python-pdf2txt/app.py

from flask import Flask, request, send_from_directory, jsonify, abort
import os
import logging
import requests
from pdf_to_word import pdf_to_word_txt, pdf_to_word_docx
from datetime import datetime
from threading import Thread

# Set up logging
logging.basicConfig(level=logging.INFO, filename='/app/logs/app.log', filemode='w',
                    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')

# Configure specific logger for potentially noisy libraries or modules
watchdog_logger = logging.getLogger('watchdog')
watchdog_logger.setLevel(logging.ERROR)  # Show only errors from watchdog

app = Flask(__name__)
UPLOAD_FOLDER = '/app/uploads'
OUTPUT_FOLDER = '/app/outputs'
CALLBACK_FOLDER = '/app/callbacks'

if not os.path.exists(UPLOAD_FOLDER):
    os.makedirs(UPLOAD_FOLDER)
if not os.path.exists(OUTPUT_FOLDER):
    os.makedirs(OUTPUT_FOLDER)
if not os.path.exists(CALLBACK_FOLDER):
    os.makedirs(CALLBACK_FOLDER)





def process_pdf_in_background(pdf_path, output_folder, callback_url=None):
    try:
        # Convert PDF to text (assuming you have a function that does this)
        output_path = pdf_to_word_docx(pdf_path, output_folder)

        if output_path and callback_url:
            with open(output_path, 'rb') as f:
                files = {'file': (os.path.basename(output_path), f)}
                response = requests.post(callback_url, files=files)
            response.raise_for_status()

    except Exception as e:
        logging.error(f'Background processing error: {e}', exc_info=True)

def start_background_task(target, *args):
    thread = Thread(target=target, args=args)
    thread.start()

@app.route('/upload-pdf', methods=['POST'])
def upload_pdf():
    file = request.files['file']
    callback_url = request.form.get('callback_url')
    if file and file.filename.endswith('.pdf'):

        pdf_filename = os.path.splitext(file.filename)[0] + '_' + datetime.now().strftime("%Y%m%d_%H%M%S") + '.pdf'
        pdf_path = os.path.join(UPLOAD_FOLDER, pdf_filename)
        file.save(pdf_path)
        logging.info(f'PDF uploaded and saved to {pdf_path}')
        
        # Start the PDF processing in a background thread
        start_background_task(process_pdf_in_background, pdf_path, OUTPUT_FOLDER, callback_url)
        
        # Respond immediately
        return jsonify({'message': 'File upload accepted, processing started'}), 202

    else:
        logging.warning('Invalid file upload attempt')
        abort(400, 'Invalid file type or no file uploaded')


    
    
    
# @app.route('/upload-pdf', methods=['POST'])
# def upload_pdf():
#     file = request.files['file']
#     callback_url = request.form.get('callback_url')
#     if file and file.filename.endswith('.pdf'):

#         pdf_filename = os.path.splitext(file.filename)[0] + '_' + datetime.now().strftime("%Y%m%d_%H%M%S") + '.pdf'
#         pdf_path = os.path.join(UPLOAD_FOLDER, pdf_filename)
#         file.save(pdf_path)

#         logging.info(f'PDF uploaded and saved to {pdf_path}')
        
#         # Convert PDF to text
#         #output_path = pdf_to_word_txt(pdf_path, OUTPUT_FOLDER)
#         output_path = pdf_to_word_docx(pdf_path, OUTPUT_FOLDER)
#         if output_path:
#             logging.info(f'Text file created at {output_path}')
#             if callback_url:
#                 try:
#                     with open(output_path, 'rb') as f:
#                         files = {'file': (os.path.basename(output_path), f)}
#                         response = requests.post(callback_url, files=files)
#                     response.raise_for_status()  # Will raise an HTTPError for bad requests
#                     return jsonify({'message': 'File processed and sent successfully'}), 200
#                 except requests.exceptions.RequestException as e:
#                     logging.error(f'Failed to send file to callback URL: {e}')
#                     abort(500, 'Failed to send file to callback URL')
#             else:
#                 return send_from_directory(OUTPUT_FOLDER, os.path.basename(output_path), as_attachment=True)
#         else:
#             logging.error('Failed to convert PDF')
#             abort(500, 'Conversion failed')
#     else:
#         logging.warning('Invalid file upload attempt')
#         abort(400, 'Invalid file type or no file uploaded')

@app.route('/callback', methods=['POST'])
def callback():
    file = request.files['file']
    if file:
        filepath = os.path.join(CALLBACK_FOLDER, file.filename)
        file.save(filepath)
        logging.info(f'File received and saved at {filepath}')
        return jsonify({'message': 'File received and saved successfully'}), 200
    else:
        logging.error('No file received at callback')
        abort(400, 'No file received')

if __name__ == '__main__':
    app.run(debug=True, host='0.0.0.0', port=5000)