MCP Code Analysis Server

"""Tests for the ranking module.""" # These tests target a high-level ranking API (RankingCriteria, RankingEngine) # Provide a thin compatibility layer if needed to satisfy these tests. from datetime import UTC, datetime, timedelta from typing import Any import pytest from src.query.ranking import RankingCriteria, RankingEngine @pytest.fixture def ranking_engine() -> RankingEngine: """Create a ranking engine instance.""" return RankingEngine() @pytest.fixture def sample_search_results() -> list[dict[str, Any]]: """Create sample search results for testing.""" return [ { "file_id": 1, "file_path": "src/core/database.py", "function_name": "connect_to_database", "similarity": 0.95, "complexity": 10, "lines": 50, "last_modified": datetime.now(UTC) - timedelta(days=1), "imports": ["psycopg2", "sqlalchemy"], "has_docstring": True, "is_test": False, }, { "file_id": 2, "file_path": "tests/test_database.py", "function_name": "test_connection", "similarity": 0.90, "complexity": 3, "lines": 20, "last_modified": datetime.now(UTC) - timedelta(days=30), "imports": ["pytest", "unittest"], "has_docstring": False, "is_test": True, }, { "file_id": 3, "file_path": "src/utils/db_helper.py", "function_name": "get_connection", "similarity": 0.85, "complexity": 15, "lines": 100, "last_modified": datetime.now(UTC), "imports": ["psycopg2", "logging"], "has_docstring": True, "is_test": False, }, ] class TestRankingEngine: """Test cases for RankingEngine class.""" def test_rank_by_similarity( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test ranking by similarity score.""" # Arrange criteria = RankingCriteria( similarity_weight=1.0, complexity_weight=0.0, recency_weight=0.0, documentation_weight=0.0, ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert assert len(ranked) == 3 assert ranked[0]["file_id"] == 1 # Highest similarity assert ranked[1]["file_id"] == 2 assert ranked[2]["file_id"] == 3 assert ranked[0]["ranking_score"] > ranked[1]["ranking_score"] assert ranked[1]["ranking_score"] > ranked[2]["ranking_score"] def test_rank_by_complexity( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test ranking by code complexity (prefer simpler code).""" # Arrange criteria = RankingCriteria( similarity_weight=0.0, complexity_weight=1.0, recency_weight=0.0, documentation_weight=0.0, prefer_simple=True, ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert assert ranked[0]["file_id"] == 2 # Lowest complexity (3) assert ranked[1]["file_id"] == 1 # Medium complexity (10) assert ranked[2]["file_id"] == 3 # Highest complexity (15) def test_rank_by_recency( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test ranking by recency (prefer newer code).""" # Arrange criteria = RankingCriteria( similarity_weight=0.0, complexity_weight=0.0, recency_weight=1.0, documentation_weight=0.0, ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert assert ranked[0]["file_id"] == 3 # Most recent (today) assert ranked[1]["file_id"] == 1 # Yesterday assert ranked[2]["file_id"] == 2 # 30 days ago def test_rank_by_documentation( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test ranking by documentation quality.""" # Arrange criteria = RankingCriteria( similarity_weight=0.0, complexity_weight=0.0, recency_weight=0.0, documentation_weight=1.0, ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert # Files with docstrings should rank higher assert ranked[0]["has_docstring"] is True assert ranked[1]["has_docstring"] is True assert ranked[2]["has_docstring"] is False def test_combined_ranking( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test ranking with multiple criteria.""" # Arrange criteria = RankingCriteria( similarity_weight=0.4, complexity_weight=0.2, recency_weight=0.2, documentation_weight=0.2, ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert assert len(ranked) == 3 # Each result should have a ranking score for result in ranked: assert "ranking_score" in result assert 0 <= result["ranking_score"] <= 1 def test_filter_test_files( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test filtering out test files.""" # Arrange criteria = RankingCriteria(exclude_tests=True) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert assert len(ranked) == 2 assert all(not r["is_test"] for r in ranked) def test_minimum_similarity_threshold( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test filtering by minimum similarity threshold.""" # Arrange criteria = RankingCriteria(min_similarity=0.9) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert assert len(ranked) == 2 assert all(r["similarity"] >= 0.9 for r in ranked) def test_file_type_preference( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test preferring certain file types.""" # Arrange criteria = RankingCriteria( preferred_paths=["src/core/"], path_weight=0.5, similarity_weight=0.5 ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert # src/core/database.py should rank higher due to path preference assert ranked[0]["file_path"] == "src/core/database.py" def test_boost_by_imports( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test boosting results that use specific imports.""" # Arrange criteria = RankingCriteria( boost_imports=["sqlalchemy"], import_boost_factor=2.0 ) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert # Result with sqlalchemy import should rank higher assert ranked[0]["file_id"] == 1 assert "sqlalchemy" in ranked[0]["imports"] def test_empty_results(self, ranking_engine: RankingEngine) -> None: """Test ranking empty results.""" # Act ranked = ranking_engine.rank_results([], RankingCriteria()) # Assert assert ranked == [] def test_single_result(self, ranking_engine: RankingEngine) -> None: """Test ranking a single result.""" # Arrange single_result = [{"file_id": 1, "similarity": 0.9, "complexity": 5}] # Act ranked = ranking_engine.rank_results(single_result, RankingCriteria()) # Assert assert len(ranked) == 1 assert ranked[0]["file_id"] == 1 assert "ranking_score" in ranked[0] def test_normalize_scores(self, ranking_engine: RankingEngine) -> None: """Test score normalization.""" # Arrange results = [ {"file_id": 1, "similarity": 0.5, "lines": 1000}, {"file_id": 2, "similarity": 0.7, "lines": 100}, {"file_id": 3, "similarity": 0.9, "lines": 10}, ] criteria = RankingCriteria( similarity_weight=0.5, size_weight=0.5, prefer_concise=True ) # Act ranked = ranking_engine.rank_results(results, criteria) # Assert # All scores should be between 0 and 1 for result in ranked: assert 0 <= result["ranking_score"] <= 1 def test_custom_scoring_function( self, ranking_engine: RankingEngine, sample_search_results: list[dict[str, Any]] ) -> None: """Test using a custom scoring function.""" # Arrange def custom_scorer(result: dict[str, Any]) -> float: # Prefer files with exactly 50 lines lines_val = float(result.get("lines", 0) or 0) distance_from_50 = abs(lines_val - 50.0) return 1.0 / (1.0 + distance_from_50) criteria = RankingCriteria(custom_scorer=custom_scorer) # Act ranked = ranking_engine.rank_results(sample_search_results, criteria) # Assert # File with 50 lines should rank first assert ranked[0]["lines"] == 50 def test_diversity_ranking(self, ranking_engine: RankingEngine) -> None: """Test promoting diversity in results.""" # Arrange results = [ {"file_id": 1, "file_path": "src/db/conn.py", "similarity": 0.95}, {"file_id": 2, "file_path": "src/db/pool.py", "similarity": 0.94}, {"file_id": 3, "file_path": "src/api/handler.py", "similarity": 0.93}, {"file_id": 4, "file_path": "src/db/manager.py", "similarity": 0.92}, ] criteria = RankingCriteria(promote_diversity=True, diversity_factor=0.3) # Act ranked = ranking_engine.rank_results(results, criteria) # Assert # Results from different directories should be interleaved paths = [r["file_path"] for r in ranked[:2]] assert not all("src/db/" in p for p in paths) def test_invalid_weights(self, ranking_engine: RankingEngine) -> None: """Test handling of invalid weight values.""" # Arrange criteria = RankingCriteria( similarity_weight=-0.5, # Invalid negative weight complexity_weight=2.0, # Weight > 1 ) # Act & Assert with pytest.raises(ValueError, match="Weights must be between 0 and 1"): ranking_engine.validate_criteria(criteria) def test_performance_with_large_dataset( self, ranking_engine: RankingEngine ) -> None: """Test ranking performance with many results.""" # Arrange large_results = [ { "file_id": i, "similarity": 0.9 - (i * 0.0001), "complexity": i % 20, "lines": (i * 10) % 500, "last_modified": datetime.now(UTC) - timedelta(days=i % 365), } for i in range(1000) ] criteria = RankingCriteria( similarity_weight=0.4, complexity_weight=0.3, recency_weight=0.3 ) # Act import time start_time = time.time() ranked = ranking_engine.rank_results(large_results, criteria) end_time = time.time() # Assert assert len(ranked) == 1000 assert end_time - start_time < 1.0 # Should complete within 1 second