Last active
October 11, 2025 16:03
-
-
Save xflr6/d106aa5b561fbac4ce1a9969eba728bb to your computer and use it in GitHub Desktop.
Recursively traverse the directory tree of a Google Drive folder as variation of os.walk()
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| """os.walk() variation with Google Drive API v3.""" | |
| from collections.abc import Iterator, Sequence | |
| import os | |
| import pathlib | |
| from typing import TypedDict | |
| # $ pip install google-api-python-client google-auth-oauthlib | |
| from apiclient import discovery | |
| from google.oauth2 import credentials | |
| from google_auth_oauthlib import flow as flow_lib | |
| SCOPES = ['https://www.googleapis.com/auth/drive.metadata.readonly'] | |
| FOLDER = 'application/vnd.google-apps.folder' | |
| def get_credentials(scopes: Sequence[str], *, | |
| secrets: os.PathLike[str] | str = '~/client_secrets.json', | |
| storage: os.PathLike[str] | str | None = '~/authorized_user.json' | |
| ) -> credentials.Credentials: | |
| creds = None | |
| if storage is not None: | |
| storage = pathlib.Path(storage).expanduser() | |
| if storage.exists(): | |
| creds = credentials.Credentials.from_authorized_user_file(storage, scopes=scopes) | |
| if creds is None or creds.token_state.name == 'INVALID': | |
| secrets = pathlib.Path(secrets).expanduser() | |
| flow = flow_lib.InstalledAppFlow.from_client_secrets_file(secrets, scopes=scopes) | |
| flow.run_local_server() | |
| creds = flow.credentials | |
| if storage is not None: | |
| authorized_user_info = creds.to_json() | |
| storage.write_text(authorized_user_info) | |
| return creds | |
| creds = get_credentials(SCOPES) | |
| service = discovery.build('drive', version='v3', credentials=creds) | |
| def iterfiles(name: str | None = None, *, | |
| is_folder: bool | None = None, | |
| parent: str | None = None, | |
| order_by: str = 'folder,name,createdTime') -> Iterator['File']: | |
| q = [] | |
| if name is not None: | |
| q.append("name = '{}'".format(name.replace("'", "\\'"))) | |
| if is_folder is not None: | |
| q.append("mimeType {} '{}'".format('=' if is_folder else '!=', FOLDER)) | |
| if parent is not None: | |
| q.append("'{}' in parents".format(parent.replace("'", "\\'"))) | |
| params = {'pageToken': None, 'orderBy': order_by} | |
| if q: | |
| params['q'] = ' and '.join(q) | |
| while True: | |
| response = service.files().list(**params).execute() | |
| for f in response['files']: | |
| yield f | |
| try: | |
| params['pageToken'] = response['nextPageToken'] | |
| except KeyError: | |
| return | |
| class File(TypedDict): | |
| id: str | |
| kind: str | |
| name: str | |
| mimeType: str | |
| resourceKey: str | |
| def walk(top: str = 'root', *, | |
| by_name: bool = False) -> Iterator[tuple[str, list[File], list[File]]]: | |
| if by_name: | |
| (top,) = iterfiles(name=top, is_folder=True) | |
| else: | |
| top = service.files().get(fileId=top).execute() | |
| if top['mimeType'] != FOLDER: | |
| raise ValueError(f'not a folder: {top!r}') | |
| stack = [((top['name'],), top)] | |
| while stack: | |
| (path, top) = stack.pop() | |
| (dirs, files) = is_file = ([], []) | |
| for f in iterfiles(parent=top['id']): | |
| is_file[f['mimeType'] != FOLDER].append(f) | |
| yield path, top, dirs, files | |
| if dirs: | |
| stack.extend((path + (d['name'],), d) for d in reversed(dirs)) | |
| for kwargs in [{'top': 'spam', 'by_name': True}, {}]: | |
| print('', f'walk(**{kwargs!r})', sep='\n') | |
| for path, root, dirs, files in walk(**kwargs): | |
| print('/'.join(path), f'{len(dirs):d}', f'{len(files):d}', sep='\t') |
Author
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Thanks, I have updated
get_credentials()in the latest version of the gist to usegoogle_auth_oauthlibinstead ofoauth2clientnow using~/authorized_user.jsoninstead of~/storage.jsonas default path forstoragesince the new library uses a slightly different format (notably'token'instead of'access_token'as key for the token in the JSON).