Files
compounding-intelligence/scripts/dependency_bloat_detector.py
Alexander Payne 899c4bee1e
Some checks failed
Test / pytest (pull_request) Failing after 8s
5.6: Add dependency bloat detector
Scans all .py files for import statements, compares against
requirements.txt, and reports unused declared packages.

Acceptance:
- Lists all imports in code
- Compares against declared deps
- Reports: unused deps
- Output: bloat report

Refs #112
2026-04-26 00:50:36 -04:00

113 lines
3.4 KiB
Python
Executable File

#!/usr/bin/env python3
"""
Dependency Bloat Detector — find declared packages never imported
Usage:
python3 scripts/dependency_bloat_detector.py
python3 scripts/dependency_bloat_detector.py --output json
"""
import ast
import json
import re
import sys
from pathlib import Path
from typing import Set, List, Tuple
def extract_imports_from_py_files(repo_path: Path) -> Set[str]:
"""Walk the repo and return the set of top-level imported module names."""
imports = set()
exclude_dirs = {".git", "venv", ".venv", "__pycache__", "node_modules",
"dist", "build", ".tox", "vendor"}
py_files = [
f for f in repo_path.rglob("*.py")
if not any(part in exclude_dirs for part in f.parts)
]
for fpath in py_files:
try:
content = fpath.read_text(errors="ignore")
tree = ast.parse(content)
except Exception:
continue
for node in ast.walk(tree):
if isinstance(node, ast.Import):
for alias in node.names:
top = alias.name.split('.')[0]
imports.add(top)
elif isinstance(node, ast.ImportFrom):
if node.module:
top = node.module.split('.')[0]
imports.add(top)
return imports
def parse_requirements_txt(req_path: Path) -> List[Tuple[str, str]]:
"""
Parse requirements.txt and return list of (package_name, raw_line).
Strips version specifiers and ignores comments.
"""
if not req_path.exists():
return []
declared = []
for line in req_path.read_text().splitlines():
line = line.strip()
if not line or line.startswith('#'):
continue
# Strip inline comments
line = line.split('#')[0].strip()
# Extract package name (before any version specifier)
pkg_match = re.match(r'^([a-zA-Z0-9_-]+)', line)
if pkg_match:
pkg = pkg_match.group(1).strip()
declared.append((pkg, line))
return declared
def main():
repo_path = Path('.').resolve()
req_path = repo_path / 'requirements.txt'
# 1. Scan imports
used = extract_imports_from_py_files(repo_path)
# 2. Parse declared deps
declared = parse_requirements_txt(req_path)
declared_names = [pkg for pkg, _ in declared]
# 3. Compare
unused = [(raw, pkg) for pkg, raw in declared if pkg not in used]
missing_from_req = [imp for imp in used if imp not in declared_names]
# 4. Output
print("=" * 60)
print(" DEPENDENCY BLOAT DETECTOR")
print("=" * 60)
print(f" Repository: {repo_path.name}")
print(f" Requirements: {req_path}")
print(f" Python files: {len(list(repo_path.rglob('*.py')))}")
print()
print(f" Declared packages ({len(declared_names)}): {declared_names}")
print(f" Imported packages ({len(used)}): {sorted(used)}")
print()
if unused:
print(" UNUSED DEPENDENCIES (bloat):")
for raw, pkg in unused:
print(f"{raw}")
else:
print(" No unused dependencies detected.")
print()
if missing_from_req:
print(" UNDECLARED IMPORTS (used but not in requirements.txt):")
for imp in missing_from_req:
print(f" ! {imp}")
print()
print("=" * 60)
# Exit code: 0 if no bloat, 1 if unused deps found
sys.exit(1 if unused else 0)
if __name__ == "__main__":
main()