get_drive.py

VISTA_FOLDER_ID = "10xTrOHOFku5C-MnatdiYuqzBKJXzt4kJ"
GEM_FOLDER_ID = "1-rmwZlTklf5w8oNr75Xeb6M1L7RvlS6C"

from pydrive.auth import GoogleAuth
import pandas as pd
from pydrive.drive import GoogleDrive
# Rename the downloaded JSON file to client_secrets.json
# The client_secrets.json file needs to be in the same directory as the script.
gauth = GoogleAuth()
drive = GoogleDrive(gauth)
# List files in Google Drive
fileList = drive.ListFile({'q': f"'{VISTA_FOLDER_ID}' in parents and trashed=false"}).GetList()
vista_files = [(file1['title'], file1['id']) for file1 in fileList]

fileList = drive.ListFile({'q': f"'{GEM_FOLDER_ID}' in parents and trashed=false"}).GetList()
gem_files = [(file1['title'], file1['id']) for file1 in fileList]

# Match files by title and create a csv
data = []

for vista_file in vista_files:
    for gem_file in gem_files:
        if vista_file[0] == gem_file[0]:
            data.append((vista_file[0], vista_file[1], gem_file[1]))

# sort by file_name
data.sort(key=lambda x: x[0])
df = pd.DataFrame(data, columns=['file_name', 'vista_id', 'gem_id'])
df.to_csv('file_pairs.csv', index=False)