minor fixes for ibims orm creation (#99)

* relative imports

+snakecasing

* fixed linting

* fixed uuid refs

* many refs fixed uuid ref

* remove unused function

Author: DeltaDaniel

pyproject.toml

@@ -19,6 +19,7 @@ classifiers = [
 dependencies = [
     "datamodel-code-generator",
     "click",
+    "autoflake"
 ]     # add all the dependencies from requirements.in here, too
 dynamic = ["readme", "version"]
 

requirements.in

@@ -2,20 +2,24 @@
 # This file is autogenerated by pip-compile with Python 3.12
 # by the following command:
 #
-#    pip-compile requirements.in
+#    pip-compile '.\pyproject.toml'
 #
 annotated-types==0.6.0
     # via pydantic
 argcomplete==3.1.2
     # via datamodel-code-generator
+autoflake==2.3.1
+    # via BO4E-Python-Generator (pyproject.toml)
 black==24.8.0
     # via datamodel-code-generator
 click==8.1.7
     # via
-    #   -r requirements.in
+    #   BO4E-Python-Generator (pyproject.toml)
     #   black
 datamodel-code-generator==0.25.9
     # via -r requirements.in
+colorama==0.4.6
+    # via click
 dnspython==2.4.2
     # via email-validator
 email-validator==2.0.0.post2
@@ -46,6 +50,8 @@ pydantic[email]==2.4.2
     # via datamodel-code-generator
 pydantic-core==2.10.1
     # via pydantic
+pyflakes==3.2.0
+    # via autoflake
 pyyaml==6.0.1
     # via datamodel-code-generator
 typing-extensions==4.8.0

requirements.txt

@@ -16,6 +16,7 @@
     parse_bo4e_schemas,
 )
 from bo4e_generator.schema import get_namespace, get_version
+from bo4e_generator.sqlparser import remove_unused_imports
 
 
 def resolve_paths(input_directory: Path, output_directory: Path) -> tuple[Path, Path]:
@@ -52,6 +53,11 @@ def generate_bo4e_schemas(
     for relative_file_path, file_content in file_contents.items():
         file_path = output_directory / relative_file_path
         file_path.parent.mkdir(parents=True, exist_ok=True)
+        if (
+            relative_file_path.name not in ["__init__.py", "__version__.py"]
+            and OutputType[output_type] == OutputType.SQL_MODEL
+        ):
+            file_content = remove_unused_imports(file_content)
         file_content = formatter.format_code(file_content)
         file_path.write_text(file_content, encoding="utf-8")
         print(f"Created {file_path}")

src/bo4e_generator/__main__.py

@@ -1,20 +1,30 @@
 {%- if SQL and SQL['imports']%}
-{%- for class_name, module_path in SQL['imports'].items() %}
+{%- for import_class_name, module_path in SQL['imports'].items() %}
 {%- if module_path[:4] == 'enum'%}
-from borm.models.{{ module_path }} import {{ class_name }}
+from ..{{ module_path }} import {{ import_class_name }}
 {%- elif module_path == 'Link'%}
-from borm.models.many import {{ class_name }}
+{% if class_name == 'ZusatzAttribut'%}
+from .many import {{ import_class_name }}
+{% else %}
+from ..many import {{ import_class_name }}
+{% endif %}
 {%- else %}
-from {{ module_path }} import {{ class_name }}
+from {{ module_path }} import {{ import_class_name }}
 {%- endif %}
 {%- endfor -%}
 {%- endif %}
 {%- if SQL and SQL['relationimports']%}
 from typing import TYPE_CHECKING
 if TYPE_CHECKING:
-{%- for class_name, module_path in SQL['relationimports'].items() %}
-    from borm.models.{{ module_path }} import {{ class_name }}
+{% if class_name == 'ZusatzAttribut'%}
+{%- for import_class_name, module_path in SQL['relationimports'].items() %}
+    from .{{ module_path }} import {{ import_class_name }}
 {%- endfor -%}
+{% else %}
+{%- for import_class_name, module_path in SQL['relationimports'].items() %}
+    from ..{{ module_path }} import {{ import_class_name }}
+{%- endfor -%}
+{%- endif %}
 {%- endif %}
 {% for decorator in decorators -%}
 {{ decorator }}

src/bo4e_generator/custom_templates/BaseModel.jinja2

@@ -14,8 +14,8 @@ class {{class1}}{{class2[1]}}Link(SQLModel, table=True):
     """
     class linking m-n relation of tables {{class1}} and {{class2[0]}} for field {{ class2[1]}}.
     """
-    {{class1.lower()}}_id: Optional[uuid_pkg.UUID] = Field(sa_column=Column(UUID(as_uuid=True), ForeignKey("{{class1.lower()}}.{{class1.lower()}}_sqlid", ondelete="CASCADE"), primary_key=True))
-    {{class2[0].lower()}}_id: Optional[uuid_pkg.UUID] = Field(sa_column=Column(UUID(as_uuid=True), ForeignKey("{{class2[0].lower()}}.{{class2[0].lower()}}_sqlid", ondelete="CASCADE"), primary_key=True))
+    {{class1.lower()}}_id: Optional[uuid_pkg.UUID] = Field(sa_column=Column(UUID(as_uuid=True), ForeignKey("{{class1.lower()}}.id", ondelete="CASCADE"), primary_key=True))
+    {{class2[0].lower()}}_id: Optional[uuid_pkg.UUID] = Field(sa_column=Column(UUID(as_uuid=True), ForeignKey("{{class2[0].lower()}}.id", ondelete="CASCADE"), primary_key=True))
 
 {%- endfor -%}
 {%- endfor -%}

src/bo4e_generator/custom_templates/ManyLinks.jinja2

@@ -4,16 +4,17 @@
 """
 
 import json
+import os
 import re
+import subprocess
+import tempfile
 from collections import defaultdict
 from pathlib import Path
 from typing import Any, DefaultDict, Union
 
-import black
-import isort
 from jinja2 import Environment, FileSystemLoader
 
-from bo4e_generator.schema import SchemaMetadata
+from bo4e_generator.schema import SchemaMetadata, camel_to_snake
 
 
 def remove_pydantic_field_import(python_code: str) -> str:
@@ -42,29 +43,32 @@ def adapt_parse_for_sql(
     for schema_metadata in namespace.values():
         if schema_metadata.module_path[0] != "enum":
             # list of fields which will be replaced by modified versions
-            del_fields = []
+            del_fields = set()
             for field, val in schema_metadata.schema_parsed["properties"].items():
                 # type Any field
                 if "type" not in str(val):
                     add_relation, relation_imports = create_sql_any(
                         field, schema_metadata.class_name, namespace, add_relation, relation_imports
                     )
-                    del_fields.append(field)
+                    del_fields.add(field)
                 # modify decimal fields
                 if "number" in str(val) and "string" in str(val):
                     relation_imports[schema_metadata.class_name + "ADD"]["Decimal"] = "decimal"
                 if "array" in str(val) and "$ref" not in str(val):
                     add_relation, relation_imports = create_sql_list(
                         field, schema_metadata.class_name, namespace, add_relation, relation_imports
                     )
-                    del_fields.append(field)
+                    del_fields.add(field)
                 if "$ref" in str(val):  # or "array" in str(val):
                     add_relation, relation_imports = create_sql_field(
                         field, schema_metadata.class_name, namespace, add_relation, relation_imports
                     )
-                    del_fields.append(field)
+                    del_fields.add(field)
             for field in del_fields:
                 del schema_metadata.schema_parsed["properties"][field]
+            # delete id field as it is replaced below
+            if schema_metadata.schema_parsed["properties"].get("_id"):
+                del schema_metadata.schema_parsed["properties"]["_id"]
             # store the reduced version. The modified fields will be added in the BaseModel.jinja2 schema
             schema_metadata.schema_text = json.dumps(schema_metadata.schema_parsed, indent=2, ensure_ascii=False)
 
@@ -104,9 +108,8 @@ def additional_sql_arguments(
         if schema_metadata.module_path[0] != "enum":
             # add primary key
             additional_sql_data[schema_metadata.class_name]["SQL"] = {
-                "primary": schema_metadata.class_name.lower()
-                + "_sqlid: uuid_pkg.UUID = Field( default_factory=uuid_pkg.uuid4, primary_key=True, index=True, "
-                "nullable=False )"
+                "primary": "id: uuid_pkg.UUID = Field( default_factory=uuid_pkg.uuid4, primary_key=True, index=True, "
+                'nullable=False, alias="_id", title=" Id" )'
             }
         if schema_metadata.class_name in add_relation:
             additional_sql_data[schema_metadata.class_name]["SQL"]["relations"] = add_relation[
@@ -184,7 +187,7 @@ def create_sql_list(
     add_imports[class_name + "ADD"]["Column, ARRAY"] = "sqlalchemy"
     add_imports[class_name + "ADD"][sa_type] = "sqlalchemy"
 
-    add_fields[class_name][f"{field_name}"] = (
+    add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
         f"List[{type_hint}] "
         + is_optional
         + f' = Field({default}, title="{field_name}", sa_column=Column( ARRAY( {sa_type} )))'
@@ -209,12 +212,14 @@ def sql_reference_enum(
     returns field which references enums.
     """
     if is_list:
-        add_fields[class_name][f"{field_name}"] = (
+        add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
             f"List[{reference_name}]" + is_optional + f" = Field({default},"
             f' sa_column=Column( ARRAY( Enum( {reference_name}, name="{reference_name.lower()}"))))'
         )
     else:
-        add_fields[class_name][f"{field_name}"] = f"{reference_name}" + is_optional + f"= Field({default})"
+        add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
+            f"{reference_name}" + is_optional + f"= Field({default})"
+        )
 
     # import enums
     if is_list:
@@ -265,23 +270,23 @@ def create_sql_field(
             add_fields["MANY"][class_name] = [[reference_name, field_name]]
         elif reference_name not in add_fields["MANY"][class_name]:
             add_fields["MANY"][class_name].append([reference_name, field_name])
-        add_fields[class_name][f"{field_name}"] = (
+        add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
             f'List["{reference_name}"] ='
             f' Relationship(back_populates="{class_name.lower()}_{field_name.lower()}_link", '
             f"link_model={class_name}{field_name}Link)"
         )
         add_fields[reference_name][f"{class_name.lower()}_{field_name.lower()}_link"] = (
             f'List["{class_name}"] ='
-            f' Relationship(back_populates="{field_name}", '
+            f' Relationship(back_populates="{camel_to_snake(field_name)}", '
             f"link_model={class_name}{field_name}Link)"
         )
         add_imports[class_name + "ADD"][f"{class_name}{field_name}Link)"] = "Link"
         add_imports[reference_name + "ADD"][f"{class_name}{field_name}Link)"] = "Link"
     else:
         # cf. https://github.com/tiangolo/sqlmodel/pull/610
-        add_fields[class_name][f"{field_name}_id"] = (
+        add_fields[class_name][f"{camel_to_snake(field_name)}_id"] = (
             "uuid_pkg.UUID " + is_optional + f" = Field(sa_column=Column(UUID(as_uuid=True),"
-            f' ForeignKey("{reference_name.lower()}.{reference_name.lower()}_sqlid"'
+            f' ForeignKey("{reference_name.lower()}.id"'
             f', ondelete="SET NULL")))'
         )
         add_imports[class_name + "ADD"]["Column"] = "sqlalchemy"
@@ -291,20 +296,20 @@ def create_sql_field(
         # pylint: disable= fixme
         # todo: check default
 
-        add_fields[class_name][f"{field_name}"] = (
+        add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
             f'"{reference_name}" ='
-            f' Relationship(back_populates="{class_name.lower()}_{field_name}",'
-            f' sa_relationship_kwargs= {{ "foreign_keys":"[{class_name}.{field_name}_id]" }})'
+            f' Relationship(back_populates="{class_name.lower()}_{camel_to_snake(field_name)}",'
+            f' sa_relationship_kwargs= {{ "foreign_keys":"[{class_name}.{camel_to_snake(field_name)}_id]" }})'
         )
 
         # cf. https://github.com/tiangolo/sqlmodel/issues/10
         # https://github.com/tiangolo/sqlmodel/issues/213
         # https://dev.to/whchi/disable-sqlmodel-foreign-key-constraint-55kp
-        add_fields[reference_name][f"{class_name.lower()}_{field_name}"] = (
-            f'List["{class_name}"] = Relationship(back_populates="{field_name}",'
+        add_fields[reference_name][f"{class_name.lower()}_{camel_to_snake(field_name)}"] = (
+            f'List["{class_name}"] = Relationship(back_populates="{camel_to_snake(field_name)}",'
             f"sa_relationship_kwargs="
             f'{{"primaryjoin":'
-            f' "{class_name}.{field_name}_id=={reference_name}.{reference_name.lower()}_sqlid",'
+            f' "{class_name}.{camel_to_snake(field_name)}_id=={reference_name}.id",'
             f' "lazy": "joined"}})'
         )
     # add_relation_import
@@ -346,11 +351,11 @@ def create_sql_any(
     if is_list:
         add_imports[class_name + "ADD"]["List"] = "typing"
         add_imports[class_name + "ADD"]["ARRAY"] = "sqlalchemy"
-        add_fields[class_name][f"{field_name}"] = (
+        add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
             "List[Any]" + is_optional + f" = Field({default}," f" sa_column=Column( ARRAY( PickleType)))"
         )
     else:
-        add_fields[class_name][f"{field_name}"] = (
+        add_fields[class_name][f"{camel_to_snake(field_name)}"] = (
             "Any" + is_optional + f" = Field({default}," f" sa_column=Column(  PickleType))"
         )
 
@@ -365,13 +370,27 @@ def write_many_many_links(links: dict[str, str]) -> str:
     environment = Environment(loader=FileSystemLoader(template_path))
     template = environment.get_template("ManyLinks.jinja2")
     python_code = template.render({"class": links})
-    python_code = format_code(python_code)
+    # python_code = format_code(python_code)
     return python_code
 
 
-def format_code(code: str) -> str:
+def remove_unused_imports(code):
     """
-    perform isort and black on code
+    Removes unused imports from the given code using autoflake.
     """
-    code = black.format_str(code, mode=black.Mode())
-    return isort.code(code, known_local_folder=["borm"])
+    # Create a temporary file
+    with tempfile.NamedTemporaryFile(suffix=".py", delete=False) as tmp_file:
+        tmp_file_name = tmp_file.name
+        tmp_file.write(code.encode("utf-8"))
+
+    # Run autoflake to remove unused imports
+    subprocess.run(["autoflake", "--remove-all-unused-imports", "--in-place", tmp_file_name], check=True)
+
+    # Read the cleaned code from the temporary file
+    with open(tmp_file_name, "r", encoding="utf-8") as tmp_file:
+        cleaned_code = tmp_file.read()
+
+    # Clean up the temporary file
+    os.remove(tmp_file_name)
+
+    return cleaned_code

src/bo4e_generator/sqlparser.py

@@ -78,7 +78,7 @@ deps =
     pre-commit
 commands =
     python -m pip install --upgrade pip
-    pip-compile requirements.in
+    pip-compile .\pyproject.toml
     pip install -r requirements.txt
     pre-commit install
 

tox.ini

@@ -8,7 +8,6 @@
 from bo4e_generator.sqlparser import (
     adapt_parse_for_sql,
     create_sql_field,
-    format_code,
     remove_pydantic_field_import,
     return_ref,
     write_many_many_links,
@@ -76,16 +75,3 @@ def test_write_many_many_links(self) -> None:
         file_contents = write_many_many_links(links)
         keywords = ["AngebotzusatzAttributeLink", "angebot_id", "zusatzattribut_id"]
         assert all(substring in file_contents for substring in keywords)
-
-    def test_format_code(self) -> None:
-        unsorted = (
-            "from sqlmodel import Field, Relationship, SQLModel\n"
-            "from typing import TYPE_CHECKING, List\n"
-            "from borm.models.enum.anrede import Anrede"
-        )
-        resorted = (
-            "from typing import TYPE_CHECKING, List\n\n"
-            "from sqlmodel import Field, Relationship, SQLModel\n\n"
-            "from borm.models.enum.anrede import Anrede\n"
-        )
-        assert resorted == format_code(unsorted)