changement vers google drive storage pour les contrats.
CI - Security, Lint & Tests / validate (push) Failing after 1m15s

Tentative d'obtenir un token de connection échoue
This commit is contained in:
cedrick2711
2026-08-26 20:52:40 -04:00
parent 38defc53a5
commit 820cccb25e
20 changed files with 476 additions and 2746 deletions
+40 -1
View File
@@ -22,7 +22,9 @@ class GoogleDriveStorageError(RuntimeError):
def _setting(name):
value = os.getenv(name)
if not value:
raise GoogleDriveStorageError(f'{name} must be configured for Google Drive document storage.')
raise GoogleDriveStorageError(
f'{name} must be configured for Google Drive document storage.'
)
return value
@@ -47,6 +49,43 @@ def _drive_service():
return build('drive', 'v3', credentials=credentials, cache_discovery=False)
def list_folder_files():
"""Return (file_id, name) tuples for every file in the configured folder.
Used by the backup script to enumerate Drive-stored contracts. Returns
only files directly in the folder (not nested inside sub-folders).
Returns:
list[tuple[str, str]]: (file_id, name) for each file found.
"""
try:
folder_id = _setting(GOOGLE_DRIVE_FOLDER_ID_ENV)
page_token = None
results = []
while True:
response = (
_drive_service()
.files()
.list(
q=f"'{folder_id}' in parents and trashed=false",
fields='nextPageToken, files(id, name)',
pageToken=page_token,
pageSize=1000,
)
.execute()
)
for f in response.get('files', []):
results.append((f['id'], f['name']))
page_token = response.get('nextPageToken')
if not page_token:
break
return results
except GoogleDriveStorageError:
raise
except Exception as exc:
raise GoogleDriveStorageError('Google Drive could not list folder contents.') from exc
def upload_file(*, stream, filename, mimetype):
"""Upload a PDF to the configured owner folder and return its Drive ID."""
try:
-1312
View File
File diff suppressed because it is too large Load Diff
+20
View File
@@ -39,6 +39,10 @@ def pdf_upload_error(file, allowed_extensions):
could put an arbitrary file on the server under a name the application
later hands back for download (SEC-021).
The check now goes beyond the header: it parses the upload with pypdf
and requires at least one parseable page, so a file with ``%PDF-``
prepended or a renamed non-PDF is rejected.
Args:
file: The uploaded FileStorage, or None.
allowed_extensions: Extensions to accept, lowercase and without dot.
@@ -53,11 +57,27 @@ def pdf_upload_error(file, allowed_extensions):
if not (stem and dot) or extension.lower() not in allowed_extensions:
return _('Only PDF files are allowed for contracts.')
# Fast-fail: the PDF signature must appear at the start of the stream.
head = file.stream.read(len(PDF_SIGNATURE))
file.stream.seek(0)
if head != PDF_SIGNATURE:
return _('That file is not a PDF, whatever its name says.')
# Full-content validation: parse with pypdf and require at least one
# page. A file with a forged header but no real PDF structure (e.g. a
# renamed executable with ``%PDF-`` prepended) will fail here.
try:
from pypdf import PdfReader
from pypdf.errors import PyPdfError
reader = PdfReader(file.stream, strict=False)
if len(reader.pages) == 0:
return _('That file is not a PDF, whatever its name says.')
except (PyPdfError, ValueError, TypeError):
return _('That file is not a PDF, whatever its name says.')
finally:
file.stream.seek(0)
return None
+3 -1
View File
@@ -135,7 +135,9 @@ def upload_contract():
try:
stored_path = store_uploaded_document(file, relative_path)
except GoogleDriveStorageError:
flash(_('Contract storage is temporarily unavailable. Please try again later.'), 'danger')
flash(
_('Contract storage is temporarily unavailable. Please try again later.'), 'danger'
)
return redirect(url_for('users.upload_contract'))
contract = Contract(
+7 -1
View File
@@ -34,8 +34,14 @@ import os
from app.google_drive import (
GoogleDriveStorageError,
)
from app.google_drive import (
delete_file as delete_google_drive_file,
)
from app.google_drive import (
download_file as download_google_drive_file,
)
from app.google_drive import (
upload_file as upload_google_drive_file,
)
@@ -112,7 +118,7 @@ def _rooted(env_name, default_name):
def document_storage_backend():
"""Return the configured backend for new document uploads."""
backend = os.getenv(DOCUMENT_STORAGE_BACKEND_ENV, 'google_drive').strip().lower()
backend = os.getenv(DOCUMENT_STORAGE_BACKEND_ENV, 'local').strip().lower()
if backend not in {'google_drive', 'local'}:
raise ValueError(
f'{DOCUMENT_STORAGE_BACKEND_ENV} must be "google_drive" or "local", not {backend!r}'
+81 -2
View File
@@ -42,7 +42,11 @@ from urllib.parse import unquote, urlparse
# to be — see DOCUMENTS_DIR below.
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))))
from app.storage import backups_root, documents_root # noqa: E402 — needs the path above
from app.storage import ( # noqa: E402 — needs the path above
backups_root,
document_storage_backend,
documents_root,
)
# Configuration
BACKUP_DIR = backups_root()
@@ -292,6 +296,75 @@ def backup_documents():
return zip_path
def backup_drive_documents():
"""Download every file from the configured Google Drive folder and archive it.
Called **in addition to** ``backup_documents()`` when the storage backend is
``google_drive``. When the backend is ``local`` this function silently
skips — local backups already handled by ``backup_documents()``.
Returns:
str | None: Path to the created archive, or None when skipped.
Raises:
BackupError: If Drive is configured but listing or downloading fails.
"""
if document_storage_backend() != 'google_drive':
print('[SKIP] Google Drive is not configured; local documents only.')
return None
try:
from app.google_drive import (
GoogleDriveStorageError,
list_folder_files,
)
from app.google_drive import (
download_file as download_drive_file,
)
except ImportError:
raise BackupError(
'Google Drive dependencies are not installed. Install requirements.txt again.'
) from None
print('[INFO] Listing Google Drive contract folder...')
try:
files = list_folder_files()
except (GoogleDriveStorageError, Exception) as exc: # noqa: BLE001 — boundary
raise BackupError(f'Google Drive listing failed: {exc}') from exc
if not files:
print('[INFO] Google Drive folder is empty; no Drive documents to back up.')
return None
timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
archive_basename = os.path.join(BACKUP_DIR, f'documents_backup_drive_{timestamp}')
download = {}
for file_id, name in files:
print(f' Downloading {name} ({file_id})...')
try:
download[name] = download_drive_file(file_id)
except (GoogleDriveStorageError, Exception) as exc: # noqa: BLE001 — boundary
raise BackupError(f'Google Drive download failed for {name}: {exc}') from exc
try:
import zipfile
zip_path = f'{archive_basename}.zip'
with zipfile.ZipFile(zip_path, 'w', zipfile.ZIP_DEFLATED) as zf:
for name, data in download.items():
zf.writestr(name, data)
except Exception as exc: # noqa: BLE001 — boundary
raise BackupError(f'Drive document archive failed: {exc}') from exc
if not os.path.exists(zip_path) or os.path.getsize(zip_path) == 0:
raise BackupError('Drive document archiver reported success but produced an empty file.')
size_mb = os.path.getsize(zip_path) / (1024 * 1024)
print(f'[OK] Drive documents backed up to: {zip_path} ({size_mb:.1f} MB)')
return zip_path
def cleanup_old_backups():
"""Remove backup files older than BACKUP_RETENTION_DAYS."""
if not os.path.exists(BACKUP_DIR):
@@ -365,10 +438,16 @@ def main(argv=None):
except BackupError as exc:
documents_ok = False
print(f'[ERROR] {exc}')
drive_ok = True
try:
backup_drive_documents()
except BackupError as exc:
drive_ok = False
print(f'[ERROR] {exc}')
cleanup_old_backups()
print()
if verified and documents_ok:
if verified and documents_ok and drive_ok:
print('=== Backup completed successfully ===')
return 0
+121
View File
@@ -0,0 +1,121 @@
"""Obtain a Google Drive refresh token for the application.
Run this once from a development machine that has a browser. It opens a local
OAuth redirect server, asks you to log in, and prints the refresh token.
Usage:
python app/supporting_scripts/get_refresh_token.py
Requires:
pip install google-auth-oauthlib
Before running:
1. Create a Google Cloud project and enable the Drive API.
2. Create a **Desktop** OAuth 2.0 client ID.
3. Add your email as a test user in the OAuth consent screen.
4. Run this script and paste the values when prompted.
"""
import os
import sys
def _prompt(label, env_name, default=""):
"""Ask the user for a value, accepting an env-var fallback or a default."""
candidate = os.getenv(env_name) or default
hint = f" [{candidate}]" if candidate else ""
raw = input(f"{label}{hint}: ").strip()
if raw:
return raw
if candidate:
return candidate
return default
def _validate_client_id(client_id):
"""Warn if the client ID doesn't look like a real OAuth client ID."""
if client_id == "YOUR_CLIENT_ID.apps.googleusercontent.com":
print("[ERROR] You must provide a real client ID from the Google Cloud console.")
return False
if not client_id.endswith(".apps.googleusercontent.com"):
print(
f"[WARNING] Client ID should end with '.apps.googleusercontent.com'.\n"
f" Got: {client_id!r}\n"
f" If this is a placeholder, the OAuth request will fail.\n"
)
return True
def _validate_client_secret(secret):
if secret == "YOUR_CLIENT_SECRET":
print("[ERROR] You must provide a real client secret from the Google Cloud console.")
return False
# The GOCSPX- prefix is now used by every client type, so it no longer
# tells us anything about whether the client was created as a Desktop app.
return True
def main():
print("=" * 60)
print("Google Drive Refresh Token — Setup")
print("=" * 60)
print()
# When both env vars are set, skip prompts entirely for headless runs.
env_id = os.getenv("GOOGLE_DRIVE_CLIENT_ID")
env_secret = os.getenv("GOOGLE_DRIVE_CLIENT_SECRET")
if env_id and env_secret:
client_id = env_id
client_secret = env_secret
print(f"Using Client ID from environment: {client_id[:30]}...")
else:
print("Enter your Google Cloud OAuth credentials.")
print("You can also set GOOGLE_DRIVE_CLIENT_ID and GOOGLE_DRIVE_CLIENT_SECRET")
print("environment variables and re-run to skip the prompts.")
print()
client_id = _prompt("Client ID", "GOOGLE_DRIVE_CLIENT_ID")
client_secret = _prompt("Client Secret", "GOOGLE_DRIVE_CLIENT_SECRET")
if not _validate_client_id(client_id) or not _validate_client_secret(client_secret):
sys.exit(1)
print()
print("Opening browser to authenticate with Google...")
print("(If it doesn't open, copy the URL below into your browser.)")
print()
from google_auth_oauthlib.flow import InstalledAppFlow
SCOPE = "https://www.googleapis.com/auth/drive.file"
client_config = {
"installed": {
"client_id": client_id,
"client_secret": client_secret,
"redirect_uris": ["http://localhost"],
"auth_uri": "https://accounts.google.com/o/oauth2/auth",
"token_uri": "https://oauth2.googleapis.com/token",
}
}
flow = InstalledAppFlow.from_client_config(client_config, scopes=[SCOPE])
credentials = flow.run_local_server(host="localhost", port=8080, open_browser=True)
print()
print("=" * 60)
print("SUCCESS — copy these values into your .env file:")
print("=" * 60)
print()
print("DOCUMENT_STORAGE_BACKEND=google_drive")
print(f"GOOGLE_DRIVE_CLIENT_ID={credentials.client_id}")
print(f"GOOGLE_DRIVE_CLIENT_SECRET={client_secret}")
print(f"GOOGLE_DRIVE_REFRESH_TOKEN={credentials.refresh_token}")
print("GOOGLE_DRIVE_FOLDER_ID=<get this from the Drive folder URL>")
print()
return 0
if __name__ == "__main__":
sys.exit(main())
+2 -2
View File
@@ -94,8 +94,8 @@
<input type="hidden" name="csrf_token" value="{{ csrf_token() }}"/>
<div class="form-group">
<label for="signed_file">{{ _('Select Signed Contract File') }}</label>
<input type="file" id="signed_file" name="signed_file" accept=".pdf,.doc,.docx,.jpg,.jpeg,.png" required>
<small class="form-text text-muted">{{ _('Accepted formats: PDF, DOC, DOCX, JPG, PNG') }}</small>
<input type="file" id="signed_file" name="signed_file" accept=".pdf" required>
<small class="form-text text-muted">{{ _('Accepted format: PDF') }}</small>
</div>
<div class="form-actions">
<button type="button" class="btn btn-secondary" data-action="hide-upload-signed">{{ _('Cancel') }}</button>
+2 -2
View File
@@ -24,8 +24,8 @@
<div class="form-group">
<label for="contract_file">{{ _('Contract File') }}</label>
<input type="file" id="contract_file" name="contract_file" accept=".pdf,.doc,.docx,.jpg,.jpeg,.png" required>
<small class="form-text text-muted">{{ _('Accepted formats: PDF, DOC, DOCX, JPG, PNG') }}</small>
<input type="file" id="contract_file" name="contract_file" accept=".pdf" required>
<small class="form-text text-muted">{{ _('Accepted format: PDF') }}</small>
</div>
<div class="form-group">
Binary file not shown.
File diff suppressed because it is too large Load Diff
Binary file not shown.
File diff suppressed because it is too large Load Diff