transcriptor-gdrive-sync/gdrive_id_loader.py

#!/usr/bin/env python3
import os
import sys
import mariadb
from google.oauth2 import service_account
from googleapiclient.discovery import build

# ---------------- Configuration ----------------
# Local directory where clips are stored.
LOCAL_DIR = "/shared/transcriptor/clips"

# Google Drive root folder ID where rclone is syncing your files.
DRIVE_ROOT_FOLDER_ID = "1qjq9XEC19g6LGw6fwcZXSQYgOO2YuAOA"

# Path to your service account credentials JSON file.
SERVICE_ACCOUNT_FILE = "service_account.json"


# ---------------- Google Drive API Setup ----------------
SCOPES = ['https://www.googleapis.com/auth/drive.readonly']
try:
    credentials = service_account.Credentials.from_service_account_file(
        SERVICE_ACCOUNT_FILE, scopes=SCOPES
    )
    drive_service = build('drive', 'v3', credentials=credentials)
except Exception as e:
    print(f"Error setting up Google Drive API: {e}")
    sys.exit(1)

# ---------------- Database Connection ----------------
try:
    conn = mariadb.connect(
        host=os.environ.get("DB_HOST", "192.168.0.187"),
        user=os.environ.get("DB_USER", "t0is"),
        password=os.environ.get("DB_PASS", "Silenceisgolden555"),
        database=os.environ.get("DB_NAME", "transcriptor"),
        port=int(os.environ.get("DB_PORT", 3306))
    )
    cursor = conn.cursor()
except mariadb.Error as e:
    print(f"Error connecting to MariaDB: {e}")
    sys.exit(1)

# ---------------- Helper Functions ----------------
def search_folder(folder_name, parent_id):
    """
    Searches for a folder with the given name under the specified parent_id.
    Returns the folder ID if found, otherwise None.
    """
    query = (
        f"'{parent_id}' in parents and "
        f"name='{folder_name}' and "
        f"mimeType='application/vnd.google-apps.folder' and "
        f"trashed=false"
    )
    try:
        results = drive_service.files().list(
            q=query,
            fields="nextPageToken, files(id, name)",
            includeItemsFromAllDrives=True,
            supportsAllDrives=True
        ).execute()
        files = results.get('files', [])
        if files:
            return files[0]['id']
    except Exception as e:
        print(f"Error searching for folder '{folder_name}' under parent '{parent_id}': {e}")
    return None

def search_file(file_name, parent_id):
    """
    Searches for a file (non-folder) with the given name under the specified parent_id.
    Returns the file ID if found, otherwise None.
    """
    query = (
        f"'{parent_id}' in parents and "
        f"name='{file_name}' and "
        f"mimeType!='application/vnd.google-apps.folder' and "
        f"trashed=false"
    )
    try:
        results = drive_service.files().list(
            q=query,
            fields="nextPageToken, files(id, name)",
            includeItemsFromAllDrives=True,
            supportsAllDrives=True
        ).execute()
        files = results.get('files', [])
        if files:
            return files[0]['id']
    except Exception as e:
        print(f"Error searching for file '{file_name}' under parent '{parent_id}': {e}")
    return None

def get_drive_file_id(relative_path):
    """
    Given a relative file path (e.g., "agraelus/from_vod/07-03-25/clip_2399595117_3500.mp4"),
    traverse the folder structure on Google Drive starting at DRIVE_ROOT_FOLDER_ID and
    return the file's ID if found.
    """
    # Split the path into components. (If rclone always uses forward slashes, you may also split on "/".)
    parts = relative_path.split(os.sep)
    if not parts:
        return None

    # All parts except the last are folder names.
    *folders, file_name = parts
    parent_id = DRIVE_ROOT_FOLDER_ID

    # Traverse the folder structure.
    for folder in folders:
        folder_id = search_folder(folder, parent_id)
        if not folder_id:
            print(f"Folder '{folder}' not found under parent '{parent_id}'.")
            return None
        parent_id = folder_id

    # Now search for the file within the final folder.
    return search_file(file_name, parent_id)

def update_database(clip_id, file_id):
    """
    Updates the clip record in the database with the Google Drive file ID.
    """
    query = "UPDATE clips SET gdrive_file_id = ? WHERE id = ?"
    try:
        cursor.execute(query, (file_id, clip_id))
        conn.commit()
        print(f"Updated clip id {clip_id} with file_id: {file_id}")
    except mariadb.Error as e:
        print(f"Database update failed for clip id {clip_id}: {e}")

# ---------------- Main Process ----------------
def main():
    try:
        cursor.execute("SELECT id, filename FROM clips where gdrive_file_id is null")
        clips = cursor.fetchall()
    except mariadb.Error as e:
        print(f"Database query failed: {e}")
        return

    for clip in clips:
        clip_id = clip[0]
        filename = clip[1]

        print(f"Processing clip id {clip_id}: {filename}")
        drive_file_id = get_drive_file_id(filename.replace('clips/', ''))
        if drive_file_id:
            update_database(clip_id, drive_file_id)
        else:
            print(f"Google Drive file ID not found for clip id {clip_id}.")

if __name__ == "__main__":
    main()
    cursor.close()
    conn.close()