Better Git Diff A Python library for extracting function-level changes from Python files in git repositories. This tool enhances standard git diffs by identifying which specific Python functions (methods/functions) were added, modified, or deleted.

Features Parse git diffs and map changes to Python functions Extract precise line numbers for affected functions Support for functions, methods, async functions, and decorator detection Built on Python's native AST parsing for robust Python code analysis Installation bash uv add GitPython Quick Start python from git_operations import get_commit_diff from function_aware_diff import parse_git_diff_with_functions

Get diff for a specific commit

repo_path = "/path/to/your/python/repo" commit_sha = "your_commit_sha_here"

diff_text = get_commit_diff(repo_path, commit_sha) enhanced_changes = parse_git_diff_with_functions(diff_text, repo_path)

Analyze function-level changes

for change in enhanced_changes: if change.is_python_file: print(f"\nPython file: {change.file_path}") for func_change in change.function_changes: func = func_change.function print(f" {func_change.change_type.upper()}: {func.name}") print(f" Lines: {func.start_line}-{func.end_line}") if func.class_name: print(f" Class: {func.class_name}") if func.is_async: print(f" (Async function)") Data Structures PythonFunction Represents a detected Python method or function:

python @dataclass class PythonFunction: name: str # Function name (e.g., "calculate_total") start_line: int # Starting line number in file
end_line: int # Ending line number in file start_byte: int # Starting byte offset (0 for AST-based) end_byte: int # Ending byte offset (0 for AST-based) class_name: str # Containing class name (e.g., "Calculator") is_method: bool # True if this is a class method is_async: bool # True if this is an async function decorator_names: List[str] # List of decorator names (e.g., ["property", "staticmethod"]) FunctionChange Links a function to the type of change that occurred:

python @dataclass
class FunctionChange: function: PythonFunction # The affected function change_type: str # "added", "modified", or "deleted" affected_lines: List[int] # Specific line numbers that changed EnhancedFileChange Enhanced version of git diff data with function information:

python @dataclass class EnhancedFileChange: original_change: FileChange # Original git diff data detected_functions: List[PythonFunction] # All functions found in file function_changes: List[FunctionChange] # Functions that were changed

# Properties:
file_path: str             # Path to the changed file
is_python_file: bool       # True if file ends with .py
hunks: List[DiffHunk]      # Git diff hunks from original change

Advanced Usage Direct Function Detection python from python_function_detector import PythonFunctionDetector

detector = PythonFunctionDetector()

Detect all functions in a Python file

with open("my_module.py", "r") as f: content = f.read()

functions = detector.detect_functions(content)

for func in functions: print(f"{func.name} ({func.start_line}:{func.end_line})") if func.is_method: print(f" Method in class: {func.class_name}") if func.is_async: print(" (Async function)") if func.decorator_names: print(f" Decorators: {', '.join(func.decorator_names)}") Finding Functions by Line Numbers python

Find which functions contain specific line numbers

changed_lines = [15, 23, 45] affected_functions = detector.find_functions_at_lines(python_content, changed_lines)

for func in affected_functions: class_info = f"{func.class_name}." if func.class_name else "" print(f"Line changes affect: {class_info}{func.name}") Example Output Python file: src/calculator.py MODIFIED: add (lines 5-9) Class: Calculator ADDED: subtract (lines 11-15)
Class: Calculator MODIFIED: init (lines 3-4) Class: Calculator

Python file: src/async_service.py ADDED: fetch_data (lines 8-12) (Async function) Decorators: retry Running Tests bash pytest tests/test_python_function_detection.py pytest tests/test_git_operations.py Supported Python Features Regular functions and methods Async functions (async def) Class methods, static methods, and properties Nested functions Functions with decorators Class constructors (init) All Python visibility patterns (private, protected, public)