Add comprehensive tests for ast-grep and tree-sitter relationship extraction

- Introduced test suite for AstGrepPythonProcessor covering pattern definitions, parsing, and relationship extraction.
- Added comparison tests between tree-sitter and ast-grep for consistency in relationship extraction.
- Implemented tests for ast-grep binding module to verify functionality and availability.
- Ensured tests cover various scenarios including inheritance, function calls, and imports.
This commit is contained in:
catlog22
2026-02-15 21:14:14 +08:00
parent 126a357aa2
commit 48a6a1f2aa
56 changed files with 10622 additions and 374 deletions

View File

@@ -0,0 +1 @@
"""Tests for codexlens.parsers modules."""

View File

@@ -0,0 +1,444 @@
"""Tests for dedicated extraction methods: extract_inherits, extract_calls, extract_imports.
Tests pattern-based relationship extraction from Python source code
using ast-grep-py bindings for INHERITS, CALL, and IMPORTS relationships.
"""
from pathlib import Path
import pytest
from codexlens.parsers.astgrep_processor import (
AstGrepPythonProcessor,
is_astgrep_processor_available,
)
from codexlens.entities import RelationshipType
# Check if ast-grep is available for conditional test skipping
ASTGREP_AVAILABLE = is_astgrep_processor_available()
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestExtractInherits:
"""Tests for extract_inherits method - INHERITS relationship extraction."""
def test_single_inheritance(self):
"""Test extraction of single inheritance relationship."""
processor = AstGrepPythonProcessor()
code = """
class Animal:
pass
class Dog(Animal):
pass
"""
relationships = processor.extract_inherits(code, "test.py")
assert len(relationships) == 1
rel = relationships[0]
assert rel.source_symbol == "Dog"
assert rel.target_symbol == "Animal"
assert rel.relationship_type == RelationshipType.INHERITS
def test_multiple_inheritance(self):
"""Test extraction of multiple inheritance relationships."""
processor = AstGrepPythonProcessor()
code = """
class A:
pass
class B:
pass
class C(A, B):
pass
"""
relationships = processor.extract_inherits(code, "test.py")
# Should have 2 relationships: C->A and C->B
assert len(relationships) == 2
targets = {r.target_symbol for r in relationships}
assert "A" in targets
assert "B" in targets
for rel in relationships:
assert rel.source_symbol == "C"
def test_no_inheritance(self):
"""Test that classes without inheritance return empty list."""
processor = AstGrepPythonProcessor()
code = """
class Standalone:
pass
"""
relationships = processor.extract_inherits(code, "test.py")
assert len(relationships) == 0
def test_nested_class_inheritance(self):
"""Test extraction of inheritance in nested classes."""
processor = AstGrepPythonProcessor()
code = """
class Outer:
class Inner(Base):
pass
"""
relationships = processor.extract_inherits(code, "test.py")
assert len(relationships) == 1
assert relationships[0].source_symbol == "Inner"
assert relationships[0].target_symbol == "Base"
def test_inheritance_with_complex_bases(self):
"""Test extraction with generic or complex base classes."""
processor = AstGrepPythonProcessor()
code = """
class Service(BaseService, mixins.Loggable):
pass
"""
relationships = processor.extract_inherits(code, "test.py")
assert len(relationships) == 2
targets = {r.target_symbol for r in relationships}
assert "BaseService" in targets
assert "mixins.Loggable" in targets
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestExtractCalls:
"""Tests for extract_calls method - CALL relationship extraction."""
def test_simple_function_call(self):
"""Test extraction of simple function calls."""
processor = AstGrepPythonProcessor()
code = """
def main():
print("hello")
len([1, 2, 3])
"""
relationships = processor.extract_calls(code, "test.py", "main")
targets = {r.target_symbol for r in relationships}
assert "print" in targets
assert "len" in targets
def test_method_call(self):
"""Test extraction of method calls."""
processor = AstGrepPythonProcessor()
code = """
def process():
obj.method()
items.append(1)
"""
relationships = processor.extract_calls(code, "test.py", "process")
targets = {r.target_symbol for r in relationships}
assert "obj.method" in targets
assert "items.append" in targets
def test_skips_self_calls(self):
"""Test that self.method() calls are filtered."""
processor = AstGrepPythonProcessor()
code = """
class Service:
def process(self):
self.internal()
external_func()
"""
relationships = processor.extract_calls(code, "test.py", "Service")
targets = {r.target_symbol for r in relationships}
# self.internal should be filtered
assert "self.internal" not in targets
assert "internal" not in targets
assert "external_func" in targets
def test_skips_cls_calls(self):
"""Test that cls.method() calls are filtered."""
processor = AstGrepPythonProcessor()
code = """
class Factory:
@classmethod
def create(cls):
cls.helper()
other_func()
"""
relationships = processor.extract_calls(code, "test.py", "Factory")
targets = {r.target_symbol for r in relationships}
assert "cls.helper" not in targets
assert "other_func" in targets
def test_alias_resolution(self):
"""Test call alias resolution using import map."""
processor = AstGrepPythonProcessor()
code = """
def main():
np.array([1, 2, 3])
"""
alias_map = {"np": "numpy"}
relationships = processor.extract_calls(code, "test.py", "main", alias_map)
assert len(relationships) >= 1
# Should resolve np.array to numpy.array
assert any("numpy.array" in r.target_symbol for r in relationships)
def test_no_calls(self):
"""Test that code without calls returns empty list."""
processor = AstGrepPythonProcessor()
code = """
x = 1
y = x + 2
"""
relationships = processor.extract_calls(code, "test.py")
assert len(relationships) == 0
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestExtractImports:
"""Tests for extract_imports method - IMPORTS relationship extraction."""
def test_simple_import(self):
"""Test extraction of simple import statements."""
processor = AstGrepPythonProcessor()
code = "import os"
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) == 1
assert relationships[0].target_symbol == "os"
assert relationships[0].relationship_type == RelationshipType.IMPORTS
assert alias_map.get("os") == "os"
def test_import_with_alias(self):
"""Test extraction of import with alias."""
processor = AstGrepPythonProcessor()
code = "import numpy as np"
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) == 1
assert relationships[0].target_symbol == "numpy"
assert alias_map.get("np") == "numpy"
def test_from_import(self):
"""Test extraction of from-import statements."""
processor = AstGrepPythonProcessor()
code = "from typing import List, Dict"
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) == 1
assert relationships[0].target_symbol == "typing"
assert alias_map.get("List") == "typing.List"
assert alias_map.get("Dict") == "typing.Dict"
def test_from_import_with_alias(self):
"""Test extraction of from-import with alias."""
processor = AstGrepPythonProcessor()
code = "from collections import defaultdict as dd"
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) == 1
# The alias map should map dd to collections.defaultcount
assert "dd" in alias_map
assert "defaultdict" in alias_map.get("dd", "")
def test_star_import(self):
"""Test extraction of star imports."""
processor = AstGrepPythonProcessor()
code = "from module import *"
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) >= 1
# Star import should be recorded
star_imports = [r for r in relationships if "*" in r.target_symbol]
assert len(star_imports) >= 1
def test_relative_import(self):
"""Test extraction of relative imports."""
processor = AstGrepPythonProcessor()
code = "from .utils import helper"
relationships, alias_map = processor.extract_imports(code, "test.py")
# Should capture the relative import
assert len(relationships) >= 1
rel_imports = [r for r in relationships if r.target_symbol.startswith(".")]
assert len(rel_imports) >= 1
def test_multiple_imports(self):
"""Test extraction of multiple import types."""
processor = AstGrepPythonProcessor()
code = """
import os
import sys
from typing import List
from collections import defaultdict as dd
"""
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) >= 4
targets = {r.target_symbol for r in relationships}
assert "os" in targets
assert "sys" in targets
assert "typing" in targets
assert "collections" in targets
def test_no_imports(self):
"""Test that code without imports returns empty list."""
processor = AstGrepPythonProcessor()
code = """
x = 1
def foo():
pass
"""
relationships, alias_map = processor.extract_imports(code, "test.py")
assert len(relationships) == 0
assert len(alias_map) == 0
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestExtractMethodsIntegration:
"""Integration tests combining multiple extraction methods."""
def test_full_file_extraction(self):
"""Test extracting all relationships from a complete file."""
processor = AstGrepPythonProcessor()
code = """
import os
from typing import List, Optional
class Base:
pass
class Service(Base):
def __init__(self):
self.data = []
def process(self):
result = os.path.join("a", "b")
items = List([1, 2, 3])
return result
def main():
svc = Service()
svc.process()
"""
source_file = "test.py"
# Extract all relationship types
imports, alias_map = processor.extract_imports(code, source_file)
inherits = processor.extract_inherits(code, source_file)
calls = processor.extract_calls(code, source_file, alias_map=alias_map)
# Verify we got all expected relationships
assert len(imports) >= 2 # os and typing
assert len(inherits) == 1 # Service -> Base
assert len(calls) >= 2 # os.path.join and others
# Verify inheritance
assert any(r.source_symbol == "Service" and r.target_symbol == "Base"
for r in inherits)
def test_alias_propagation(self):
"""Test that import aliases propagate to call resolution."""
processor = AstGrepPythonProcessor()
code = """
import numpy as np
def compute():
arr = np.array([1, 2, 3])
return np.sum(arr)
"""
source_file = "test.py"
imports, alias_map = processor.extract_imports(code, source_file)
calls = processor.extract_calls(code, source_file, alias_map=alias_map)
# Alias map should have np -> numpy
assert alias_map.get("np") == "numpy"
# Calls should resolve np.array and np.sum
resolved_targets = {r.target_symbol for r in calls}
# At minimum, np.array and np.sum should be captured
np_calls = [t for t in resolved_targets if "np" in t or "numpy" in t]
assert len(np_calls) >= 2
class TestExtractMethodFallback:
"""Tests for fallback behavior when ast-grep unavailable."""
def test_extract_inherits_empty_when_unavailable(self):
"""Test extract_inherits returns empty list when unavailable."""
processor = AstGrepPythonProcessor()
if not processor.is_available():
code = "class Dog(Animal): pass"
relationships = processor.extract_inherits(code, "test.py")
assert relationships == []
def test_extract_calls_empty_when_unavailable(self):
"""Test extract_calls returns empty list when unavailable."""
processor = AstGrepPythonProcessor()
if not processor.is_available():
code = "print('hello')"
relationships = processor.extract_calls(code, "test.py")
assert relationships == []
def test_extract_imports_empty_when_unavailable(self):
"""Test extract_imports returns empty tuple when unavailable."""
processor = AstGrepPythonProcessor()
if not processor.is_available():
code = "import os"
relationships, alias_map = processor.extract_imports(code, "test.py")
assert relationships == []
assert alias_map == {}
class TestHelperMethods:
"""Tests for internal helper methods."""
def test_parse_base_classes_single(self):
"""Test _parse_base_classes with single base."""
processor = AstGrepPythonProcessor()
result = processor._parse_base_classes("BaseClass")
assert result == ["BaseClass"]
def test_parse_base_classes_multiple(self):
"""Test _parse_base_classes with multiple bases."""
processor = AstGrepPythonProcessor()
result = processor._parse_base_classes("A, B, C")
assert result == ["A", "B", "C"]
def test_parse_base_classes_with_generics(self):
"""Test _parse_base_classes with generic types."""
processor = AstGrepPythonProcessor()
result = processor._parse_base_classes("Generic[T], Mixin")
assert "Generic[T]" in result
assert "Mixin" in result
def test_resolve_call_alias_simple(self):
"""Test _resolve_call_alias with simple name."""
processor = AstGrepPythonProcessor()
alias_map = {"np": "numpy"}
result = processor._resolve_call_alias("np", alias_map)
assert result == "numpy"
def test_resolve_call_alias_qualified(self):
"""Test _resolve_call_alias with qualified name."""
processor = AstGrepPythonProcessor()
alias_map = {"np": "numpy"}
result = processor._resolve_call_alias("np.array", alias_map)
assert result == "numpy.array"
def test_resolve_call_alias_no_match(self):
"""Test _resolve_call_alias when no alias exists."""
processor = AstGrepPythonProcessor()
alias_map = {}
result = processor._resolve_call_alias("myfunc", alias_map)
assert result == "myfunc"

View File

@@ -0,0 +1,402 @@
"""Tests for AstGrepPythonProcessor.
Tests pattern-based relationship extraction from Python source code
using ast-grep-py bindings.
"""
from pathlib import Path
import pytest
from codexlens.parsers.astgrep_processor import (
AstGrepPythonProcessor,
BaseAstGrepProcessor,
is_astgrep_processor_available,
)
from codexlens.parsers.patterns.python import (
PYTHON_PATTERNS,
METAVARS,
RELATIONSHIP_PATTERNS,
get_pattern,
get_patterns_for_relationship,
get_metavar,
)
# Check if ast-grep is available for conditional test skipping
ASTGREP_AVAILABLE = is_astgrep_processor_available()
class TestPatternDefinitions:
"""Tests for Python pattern definitions."""
def test_python_patterns_exist(self):
"""Verify all expected patterns are defined."""
expected_patterns = [
"class_def",
"class_with_bases",
"func_def",
"async_func_def",
"import_stmt",
"import_from",
"call",
"method_call",
]
for pattern_name in expected_patterns:
assert pattern_name in PYTHON_PATTERNS, f"Missing pattern: {pattern_name}"
def test_get_pattern_returns_correct_pattern(self):
"""Test get_pattern returns expected pattern strings."""
# Note: ast-grep-py 0.40+ uses $$$ for zero-or-more multi-match
assert get_pattern("class_def") == "class $NAME $$$BODY"
assert get_pattern("func_def") == "def $NAME($$$PARAMS): $$$BODY"
assert get_pattern("import_stmt") == "import $MODULE"
def test_get_pattern_raises_for_unknown(self):
"""Test get_pattern raises KeyError for unknown patterns."""
with pytest.raises(KeyError):
get_pattern("nonexistent_pattern")
def test_metavars_defined(self):
"""Verify metavariable mappings are defined."""
expected_metavars = [
"class_name",
"func_name",
"import_module",
"call_func",
]
for var in expected_metavars:
assert var in METAVARS, f"Missing metavar: {var}"
def test_get_metavar(self):
"""Test get_metavar returns correct values."""
assert get_metavar("class_name") == "NAME"
assert get_metavar("func_name") == "NAME"
assert get_metavar("import_module") == "MODULE"
def test_relationship_patterns_mapping(self):
"""Test relationship type to pattern mapping."""
assert "class_with_bases" in get_patterns_for_relationship("inheritance")
assert "import_stmt" in get_patterns_for_relationship("imports")
assert "import_from" in get_patterns_for_relationship("imports")
assert "call" in get_patterns_for_relationship("calls")
class TestAstGrepPythonProcessorAvailability:
"""Tests for processor availability."""
def test_is_available_returns_bool(self):
"""Test is_available returns a boolean."""
processor = AstGrepPythonProcessor()
assert isinstance(processor.is_available(), bool)
def test_is_available_matches_global_check(self):
"""Test is_available matches is_astgrep_processor_available."""
processor = AstGrepPythonProcessor()
assert processor.is_available() == is_astgrep_processor_available()
def test_module_level_check(self):
"""Test module-level availability function."""
assert isinstance(is_astgrep_processor_available(), bool)
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestAstGrepPythonProcessorParsing:
"""Tests for Python parsing with ast-grep."""
def test_parse_simple_function(self):
"""Test parsing a simple function definition."""
processor = AstGrepPythonProcessor()
code = "def hello():\n pass"
result = processor.parse(code, Path("test.py"))
assert result is not None
assert result.language == "python"
assert len(result.symbols) == 1
assert result.symbols[0].name == "hello"
assert result.symbols[0].kind == "function"
def test_parse_class(self):
"""Test parsing a class definition."""
processor = AstGrepPythonProcessor()
code = "class MyClass:\n pass"
result = processor.parse(code, Path("test.py"))
assert result is not None
assert len(result.symbols) == 1
assert result.symbols[0].name == "MyClass"
assert result.symbols[0].kind == "class"
def test_parse_async_function(self):
"""Test parsing an async function definition."""
processor = AstGrepPythonProcessor()
code = "async def fetch_data():\n pass"
result = processor.parse(code, Path("test.py"))
assert result is not None
assert len(result.symbols) == 1
assert result.symbols[0].name == "fetch_data"
def test_parse_class_with_inheritance(self):
"""Test parsing class with inheritance."""
processor = AstGrepPythonProcessor()
code = """
class Base:
pass
class Child(Base):
pass
"""
result = processor.parse(code, Path("test.py"))
assert result is not None
names = [s.name for s in result.symbols]
assert "Base" in names
assert "Child" in names
# Check inheritance relationship
inherits = [
r for r in result.relationships
if r.relationship_type.value == "inherits"
]
assert any(r.source_symbol == "Child" for r in inherits)
def test_parse_imports(self):
"""Test parsing import statements."""
processor = AstGrepPythonProcessor()
code = """
import os
from sys import path
"""
result = processor.parse(code, Path("test.py"))
assert result is not None
imports = [
r for r in result.relationships
if r.relationship_type.value == "imports"
]
assert len(imports) >= 1
targets = {r.target_symbol for r in imports}
assert "os" in targets
def test_parse_function_calls(self):
"""Test parsing function calls."""
processor = AstGrepPythonProcessor()
code = """
def main():
print("hello")
len([1, 2, 3])
"""
result = processor.parse(code, Path("test.py"))
assert result is not None
calls = [
r for r in result.relationships
if r.relationship_type.value == "calls"
]
targets = {r.target_symbol for r in calls}
assert "print" in targets
assert "len" in targets
def test_parse_empty_file(self):
"""Test parsing an empty file."""
processor = AstGrepPythonProcessor()
result = processor.parse("", Path("test.py"))
assert result is not None
assert len(result.symbols) == 0
def test_parse_returns_indexed_file(self):
"""Test that parse returns proper IndexedFile structure."""
processor = AstGrepPythonProcessor()
code = "def test():\n pass"
result = processor.parse(code, Path("test.py"))
assert result is not None
assert result.path.endswith("test.py")
assert result.language == "python"
assert isinstance(result.symbols, list)
assert isinstance(result.chunks, list)
assert isinstance(result.relationships, list)
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestAstGrepPythonProcessorRelationships:
"""Tests for relationship extraction."""
def test_inheritance_extraction(self):
"""Test extraction of inheritance relationships."""
processor = AstGrepPythonProcessor()
code = """
class Animal:
pass
class Dog(Animal):
pass
class Cat(Animal):
pass
"""
result = processor.parse(code, Path("test.py"))
assert result is not None
inherits = [
r for r in result.relationships
if r.relationship_type.value == "inherits"
]
# Should have 2 inheritance relationships
assert len(inherits) >= 2
sources = {r.source_symbol for r in inherits}
assert "Dog" in sources
assert "Cat" in sources
def test_call_extraction_skips_self(self):
"""Test that self.method() calls are filtered."""
processor = AstGrepPythonProcessor()
code = """
class Service:
def process(self):
self.internal()
external_call()
def external_call():
pass
"""
result = processor.parse(code, Path("test.py"))
assert result is not None
calls = [
r for r in result.relationships
if r.relationship_type.value == "calls"
]
targets = {r.target_symbol for r in calls}
# self.internal should be filtered
assert "self.internal" not in targets
assert "external_call" in targets
def test_import_with_alias_resolution(self):
"""Test import alias resolution in calls."""
processor = AstGrepPythonProcessor()
code = """
import os.path as osp
def main():
osp.join("a", "b")
"""
result = processor.parse(code, Path("test.py"))
assert result is not None
calls = [
r for r in result.relationships
if r.relationship_type.value == "calls"
]
targets = {r.target_symbol for r in calls}
# Should resolve osp to os.path
assert any("os.path" in t for t in targets)
@pytest.mark.skipif(not ASTGREP_AVAILABLE, reason="ast-grep-py not installed")
class TestAstGrepPythonProcessorRunAstGrep:
"""Tests for run_ast_grep method."""
def test_run_ast_grep_returns_list(self):
"""Test run_ast_grep returns a list."""
processor = AstGrepPythonProcessor()
code = "def hello():\n pass"
processor._binding.parse(code) if processor._binding else None
matches = processor.run_ast_grep(code, "def $NAME($$$PARAMS) $$$BODY")
assert isinstance(matches, list)
def test_run_ast_grep_finds_matches(self):
"""Test run_ast_grep finds expected matches."""
processor = AstGrepPythonProcessor()
code = "def hello():\n pass"
matches = processor.run_ast_grep(code, "def $NAME($$$PARAMS) $$$BODY")
assert len(matches) >= 1
def test_run_ast_grep_empty_code(self):
"""Test run_ast_grep with empty code."""
processor = AstGrepPythonProcessor()
matches = processor.run_ast_grep("", "def $NAME($$$PARAMS) $$$BODY")
assert matches == []
def test_run_ast_grep_no_matches(self):
"""Test run_ast_grep when pattern doesn't match."""
processor = AstGrepPythonProcessor()
code = "x = 1"
matches = processor.run_ast_grep(code, "class $NAME $$$BODY")
assert matches == []
class TestAstGrepPythonProcessorFallback:
"""Tests for fallback behavior when ast-grep unavailable."""
def test_parse_returns_none_when_unavailable(self):
"""Test parse returns None when ast-grep unavailable."""
# This test runs regardless of availability
# When unavailable, should gracefully return None
processor = AstGrepPythonProcessor()
if not processor.is_available():
code = "def test():\n pass"
result = processor.parse(code, Path("test.py"))
assert result is None
def test_run_ast_grep_empty_when_unavailable(self):
"""Test run_ast_grep returns empty list when unavailable."""
processor = AstGrepPythonProcessor()
if not processor.is_available():
matches = processor.run_ast_grep("code", "pattern")
assert matches == []
class TestBaseAstGrepProcessor:
"""Tests for abstract base class."""
def test_cannot_instantiate_base_class(self):
"""Test that BaseAstGrepProcessor cannot be instantiated directly."""
with pytest.raises(TypeError):
BaseAstGrepProcessor("python") # type: ignore[abstract]
def test_subclass_implements_abstract_methods(self):
"""Test that AstGrepPythonProcessor implements all abstract methods."""
processor = AstGrepPythonProcessor()
# Should have process_matches method
assert hasattr(processor, "process_matches")
# Should have parse method
assert hasattr(processor, "parse")
# Check methods are callable
assert callable(processor.process_matches)
assert callable(processor.parse)
class TestPatternIntegration:
"""Tests for pattern module integration with processor."""
def test_processor_uses_pattern_module(self):
"""Verify processor uses patterns from pattern module."""
# The processor should import and use patterns from patterns/python/
from codexlens.parsers.astgrep_processor import get_pattern
# Verify pattern access works
assert get_pattern("class_def") is not None
assert get_pattern("func_def") is not None
def test_pattern_consistency(self):
"""Test pattern definitions are consistent."""
# Patterns used by processor should exist in pattern module
patterns_needed = [
"class_def",
"class_with_bases",
"func_def",
"async_func_def",
"import_stmt",
"import_from",
"call",
]
for pattern_name in patterns_needed:
# Should not raise KeyError
pattern = get_pattern(pattern_name)
assert pattern is not None
assert len(pattern) > 0

View File

@@ -0,0 +1,526 @@
"""Comparison tests for tree-sitter vs ast-grep Python relationship extraction.
Validates that both parsers produce consistent output for Python relationship
extraction (INHERITS, CALL, IMPORTS).
"""
from __future__ import annotations
from pathlib import Path
from typing import List, Set, Tuple
import pytest
from codexlens.config import Config
from codexlens.entities import CodeRelationship, RelationshipType
from codexlens.parsers.treesitter_parser import TreeSitterSymbolParser
# Sample Python code for testing relationship extraction
SAMPLE_PYTHON_CODE = '''
"""Module docstring."""
import os
import sys
from typing import List, Dict, Optional
from collections import defaultdict as dd
from pathlib import Path as PPath
class BaseClass:
"""Base class."""
def base_method(self):
pass
def another_method(self):
return self.base_method()
class Mixin:
"""Mixin class."""
def mixin_func(self):
return "mixin"
class ChildClass(BaseClass, Mixin):
"""Child class with multiple inheritance."""
def __init__(self):
super().__init__()
self.data = dd(list)
def process(self, items: List[str]) -> Dict[str, int]:
result = {}
for item in items:
result[item] = len(item)
return result
def call_external(self, path: str) -> Optional[str]:
p = PPath(path)
if p.exists():
return str(p.read_text())
return None
def standalone_function():
"""Standalone function."""
data = [1, 2, 3]
return sum(data)
async def async_function():
"""Async function."""
import asyncio
await asyncio.sleep(1)
'''
def relationship_to_tuple(rel: CodeRelationship) -> Tuple[str, str, str, int]:
"""Convert relationship to a comparable tuple.
Returns:
(source_symbol, target_symbol, relationship_type, source_line)
"""
return (
rel.source_symbol,
rel.target_symbol,
rel.relationship_type.value,
rel.source_line,
)
def extract_relationship_tuples(
relationships: List[CodeRelationship],
) -> Set[Tuple[str, str, str]]:
"""Extract relationship tuples without line numbers for comparison.
Returns:
Set of (source_symbol, target_symbol, relationship_type) tuples
"""
return {
(rel.source_symbol, rel.target_symbol, rel.relationship_type.value)
for rel in relationships
}
def filter_by_type(
relationships: List[CodeRelationship],
rel_type: RelationshipType,
) -> List[CodeRelationship]:
"""Filter relationships by type."""
return [r for r in relationships if r.relationship_type == rel_type]
class TestTreeSitterVsAstGrepComparison:
"""Compare tree-sitter and ast-grep Python relationship extraction."""
@pytest.fixture
def sample_path(self, tmp_path: Path) -> Path:
"""Create a temporary Python file with sample code."""
py_file = tmp_path / "sample.py"
py_file.write_text(SAMPLE_PYTHON_CODE)
return py_file
@pytest.fixture
def ts_parser_default(self) -> TreeSitterSymbolParser:
"""Create tree-sitter parser with default config (use_astgrep=False)."""
config = Config()
assert config.use_astgrep is False
return TreeSitterSymbolParser("python", config=config)
@pytest.fixture
def ts_parser_astgrep(self) -> TreeSitterSymbolParser:
"""Create tree-sitter parser with ast-grep enabled."""
config = Config()
config.use_astgrep = True
return TreeSitterSymbolParser("python", config=config)
def test_parser_availability(self, ts_parser_default: TreeSitterSymbolParser) -> None:
"""Test that tree-sitter parser is available."""
assert ts_parser_default.is_available()
def test_astgrep_processor_initialization(
self, ts_parser_astgrep: TreeSitterSymbolParser
) -> None:
"""Test that ast-grep processor is initialized when config enables it."""
# The processor should be initialized (may be None if ast-grep-py not installed)
# This test just verifies the initialization path works
assert ts_parser_astgrep._config is not None
assert ts_parser_astgrep._config.use_astgrep is True
def _skip_if_astgrep_unavailable(
self, ts_parser_astgrep: TreeSitterSymbolParser
) -> None:
"""Skip test if ast-grep is not available."""
if ts_parser_astgrep._astgrep_processor is None:
pytest.skip("ast-grep-py not installed")
def test_parse_returns_valid_result(
self,
ts_parser_default: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test that parsing returns a valid IndexedFile."""
source_code = sample_path.read_text()
result = ts_parser_default.parse(source_code, sample_path)
assert result is not None
assert result.language == "python"
assert len(result.symbols) > 0
assert len(result.relationships) > 0
def test_extracted_symbols_match(
self,
ts_parser_default: TreeSitterSymbolParser,
ts_parser_astgrep: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test that both parsers extract similar symbols."""
self._skip_if_astgrep_unavailable(ts_parser_astgrep)
source_code = sample_path.read_text()
result_ts = ts_parser_default.parse(source_code, sample_path)
result_astgrep = ts_parser_astgrep.parse(source_code, sample_path)
assert result_ts is not None
assert result_astgrep is not None
# Compare symbol names
ts_symbols = {s.name for s in result_ts.symbols}
astgrep_symbols = {s.name for s in result_astgrep.symbols}
# Should have the same symbols (classes, functions, methods)
assert ts_symbols == astgrep_symbols
def test_inheritance_relationships(
self,
ts_parser_default: TreeSitterSymbolParser,
ts_parser_astgrep: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test INHERITS relationship extraction consistency."""
self._skip_if_astgrep_unavailable(ts_parser_astgrep)
source_code = sample_path.read_text()
result_ts = ts_parser_default.parse(source_code, sample_path)
result_astgrep = ts_parser_astgrep.parse(source_code, sample_path)
assert result_ts is not None
assert result_astgrep is not None
# Extract inheritance relationships
ts_inherits = filter_by_type(result_ts.relationships, RelationshipType.INHERITS)
astgrep_inherits = filter_by_type(
result_astgrep.relationships, RelationshipType.INHERITS
)
ts_tuples = extract_relationship_tuples(ts_inherits)
astgrep_tuples = extract_relationship_tuples(astgrep_inherits)
# Both should detect ChildClass(BaseClass, Mixin)
assert ts_tuples == astgrep_tuples
# Verify specific inheritance relationships
expected_inherits = {
("ChildClass", "BaseClass", "inherits"),
("ChildClass", "Mixin", "inherits"),
}
assert ts_tuples == expected_inherits
def test_import_relationships(
self,
ts_parser_default: TreeSitterSymbolParser,
ts_parser_astgrep: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test IMPORTS relationship extraction consistency."""
self._skip_if_astgrep_unavailable(ts_parser_astgrep)
source_code = sample_path.read_text()
result_ts = ts_parser_default.parse(source_code, sample_path)
result_astgrep = ts_parser_astgrep.parse(source_code, sample_path)
assert result_ts is not None
assert result_astgrep is not None
# Extract import relationships
ts_imports = filter_by_type(result_ts.relationships, RelationshipType.IMPORTS)
astgrep_imports = filter_by_type(
result_astgrep.relationships, RelationshipType.IMPORTS
)
ts_tuples = extract_relationship_tuples(ts_imports)
astgrep_tuples = extract_relationship_tuples(astgrep_imports)
# Compare - should be similar (may differ in exact module representation)
# At minimum, both should detect the top-level imports
ts_modules = {t[1].split(".")[0] for t in ts_tuples}
astgrep_modules = {t[1].split(".")[0] for t in astgrep_tuples}
# Should have imports from: os, sys, typing, collections, pathlib
expected_modules = {"os", "sys", "typing", "collections", "pathlib", "asyncio"}
assert ts_modules >= expected_modules or astgrep_modules >= expected_modules
def test_call_relationships(
self,
ts_parser_default: TreeSitterSymbolParser,
ts_parser_astgrep: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test CALL relationship extraction consistency."""
self._skip_if_astgrep_unavailable(ts_parser_astgrep)
source_code = sample_path.read_text()
result_ts = ts_parser_default.parse(source_code, sample_path)
result_astgrep = ts_parser_astgrep.parse(source_code, sample_path)
assert result_ts is not None
assert result_astgrep is not None
# Extract call relationships
ts_calls = filter_by_type(result_ts.relationships, RelationshipType.CALL)
astgrep_calls = filter_by_type(
result_astgrep.relationships, RelationshipType.CALL
)
# Calls may differ due to scope tracking differences
# Just verify both parsers find call relationships
assert len(ts_calls) > 0
assert len(astgrep_calls) > 0
# Verify specific calls that should be detected
ts_call_targets = {r.target_symbol for r in ts_calls}
astgrep_call_targets = {r.target_symbol for r in astgrep_calls}
# Both should detect at least some common calls
# (exact match not required due to scope tracking differences)
common_targets = ts_call_targets & astgrep_call_targets
assert len(common_targets) > 0
def test_relationship_count_similarity(
self,
ts_parser_default: TreeSitterSymbolParser,
ts_parser_astgrep: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test that relationship counts are similar (>95% consistency)."""
self._skip_if_astgrep_unavailable(ts_parser_astgrep)
source_code = sample_path.read_text()
result_ts = ts_parser_default.parse(source_code, sample_path)
result_astgrep = ts_parser_astgrep.parse(source_code, sample_path)
assert result_ts is not None
assert result_astgrep is not None
ts_count = len(result_ts.relationships)
astgrep_count = len(result_astgrep.relationships)
# Calculate consistency percentage
if max(ts_count, astgrep_count) == 0:
consistency = 100.0
else:
consistency = (
min(ts_count, astgrep_count) / max(ts_count, astgrep_count) * 100
)
# Require >95% consistency
assert consistency >= 95.0, (
f"Relationship consistency {consistency:.1f}% below 95% threshold "
f"(tree-sitter: {ts_count}, ast-grep: {astgrep_count})"
)
def test_config_switch_affects_parser(
self, sample_path: Path
) -> None:
"""Test that config.use_astgrep affects which parser is used."""
config_default = Config()
config_astgrep = Config()
config_astgrep.use_astgrep = True
parser_default = TreeSitterSymbolParser("python", config=config_default)
parser_astgrep = TreeSitterSymbolParser("python", config=config_astgrep)
# Default parser should not have ast-grep processor
assert parser_default._astgrep_processor is None
# Ast-grep parser may have processor if ast-grep-py is installed
# (could be None if not installed, which is fine)
if parser_astgrep._astgrep_processor is not None:
# If available, verify it's the right type
from codexlens.parsers.astgrep_processor import AstGrepPythonProcessor
assert isinstance(
parser_astgrep._astgrep_processor, AstGrepPythonProcessor
)
def test_fallback_to_treesitter_on_astgrep_failure(
self,
ts_parser_astgrep: TreeSitterSymbolParser,
sample_path: Path,
) -> None:
"""Test that parser falls back to tree-sitter if ast-grep fails."""
source_code = sample_path.read_text()
# Even with use_astgrep=True, should get valid results
result = ts_parser_astgrep.parse(source_code, sample_path)
# Should always return a valid result (either from ast-grep or tree-sitter fallback)
assert result is not None
assert result.language == "python"
assert len(result.relationships) > 0
class TestSimpleCodeSamples:
"""Test with simple code samples for precise comparison."""
def test_simple_inheritance(self) -> None:
"""Test simple single inheritance."""
code = """
class Parent:
pass
class Child(Parent):
pass
"""
self._compare_parsers(code, expected_inherits={("Child", "Parent")})
def test_multiple_inheritance(self) -> None:
"""Test multiple inheritance."""
code = """
class A:
pass
class B:
pass
class C(A, B):
pass
"""
self._compare_parsers(
code, expected_inherits={("C", "A"), ("C", "B")}
)
def test_simple_imports(self) -> None:
"""Test simple import statements."""
code = """
import os
import sys
"""
config_ts = Config()
config_ag = Config()
config_ag.use_astgrep = True
parser_ts = TreeSitterSymbolParser("python", config=config_ts)
parser_ag = TreeSitterSymbolParser("python", config=config_ag)
tmp_path = Path("test.py")
result_ts = parser_ts.parse(code, tmp_path)
result_ag = parser_ag.parse(code, tmp_path)
assert result_ts is not None
# ast-grep result may be None if not installed
if result_ag is not None:
ts_imports = {
r.target_symbol
for r in result_ts.relationships
if r.relationship_type == RelationshipType.IMPORTS
}
ag_imports = {
r.target_symbol
for r in result_ag.relationships
if r.relationship_type == RelationshipType.IMPORTS
}
assert ts_imports == ag_imports
def test_imports_inside_function(self) -> None:
"""Test simple import inside a function scope is recorded.
Note: tree-sitter parser requires a scope to record imports.
Module-level imports without any function/class are not recorded
because scope_stack is empty at module level.
"""
code = """
def my_function():
import collections
return collections
"""
config_ts = Config()
config_ag = Config()
config_ag.use_astgrep = True
parser_ts = TreeSitterSymbolParser("python", config=config_ts)
parser_ag = TreeSitterSymbolParser("python", config=config_ag)
tmp_path = Path("test.py")
result_ts = parser_ts.parse(code, tmp_path)
result_ag = parser_ag.parse(code, tmp_path)
assert result_ts is not None
# Get import relationship targets
ts_imports = [
r.target_symbol
for r in result_ts.relationships
if r.relationship_type == RelationshipType.IMPORTS
]
# Should have collections
ts_has_collections = any("collections" in t for t in ts_imports)
assert ts_has_collections, f"Expected collections import, got: {ts_imports}"
# If ast-grep is available, verify it also finds the imports
if result_ag is not None:
ag_imports = [
r.target_symbol
for r in result_ag.relationships
if r.relationship_type == RelationshipType.IMPORTS
]
ag_has_collections = any("collections" in t for t in ag_imports)
assert ag_has_collections, f"Expected collections import in ast-grep, got: {ag_imports}"
def _compare_parsers(
self,
code: str,
expected_inherits: Set[Tuple[str, str]],
) -> None:
"""Helper to compare parser outputs for inheritance."""
config_ts = Config()
config_ag = Config()
config_ag.use_astgrep = True
parser_ts = TreeSitterSymbolParser("python", config=config_ts)
parser_ag = TreeSitterSymbolParser("python", config=config_ag)
tmp_path = Path("test.py")
result_ts = parser_ts.parse(code, tmp_path)
assert result_ts is not None
# Verify tree-sitter finds expected inheritance
ts_inherits = {
(r.source_symbol, r.target_symbol)
for r in result_ts.relationships
if r.relationship_type == RelationshipType.INHERITS
}
assert ts_inherits == expected_inherits
# If ast-grep is available, verify it matches
result_ag = parser_ag.parse(code, tmp_path)
if result_ag is not None:
ag_inherits = {
(r.source_symbol, r.target_symbol)
for r in result_ag.relationships
if r.relationship_type == RelationshipType.INHERITS
}
assert ag_inherits == expected_inherits
if __name__ == "__main__":
pytest.main([__file__, "-v"])

View File

@@ -0,0 +1,191 @@
"""Tests for ast-grep binding module.
Verifies basic import and functionality of AstGrepBinding.
Run with: python -m pytest tests/test_astgrep_binding.py -v
"""
from __future__ import annotations
import pytest
from pathlib import Path
class TestAstGrepBindingAvailability:
"""Test availability checks."""
def test_is_astgrep_available_function(self):
"""Test is_astgrep_available function returns boolean."""
from codexlens.parsers.astgrep_binding import is_astgrep_available
result = is_astgrep_available()
assert isinstance(result, bool)
def test_get_supported_languages(self):
"""Test get_supported_languages returns expected languages."""
from codexlens.parsers.astgrep_binding import get_supported_languages
languages = get_supported_languages()
assert isinstance(languages, list)
assert "python" in languages
assert "javascript" in languages
assert "typescript" in languages
class TestAstGrepBindingInit:
"""Test AstGrepBinding initialization."""
def test_init_python(self):
"""Test initialization with Python language."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("python")
assert binding.language_id == "python"
def test_init_typescript_with_tsx(self):
"""Test TSX detection from file extension."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("typescript", Path("component.tsx"))
assert binding.language_id == "typescript"
def test_is_available_returns_boolean(self):
"""Test is_available returns boolean."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("python")
result = binding.is_available()
assert isinstance(result, bool)
def _is_astgrep_installed():
"""Check if ast-grep-py is installed."""
try:
import ast_grep_py # noqa: F401
return True
except ImportError:
return False
@pytest.mark.skipif(
not _is_astgrep_installed(),
reason="ast-grep-py not installed"
)
class TestAstGrepBindingWithAstGrep:
"""Tests that require ast-grep-py to be installed."""
def test_parse_simple_python(self):
"""Test parsing simple Python code."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("python")
if not binding.is_available():
pytest.skip("ast-grep not available")
source = "x = 1"
result = binding.parse(source)
assert result is True
def test_find_inheritance(self):
"""Test finding class inheritance."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("python")
if not binding.is_available():
pytest.skip("ast-grep not available")
source = """
class MyClass(BaseClass):
pass
"""
binding.parse(source)
results = binding.find_inheritance()
assert len(results) >= 0 # May or may not find depending on pattern match
def test_find_calls(self):
"""Test finding function calls."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("python")
if not binding.is_available():
pytest.skip("ast-grep not available")
source = """
def foo():
bar()
baz.qux()
"""
binding.parse(source)
results = binding.find_calls()
assert isinstance(results, list)
def test_find_imports(self):
"""Test finding import statements."""
from codexlens.parsers.astgrep_binding import AstGrepBinding
binding = AstGrepBinding("python")
if not binding.is_available():
pytest.skip("ast-grep not available")
source = """
import os
from typing import List
"""
binding.parse(source)
results = binding.find_imports()
assert isinstance(results, list)
def test_basic_import():
"""Test that the module can be imported."""
try:
from codexlens.parsers.astgrep_binding import (
AstGrepBinding,
is_astgrep_available,
get_supported_languages,
ASTGREP_AVAILABLE,
)
assert True
except ImportError as e:
pytest.fail(f"Failed to import astgrep_binding: {e}")
def test_availability_flag():
"""Test ASTGREP_AVAILABLE flag is defined."""
from codexlens.parsers.astgrep_binding import ASTGREP_AVAILABLE
assert isinstance(ASTGREP_AVAILABLE, bool)
if __name__ == "__main__":
# Run basic verification
print("Testing astgrep_binding module...")
from codexlens.parsers.astgrep_binding import (
AstGrepBinding,
is_astgrep_available,
get_supported_languages,
)
print(f"ast-grep available: {is_astgrep_available()}")
print(f"Supported languages: {get_supported_languages()}")
binding = AstGrepBinding("python")
print(f"Python binding available: {binding.is_available()}")
if binding.is_available():
test_code = """
import os
from typing import List
class MyClass(BaseClass):
def method(self):
self.helper()
external_func()
def helper():
pass
"""
binding.parse(test_code)
print(f"Inheritance found: {binding.find_inheritance()}")
print(f"Calls found: {binding.find_calls()}")
print(f"Imports found: {binding.find_imports()}")
else:
print("Note: ast-grep-py not installed. To install:")
print(" pip install ast-grep-py")
print(" Note: May have compatibility issues with Python 3.13")
print("Basic verification complete!")