Skip to content

Latest commit

 

History

History
41 lines (33 loc) · 1.85 KB

File metadata and controls

41 lines (33 loc) · 1.85 KB

Perform a systematic refactoring of the specified code area.

Steps

1. Identify Code Smells

Read the target code and identify issues:

  • Long functions (>40 lines): break into smaller functions with descriptive names.
  • Deep nesting (>3 levels): use early returns, guard clauses, or extract functions.
  • Duplicate logic: extract shared functions or use parameterization.
  • God objects/modules: split into focused, single-responsibility units.
  • Primitive obsession: introduce domain types or value objects.
  • Feature envy: move logic to the class/module that owns the data.
  • Unclear naming: rename variables, functions, and files to reveal intent.

2. Verify Test Coverage

  • Run existing tests to establish the green baseline.
  • If coverage is insufficient, write characterization tests that capture current behavior before changing anything.
  • Every refactored path must have a test that would catch a regression.

3. Plan Changes

  • List each refactoring step in order.
  • Each step should be a single, atomic change (one rename, one extract, one move).
  • Order steps to minimize risk: rename before restructure, restructure before optimize.

4. Execute

  • Apply one refactoring at a time.
  • Run tests after each step to confirm behavior is preserved.
  • Commit after each successful step if the changes are significant.

5. Verify

  • Run the full test suite.
  • Check that type checking passes.
  • Compare behavior before and after (same inputs produce same outputs).

Rules

  • Refactoring changes structure, not behavior. If behavior changes, that is a feature or a fix.
  • Never refactor and add features in the same step.
  • If tests are missing, add them before refactoring, not after.
  • Keep diffs reviewable: prefer multiple small commits over one massive change.
  • If a refactoring reveals a bug, note it and fix it in a separate commit.