+[^\n{;]*)").unwrap());
         let text = &self.content[node.byte_range()];
@@ -387,21 +425,30 @@ impl<'a> CExtractor<'a> {
                 self.components.push(m.as_str().trim_end().to_string());
             }
         }
-        // ordered pass: salvage loose function declarators, recurse the rest
+        // ordered pass: salvage loose function declarators, descend the rest
         let mut cursor = node.walk();
-        for child in node.named_children(&mut cursor) {
-            if child.kind() == "function_declarator" {
-                let start = line_start(self.content, child.start_byte());
-                let candidate = self.slice_clean(start, child.end_byte());
-                if FUNCTION_FORM_RE.is_match(&candidate) {
-                    let after = &self.content[child.end_byte()..];
-                    let mut chars = after.chars();
-                    if chars.next().is_some_and(char::is_whitespace) && chars.next() == Some('{') {
-                        self.components.push(candidate);
-                    }
+        let work: Vec<Work<'a>> = node
+            .named_children(&mut cursor)
+            .map(|child| {
+                if child.kind() == "function_declarator" {
+                    Work::SalvageFnDecl(child)
+                } else {
+                    Work::Visit(child)
                 }
-            } else {
-                self.visit(child);
+            })
+            .collect();
+        stack.extend(work.into_iter().rev());
+    }
+
+    /// A loose `function_declarator` found inside an ERROR region.
+    fn salvage_fn_declarator(&mut self, child: Node<'a>) {
+        let start = line_start(self.content, child.start_byte());
+        let candidate = self.slice_clean(start, child.end_byte());
+        if FUNCTION_FORM_RE.is_match(&candidate) {
+            let after = &self.content[child.end_byte()..];
+            let mut chars = after.chars();
+            if chars.next().is_some_and(char::is_whitespace) && chars.next() == Some('{') {
+                self.components.push(candidate);
             }
         }
     }
@@ -454,7 +501,7 @@ impl<'a> CExtractor<'a> {
             .push(self.slice_clean(start, cond.end_byte()).replace('\t', "  "));
     }
 
-    fn emit_template(&mut self, node: Node<'a>) {
+    fn emit_template(&mut self, node: Node<'a>, stack: &mut Vec<Work<'a>>) {
         // slice from `template` through the inner declaration's signature
         let start = line_start(self.content, node.start_byte());
         let mut cursor = node.walk();
@@ -464,7 +511,7 @@ impl<'a> CExtractor<'a> {
                     if let Some(body) = child.child_by_field_name("body") {
                         let text = self.slice_clean(start, body.start_byte());
                         self.components.push(text);
-                        self.visit(body);
+                        stack.push(Work::Visit(body));
                     }
                     return;
                 }
@@ -483,9 +530,10 @@ impl<'a> CExtractor<'a> {
                     if let Some(body) = child.child_by_field_name("body") {
                         let text = self.slice_clean(start, body.start_byte());
                         self.components.push(text);
+                        // suppression spans exactly the deferred body visit
                         self.suppress_plain_fields += 1;
-                        self.visit(body);
-                        self.suppress_plain_fields -= 1;
+                        stack.push(Work::EndSuppress);
+                        stack.push(Work::Visit(body));
                     } else {
                         let text = self.slice_clean(start, child.end_byte());
                         self.components.push(text);
diff --git a/crates/tree_plus_core/src/extract/treesitter/go.rs b/crates/tree_plus_core/src/extract/treesitter/go.rs
index 9efde45..8fe9bc8 100644
--- a/crates/tree_plus_core/src/extract/treesitter/go.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/go.rs
@@ -31,25 +31,31 @@ fn line_start(content: &str, byte: usize) -> usize {
     content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
 }
 
-fn visit(node: Node<'_>, content: &str, out: &mut Vec<String>) {
-    let mut cursor = node.walk();
-    for child in node.named_children(&mut cursor) {
-        match child.kind() {
+/// Depth-first via an explicit stack: AST depth is input-controlled, and
+/// deep expression nesting must not overflow small worker-thread stacks.
+fn visit(root: Node<'_>, content: &str, out: &mut Vec<String>) {
+    let mut stack = vec![root];
+    while let Some(node) = stack.pop() {
+        match node.kind() {
             "type_declaration" => {
-                let mut tcursor = child.walk();
-                for spec in child.named_children(&mut tcursor) {
+                let mut tcursor = node.walk();
+                for spec in node.named_children(&mut tcursor) {
                     if spec.kind() == "type_spec" {
                         emit_type_spec(spec, content, out);
                     }
                 }
             }
             "function_declaration" | "method_declaration" => {
-                emit_func(child, content, out);
-                if let Some(body) = child.child_by_field_name("body") {
-                    visit(body, content, out);
+                emit_func(node, content, out);
+                if let Some(body) = node.child_by_field_name("body") {
+                    stack.push(body);
                 }
             }
-            _ => visit(child, content, out),
+            _ => {
+                let mut cursor = node.walk();
+                let children: Vec<Node<'_>> = node.named_children(&mut cursor).collect();
+                stack.extend(children.into_iter().rev());
+            }
         }
     }
 }
diff --git a/crates/tree_plus_core/src/extract/treesitter/python.rs b/crates/tree_plus_core/src/extract/treesitter/python.rs
index 57c39fe..9416412 100644
--- a/crates/tree_plus_core/src/extract/treesitter/python.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/python.rs
@@ -57,7 +57,7 @@ pub fn extract(content: &str) -> ExtractResult {
         pending_decorators: Vec::new(),
         in_class: false,
     };
-    extractor.visit(tree.root_node());
+    extractor.run(tree.root_node());
     Ok(extractor.components)
 }
 
@@ -99,33 +99,40 @@ impl<'a> PyExtractor<'a> {
         }
     }
 
-    fn visit(&mut self, node: Node<'a>) {
-        match node.kind() {
-            "decorated_definition" => {
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    if child.kind() == "decorator" {
-                        self.collect_decorator(child);
-                    } else {
-                        self.visit(child);
+    /// Depth-first via an explicit stack: AST depth is input-controlled
+    /// (deeply nested expressions), and extraction runs on small
+    /// worker-thread stacks.
+    fn run(&mut self, root: Node<'a>) {
+        let mut stack = vec![root];
+        while let Some(node) = stack.pop() {
+            match node.kind() {
+                "decorated_definition" => {
+                    // decorators collect now; the definition pops next
+                    let mut cursor = node.walk();
+                    let mut deferred: Vec<Node<'a>> = Vec::new();
+                    for child in node.named_children(&mut cursor) {
+                        if child.kind() == "decorator" {
+                            self.collect_decorator(child);
+                        } else {
+                            deferred.push(child);
+                        }
                     }
+                    stack.extend(deferred.into_iter().rev());
                 }
-            }
-            "function_definition" => self.handle_function(node),
-            "class_definition" => self.handle_class(node),
-            "expression_statement" => {
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    if child.kind() == "assignment" {
-                        self.handle_assignment(child);
+                "function_definition" => self.handle_function(node, &mut stack),
+                "class_definition" => self.handle_class(node, &mut stack),
+                "expression_statement" => {
+                    let mut cursor = node.walk();
+                    for child in node.named_children(&mut cursor) {
+                        if child.kind() == "assignment" {
+                            self.handle_assignment(child);
+                        }
                     }
                 }
-            }
-            // expressions cannot contain statements; recursion elsewhere is safe
-            _ => {
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    self.visit(child);
+                _ => {
+                    let mut cursor = node.walk();
+                    let children: Vec<Node<'a>> = node.named_children(&mut cursor).collect();
+                    stack.extend(children.into_iter().rev());
                 }
             }
         }
@@ -153,7 +160,7 @@ impl<'a> PyExtractor<'a> {
             .push(self.content[indent_start..node.end_byte()].to_string());
     }
 
-    fn handle_function(&mut self, node: Node<'a>) {
+    fn handle_function(&mut self, node: Node<'a>, stack: &mut Vec<Node<'a>>) {
         let body = node.child_by_field_name("body");
         let is_async = node.child(0).map(|c| c.kind() == "async").unwrap_or(false);
         if !is_async {
@@ -163,7 +170,7 @@ impl<'a> PyExtractor<'a> {
             }
         }
         if let Some(body) = body {
-            self.visit(body);
+            stack.push(body);
         }
     }
 
@@ -208,14 +215,14 @@ impl<'a> PyExtractor<'a> {
         Some(cleaned)
     }
 
-    fn handle_class(&mut self, node: Node<'a>) {
+    fn handle_class(&mut self, node: Node<'a>, stack: &mut Vec<Node<'a>>) {
         let body = node.child_by_field_name("body");
         if let Some(component) = self.build_class_signature(node) {
             self.emit_with_decorators(component);
             self.in_class = true;
         }
         if let Some(body) = body {
-            self.visit(body);
+            stack.push(body);
         }
     }
 
diff --git a/crates/tree_plus_core/src/extract/treesitter/rust.rs b/crates/tree_plus_core/src/extract/treesitter/rust.rs
index 3befc1d..d83006a 100644
--- a/crates/tree_plus_core/src/extract/treesitter/rust.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/rust.rs
@@ -100,55 +100,58 @@ pub fn extract(content: &str, syntax: bool) -> ExtractResult {
     Ok(components)
 }
 
-fn visit(node: Node<'_>, content: &str, syntax: bool, out: &mut Vec<String>) {
-    let mut cursor = node.walk();
-    for child in node.named_children(&mut cursor) {
-        match child.kind() {
+/// Depth-first via an explicit stack: AST depth is input-controlled, and
+/// deep expression nesting must not overflow small worker-thread stacks.
+fn visit(root: Node<'_>, content: &str, syntax: bool, out: &mut Vec<String>) {
+    let mut stack = vec![root];
+    while let Some(node) = stack.pop() {
+        match node.kind() {
             "function_item" | "function_signature_item" => {
-                if let Some(c) = format_fn(child, content) {
+                if let Some(c) = format_fn(node, content) {
                     out.push(c);
                 }
                 // nested items inside fn bodies still match
-                visit(child, content, syntax, out);
+                descend(node, &mut stack);
             }
             "struct_item" => {
-                if let Some(c) = format_struct_impl(child, content, false) {
+                if let Some(c) = format_struct_impl(node, content, false) {
                     out.push(c);
                 }
             }
             "impl_item" => {
-                if let Some(c) = format_struct_impl(child, content, true) {
+                if let Some(c) = format_struct_impl(node, content, true) {
                     out.push(c);
                 }
-                visit(child, content, syntax, out);
+                descend(node, &mut stack);
             }
             "enum_item" => {
-                if let Some(c) = format_enum(child, content, syntax) {
-                    out.push(c);
-                }
-            }
-            "trait_item" => {
-                if let Some(c) = format_trait_mod(child, content) {
+                if let Some(c) = format_enum(node, content, syntax) {
                     out.push(c);
                 }
-                visit(child, content, syntax, out);
             }
-            "mod_item" => {
-                if let Some(c) = format_trait_mod(child, content) {
+            "trait_item" | "mod_item" => {
+                if let Some(c) = format_trait_mod(node, content) {
                     out.push(c);
                 }
-                visit(child, content, syntax, out);
+                descend(node, &mut stack);
             }
             "macro_definition" => {
-                if let Some(c) = format_macro(child, content) {
+                if let Some(c) = format_macro(node, content) {
                     out.push(c);
                 }
             }
-            _ => visit(child, content, syntax, out),
+            _ => descend(node, &mut stack),
         }
     }
 }
 
+/// Push `node`'s named children so they pop in source order.
+fn descend<'t>(node: Node<'t>, stack: &mut Vec<Node<'t>>) {
+    let mut cursor = node.walk();
+    let children: Vec<Node<'t>> = node.named_children(&mut cursor).collect();
+    stack.extend(children.into_iter().rev());
+}
+
 /// Start of the line containing `byte` (offset just after the previous `\n`).
 fn line_start(content: &str, byte: usize) -> usize {
     content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
diff --git a/crates/tree_plus_core/src/extract/treesitter/typescript.rs b/crates/tree_plus_core/src/extract/treesitter/typescript.rs
index 55dd887..dfd76ee 100644
--- a/crates/tree_plus_core/src/extract/treesitter/typescript.rs
+++ b/crates/tree_plus_core/src/extract/treesitter/typescript.rs
@@ -51,7 +51,7 @@ pub fn extract(content: &str, tsx: bool) -> ExtractResult {
         content,
         components: Vec::new(),
     };
-    extractor.visit(tree.root_node());
+    extractor.run(tree.root_node());
     Ok(extractor.components)
 }
 
@@ -60,6 +60,13 @@ struct TsExtractor<'a> {
     components: Vec<String>,
 }
 
+/// Push `node`'s named children so they pop in source order.
+fn push_children_rev<'t>(node: Node<'t>, stack: &mut Vec<Node<'t>>) {
+    let mut cursor = node.walk();
+    let children: Vec<Node<'t>> = node.named_children(&mut cursor).collect();
+    stack.extend(children.into_iter().rev());
+}
+
 fn line_start(content: &str, byte: usize) -> usize {
     content[..byte].rfind('\n').map(|i| i + 1).unwrap_or(0)
 }
@@ -121,12 +128,22 @@ impl<'a> TsExtractor<'a> {
             .to_string()
     }
 
-    fn visit(&mut self, node: Node<'a>) {
+    /// Depth-first via an explicit stack: AST depth is input-controlled
+    /// (deeply nested expressions), and extraction runs on small
+    /// worker-thread stacks.
+    fn run(&mut self, root: Node<'a>) {
+        let mut stack = vec![root];
+        while let Some(node) = stack.pop() {
+            self.visit(node, &mut stack);
+        }
+    }
+
+    fn visit(&mut self, node: Node<'a>, stack: &mut Vec<Node<'a>>) {
         match node.kind() {
             "class_declaration" | "abstract_class_declaration" => {
                 self.emit_class_like(node);
                 if let Some(body) = node.child_by_field_name("body") {
-                    self.visit(body);
+                    stack.push(body);
                 }
             }
             "interface_declaration" => {
@@ -144,7 +161,7 @@ impl<'a> TsExtractor<'a> {
             "function_declaration" | "generator_function_declaration" => {
                 self.emit_function(node);
                 if let Some(body) = node.child_by_field_name("body") {
-                    self.visit(body);
+                    stack.push(body);
                 }
             }
             "function_expression" | "generator_function" => {
@@ -152,13 +169,13 @@ impl<'a> TsExtractor<'a> {
                     self.emit_function(node);
                 }
                 if let Some(body) = node.child_by_field_name("body") {
-                    self.visit(body);
+                    stack.push(body);
                 }
             }
             "method_definition" | "method_signature" | "abstract_method_signature" => {
                 self.emit_signature_to_params_or_return(node);
                 if let Some(body) = node.child_by_field_name("body") {
-                    self.visit(body);
+                    stack.push(body);
                 }
             }
             "arrow_function" => {
@@ -166,7 +183,7 @@ impl<'a> TsExtractor<'a> {
                     self.emit_arrow(node);
                 }
                 if let Some(body) = node.child_by_field_name("body") {
-                    self.visit(body);
+                    stack.push(body);
                 }
             }
             "expression_statement" => {
@@ -177,10 +194,7 @@ impl<'a> TsExtractor<'a> {
                 {
                     self.maybe_emit_bare_call(node, call);
                 }
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    self.visit(child);
-                }
+                push_children_rev(node, stack);
             }
             "variable_declarator" => {
                 // object scope: `const name = {` when the object holds functions
@@ -192,14 +206,11 @@ impl<'a> TsExtractor<'a> {
                         self.components
                             .push(self.content[start..=brace].trim_end().to_string());
                     }
-                    self.visit(value);
+                    stack.push(value);
                 }
             }
             _ => {
-                let mut cursor = node.walk();
-                for child in node.named_children(&mut cursor) {
-                    self.visit(child);
-                }
+                push_children_rev(node, stack);
             }
         }
     }
diff --git a/crates/tree_plus_core/src/walk.rs b/crates/tree_plus_core/src/walk.rs
index c519409..c4d4db7 100644
--- a/crates/tree_plus_core/src/walk.rs
+++ b/crates/tree_plus_core/src/walk.rs
@@ -14,8 +14,26 @@ use crate::ignore::{
 use crate::model::{Category, TreePlus};
 use crate::sort::os_sort_key;
 
+/// Dedicated rayon pool with large worker stacks. Work-stealing nests
+/// `from_folder` frames on a single worker while it waits in `join`, and
+/// tree-sitter's C parser needs headroom below that; rayon's default 2 MiB
+/// worker stacks segfault on big trees (e.g. torvalds/linux `drivers/`).
+fn pool() -> &'static rayon::ThreadPool {
+    static POOL: std::sync::LazyLock<rayon::ThreadPool> = std::sync::LazyLock::new(|| {
+        rayon::ThreadPoolBuilder::new()
+            .stack_size(16 * 1024 * 1024)
+            .build()
+            .expect("build worker pool")
+    });
+    &POOL
+}
+
 /// Build a `TreePlus` from seed path/glob strings (legacy `from_seeds`).
 pub fn from_seeds(seeds: &[String], config: &TreePlusConfig) -> TreePlus {
+    pool().install(|| from_seeds_inner(seeds, config))
+}
+
+fn from_seeds_inner(seeds: &[String], config: &TreePlusConfig) -> TreePlus {
     let mut seeds: Vec<String> = if seeds.is_empty() {
         vec![std::env::current_dir()
             .map(|p| p.to_string_lossy().into_owned())
diff --git a/crates/tree_plus_core/tests/robustness.rs b/crates/tree_plus_core/tests/robustness.rs
index 79bb957..64e42dd 100644
--- a/crates/tree_plus_core/tests/robustness.rs
+++ b/crates/tree_plus_core/tests/robustness.rs
@@ -84,3 +84,60 @@ fn invalid_utf8_yields_no_components() {
     let path = write_temp("invalid.py", &[0xC3, 0x28, b'\n', b'd', b'e', b'f', b' ']);
     assert!(extract_components(&path, false).is_empty());
 }
+
+/// Deep ASTs must not overflow small (rayon-sized) thread stacks.
+///
+/// Regression: arch/x86/kernel/cpu/microcode/intel-ucode-defs.h in
+/// torvalds/linux is a headerless initializer-list fragment; tree-sitter
+/// parses it as a deeply nested ERROR tree, and the recursive extractor
+/// walk aborted with a stack overflow inside the rayon pool.
+#[test]
+fn deep_nesting_never_overflows_worker_stacks() {
+    let cases: Vec<(&str, String)> = vec![
+        (
+            "h",
+            "{ .flags = X86_CPU_ID_FLAG_ENTRY_VALID, .vendor = 1, .family = 0x6 },\n".repeat(5_000),
+        ),
+        (
+            "c",
+            format!("int x = {}1{};\n", "(".repeat(20_000), ")".repeat(20_000)),
+        ),
+        (
+            "py",
+            format!("x = {}{}\n", "[".repeat(20_000), "]".repeat(20_000)),
+        ),
+        (
+            "rs",
+            format!(
+                "fn f() {{ let x = {}1{}; }}\n",
+                "(".repeat(20_000),
+                ")".repeat(20_000)
+            ),
+        ),
+        (
+            "ts",
+            format!("const x = {}{};\n", "[".repeat(20_000), "]".repeat(20_000)),
+        ),
+        (
+            "go",
+            format!(
+                "func f() {{\n\tx := {}1{}\n}}\n",
+                "(".repeat(20_000),
+                ")".repeat(20_000)
+            ),
+        ),
+    ];
+    for (ext, content) in cases {
+        let path = write_temp(&format!("deep.{ext}"), content.as_bytes());
+        // rayon workers default to 2 MiB stacks; use a harsher 512 KiB
+        let handle = std::thread::Builder::new()
+            .stack_size(512 * 1024)
+            .spawn(move || {
+                let _ = extract_components(&path, false); // must not abort
+            })
+            .unwrap();
+        handle
+            .join()
+            .unwrap_or_else(|_| panic!("deep .{ext} input crashed the extractor"));
+    }
+}
diff --git a/docs/performance.md b/docs/performance.md
index b273e28..623574f 100644
--- a/docs/performance.md
+++ b/docs/performance.md
@@ -31,9 +31,35 @@ hyperfine './target/release/tree_plus /path/to/repo > /dev/null' \
           'python tree_plus_cli.py /path/to/repo > /dev/null'
 ```
 
-A Linux-kernel-sized tree was not available locally; the command above is
-the reproducible benchmark for one (also:
-`TREE_PLUS_BENCH_PATH=/path/to/linux cargo bench -p tree_plus_core`).
+## Linux kernel (torvalds/linux @ ~6.15, 101,136 files, 44.77M lines)
+
+Single runs (`/usr/bin/time -v`), page cache warm:
+
+| Invocation | Rust wall | Peak RSS | Notes |
+|---|---|---|---|
+| `tprs -c linux` | 0.50 s | 178 MB | 6,285 folders, 101,136 files, 44.77M lines, 409.2M tokens |
+| `tprs linux` | 12.4 s | 1.5 GB | full extraction; 8.42M-line render (598 MB of output) |
+| `tprs linux/kernel` | 0.23 s | — | 711 files, 553,855 lines; Python: 12.6 s → **~56×** |
+
+Counts (lines, tokens) agree byte-for-byte with the legacy Python CLI on
+`linux/kernel`. Peak RSS on the full run is dominated by the final render
+string (~600 MB) plus the assembled tree, not by extraction.
+
+The kernel run also surfaced two stack-overflow bugs, both fixed:
+
+1. Extractor visitors recursed on AST depth.
+   `arch/x86/kernel/cpu/microcode/intel-ucode-defs.h` is a headerless
+   initializer-list fragment that tree-sitter parses as a deeply nested
+   ERROR tree; the recursive walk blew rayon's 2 MiB worker stacks. All
+   tree-sitter extractors now traverse with explicit heap stacks
+   (regression test: `deep_nesting_never_overflows_worker_stacks`).
+2. Rayon work-stealing nests `from_folder` frames on a worker while it
+   waits in `join`, leaving too little headroom for tree-sitter's C
+   frames on big trees (`drivers/` segfaulted). Extraction now runs in a
+   dedicated pool with 16 MiB worker stacks.
+
+Also reproducible via
+`TREE_PLUS_BENCH_PATH=/path/to/linux cargo bench -p tree_plus_core`.
 
 ## Single-file extraction latency (criterion, mean)
 

From 99c3117968c89c8ac141c22d2674f24eca3c07e9 Mon Sep 17 00:00:00 2001
From: Bion Howard <flesheatingemu@gmail.com>
Date: Tue, 9 Jun 2026 18:25:13 -0400
Subject: [PATCH 8/8] docs: record full Linux-kernel Python baseline (10m02s /
 6.7GB vs 12.4s / 1.5GB)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 docs/performance.md | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/docs/performance.md b/docs/performance.md
index 623574f..4348648 100644
--- a/docs/performance.md
+++ b/docs/performance.md
@@ -41,9 +41,17 @@ Single runs (`/usr/bin/time -v`), page cache warm:
 | `tprs linux` | 12.4 s | 1.5 GB | full extraction; 8.42M-line render (598 MB of output) |
 | `tprs linux/kernel` | 0.23 s | — | 711 files, 553,855 lines; Python: 12.6 s → **~56×** |
 
-Counts (lines, tokens) agree byte-for-byte with the legacy Python CLI on
-`linux/kernel`. Peak RSS on the full run is dominated by the final render
-string (~600 MB) plus the assembled tree, not by extraction.
+The legacy Python CLI on the full tree: **10 min 2 s, 6.7 GB RSS** — and
+its render timed out internally, so it printed only the stats footer,
+while the Rust run includes the full 8.42M-line render. End-to-end that
+is **~48×** faster with ~4.5× less memory.
+
+Counts agree byte-for-byte with the legacy CLI on `linux/kernel`. On the
+full tree they drift by 0.0005% (Python +208 lines, Rust +6,373 tokens of
+409M) from a handful of invalid-UTF-8 kernel fixtures, consistent with
+the documented decoding differences. Peak RSS on the Rust run is
+dominated by the final render string (~600 MB) plus the assembled tree,
+not by extraction.
 
 The kernel run also surfaced two stack-overflow bugs, both fixed: