feat(parser): add Vitest/Jest test detection for describe/it/test blocks

[JF10R]

Summary

• Detect Vitest/Jest test runner calls (describe, it, test, beforeEach, etc.) in test files and create proper Test nodes
• Recurse into callback bodies with correct enclosing_func context so CALLS edges inside tests are properly attributed
• Enables automatic TESTED_BY edge generation via the existing reverse-edge logic

Before: 1 Test node, 2 TESTED_BY edges across 69 test files
After: 3,205 Test nodes, 18,331 TESTED_BY edges

Root cause

Tree-sitter parses describe("suite", () => { ... }) as a call_expression containing an anonymous arrow_function. The parser found the arrow function but skipped it because _get_name() returns None for anonymous functions. No Test node was ever created.

Changes

• Modified: code_review_graph/parser.py
  • Added _TEST_RUNNER_NAMES constant
  • Added _get_test_description() method to extract the description string from test calls
  • New branch in _extract_from_tree() that intercepts test runner call_expression nodes in test files, creates Test nodes with synthetic names (e.g., describe:UserService), and recurses with proper enclosing_func
  • Deduplicated _is_test_function() to reuse _TEST_RUNNER_NAMES
• New fixture: tests/fixtures/sample_vitest.test.ts
• New tests: 5 test methods covering detection, CONTAINS edges, CALLS edges, TESTED_BY edges, and non-test-file guard

Design decisions

• Test runner calls are intercepted at the call_expression level (not by modifying arrow function naming), keeping the change localized
• Synthetic node names use call_name:description format (e.g., it:should create a user)
• Only applies to JS/TS/TSX files that match test file patterns — describe() in production code is not affected
• Nesting is handled naturally by recursion: describe → it → function calls

Test plan

• All 33 parser tests pass (pytest tests/test_parser.py -v)
• Full test suite: 196 passed, 26 pre-existing Windows-only SQLite teardown errors
• Validated on a real-world TypeScript project (357 files, 69 test files): 3,205 Test nodes and 18,331 TESTED_BY edges detected

🤖 Generated with Claude Code

[Copilot]

Pull request overview

Adds Vitest/Jest-style test block detection to the Tree-sitter parser so describe/it/test (and hooks) in JS/TS test files produce Test nodes and proper containment/call attribution, enabling much broader automatic TESTED_BY edge generation.

Changes:

• Extend code_review_graph/parser.py to detect test-runner call_expression nodes in JS/TS test files, create synthetic Test nodes, and recurse into callback bodies with the correct enclosing_func.
• Add a Vitest fixture (sample_vitest.test.ts) that exercises nested describe + it/test blocks and calls into imported production symbols.
• Add parser tests validating Test node creation plus CONTAINS, CALLS, and TESTED_BY edge generation, and ensure non-test files don’t get special handling.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 5 comments.

File	Description
code_review_graph/parser.py	Adds test-runner call interception and description extraction to build Test nodes and attribute internal calls to those tests.
tests/fixtures/sample_vitest.test.ts	New TS fixture containing describe/it/test blocks and calls into imported code.
tests/test_parser.py	New pytest cases covering test detection and edge generation for Vitest/Jest-style tests.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[JF10R]

All 5 review comments addressed in 460f516:

parser.py:

• _get_test_description() now normalizes whitespace and truncates to 200 chars
• Synthetic test names include @l{line} suffix for uniqueness (e.g., it:should work@L42)
• Added _get_base_call_name() to detect describe.only(), it.skip(), test.each() and other modifier patterns

Tests:

• All assertions use startswith() instead of substring in checks to avoid false positives

[tirth8205]

Integrated into main via PR #68. Thank you for the contribution! 🎉