app.py

from pydrive2.auth import GoogleAuth
from pydrive2.drive import GoogleDrive
import os
import gradio as gr
from datasets import load_dataset, Dataset, concatenate_datasets
import pandas as pd
from PIL import Image
from tqdm import tqdm
import logging
import yaml

# Set up logging
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(levelname)s - %(message)s'
)
logger = logging.getLogger(__name__)

# Load settings
if not os.path.exists("settings.yaml"):
    raise FileNotFoundError("settings.yaml file is missing. Please add it with 'client_secrets_file'.")

with open('settings.yaml', 'r') as file:
    settings = yaml.safe_load(file)

[... keep all the utility functions and DatasetManager class the same ...]

def process_pipeline(folder_id, naming_convention):
    """Main pipeline for processing images and updating dataset."""
    # Validate input
    if not folder_id or not naming_convention:
        return "Please provide both folder ID and naming convention", []

    manager = DatasetManager()

    # Step 1: Authenticate Google Drive
    auth_success, auth_message = manager.authenticate_drive()
    if not auth_success:
        return auth_message, []

    # Step 2: Download and rename files
    success, message, renamed_files = manager.download_and_rename_files(folder_id, naming_convention)
    if not success:
        return message, []

    # Step 3: Update Hugging Face dataset
    success, hf_message = manager.update_huggingface_dataset(renamed_files)
    return f"{message}\n{hf_message}", renamed_files

def process_ui(folder_id, naming_convention):
    """UI handler for the process pipeline"""
    status, renamed_files = process_pipeline(folder_id, naming_convention)
    table_data = [[file['original_name'], file['new_name'], file['file_path']] 
                 for file in renamed_files] if renamed_files else []
    return status, table_data

# Simplified Gradio interface
demo = gr.Interface(
    fn=process_ui,
    inputs=[
        gr.Textbox(
            label="Google Drive Folder ID",
            placeholder="Enter the folder ID from the URL"
        ),
        gr.Textbox(
            label="Naming Convention",
            placeholder="e.g., sports_card",
            value="sports_card"
        )
    ],
    outputs=[
        gr.Textbox(label="Status"),
        gr.Dataframe(
            headers=["Original Name", "New Name", "File Path"]
        )
    ],
    title="Sports Cards Dataset Processor",
    description="""
    Instructions:
    1. Enter the Google Drive folder ID (found in the folder's URL)
    2. Specify a naming convention for the files (e.g., 'sports_card')
    3. Click submit to start processing
    
    Note: Only image files will be processed. Invalid images will be skipped.
    """
)

if __name__ == "__main__":
    demo.launch()