gmakstutis/cognee
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix: add allowed extensions

Browse source
This commit is contained in:
lxobr 2025-01-06 11:22:45 +01:00
parent 262deee26e
commit 4cee9a16ce
1 changed files with 20 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

21
cognee/tasks/repo_processor/get_non_code_files.py
View file

@ -25,8 +25,27 @@ async def get_non_py_files(repo_path):
'node_modules', '*.egg-info'
}
ALLOWED_EXTENSIONS = {
'.txt', '.md', '.csv', '.json', '.xml', '.yaml', '.yml', '.html',
'.css', '.js', '.ts', '.jsx', '.tsx', '.sql', '.log', '.ini',
'.toml', '.properties', '.sh', '.bash', '.dockerfile', '.gitignore',
'.gitattributes', '.makefile', '.pyproject', '.requirements',
'.env', '.pdf', '.doc', '.docx', '.dot', '.dotx', '.rtf',
'.wps', '.wpd', '.odt', '.ott', '.ottx', '.txt', '.wp',
'.sdw', '.sdx', '.docm', '.dotm',
# Additional extensions for other programming languages
'.java', '.c', '.cpp', '.h', '.cs', '.go', '.php', '.rb',
'.swift', '.pl', '.lua', '.rs', '.scala', '.kt', '.sh',
'.sql', '.v', '.asm', '.pas', '.d', '.ml', '.clj', '.cljs',
'.erl', '.ex', '.exs', '.f', '.fs', '.r', '.pyi',
'.pdb', '.ipynb', '.rmd', '.cabal', '.hs', '.nim',
'.vhdl', '.verilog', '.svelte', '.html', '.css', '.scss',
'.less', '.json5', '.yaml', '.yml'
}
def should_process(path):
return not any(pattern in path for pattern in IGNORED_PATTERNS)
_, ext = os.path.splitext(path)
return ext in ALLOWED_EXTENSIONS and not any(pattern in path for pattern in IGNORED_PATTERNS)
non_py_files_paths = [
os.path.join(root, file)