Script to Rename TubeArchivist Downloads Using the TubeArchivist API

archivist_renamer.py

import os
import requests
import sys
from datetime import datetime

# Configurable variables
archivist_api_token = os.getenv('ARCHIVIST_API_TOKEN')
archivist_host = os.getenv('ARCHIVIST_HOST')

# Function to sanitize file name
def sanitize_filename(filename):
    return filename.replace(':', '').replace('?', '').replace('\'', '').replace('/', '')

# Function to reformat date in title
def reformat_date_in_title(title):
    try:
        parsed_date = datetime.strptime(title, "%B %d %Y")
        return parsed_date.strftime("%Y-%m-%d")
    except ValueError:
        return title

# Function to get video title and last refresh date from TubeArchivist
def get_video_info_from_tubearchivist(video_id):
    url = f"{archivist_host}/api/video/{video_id}"
    headers = {'Authorization': f'Token {archivist_api_token}'}
    
    try:
        response = requests.get(url, headers=headers)
        response.raise_for_status()
        data = response.json()
        title = data.get('data', {}).get('title')
        published = data.get('data', {}).get('published')

        if title and published:
            formatted_title = reformat_date_in_title(title)
            return formatted_title, published
        else:
            return None, None
    except Exception as e:
        print(f"Error fetching data for video ID {video_id}: {e}")
        return None, None

# Main function
def main():
    if len(sys.argv) != 2:
        print("Usage: python script.py <directory>")
        sys.exit(1)

    directory = sys.argv[1]

    for filename in os.listdir(directory):
        if filename.endswith('.mp4') and len(filename.split('.')[0]) == 11:
            video_id = filename.split('.')[0]
            video_title, published_date = get_video_info_from_tubearchivist(video_id)

            if video_title and published_date:
                base_new_filename = f"{published_date} - {sanitize_filename(video_title)}"
                new_mp4_filename = f"{base_new_filename}.mp4"
                new_att_filename = f"{base_new_filename}.en.vtt"

                # Rename .mp4 file
                # os.rename(os.path.join(directory, filename), os.path.join(directory, new_mp4_filename))
                print(f'Renamed "{filename}" to "{new_mp4_filename}"')

                # Rename .en.vtt file if exists
                att_filename = filename.replace('.mp4', '.en.vtt')
                if os.path.exists(os.path.join(directory, att_filename)):
                    # os.rename(os.path.join(directory, att_filename), os.path.join(directory, new_att_filename))
                    print(f'Renamed "{att_filename}" to "{new_att_filename}"')
            else:
                print(f'Video title or last refresh date not found for "{filename}"')

if __name__ == "__main__":
    main()