windows - Python flask API to convert DOCX to PDF - Stack Overflow

admin管理员组

文章数量:1125579

I created a Python Flask API to convert DOCX files to PDF. It works perfectly when I start it manually via the command line. However, I’d like to turn it into a Windows service so that it can start automatically and restart if needed.

I tried using NSSM to set it up as a service, and while everything seemed to work, the service couldn't open the DOCX files no matter what I tried.

I also experimented with pywin32, but the requests completely failed once the service was started.

I don’t think this is a permissions issue. I added logging to my code and verified that the DOCX files are received correctly, have the right permissions, and are in the correct format.

Does anyone have an idea of how to resolve this?

Here is my UPDATED code :

import win32com.client
import pythoncom
import inspect, os
from flask import Flask, request, send_file
import tempfile
import logging
import time
import stat
import getpass

# Set up logging
logging.basicConfig(level=logging.DEBUG)
logger = logging.getLogger(__name__)

app = Flask(__name__)

# WordHandler class to handle opening and closing Word application
class WordHandler:
    def __init__(self):
        self.word = None
        logger.debug("WordHandler initialized")

    def __enter__(self):
        logger.debug("Starting Word application...")
        start_time = time.time()
        pythoncom.CoInitialize()
        try:
            self.word = win32com.client.DispatchEx("Word.Application")
            logger.debug(f"Word application started in {time.time() - start_time:.2f} seconds")
        except Exception as ex:
            logger.error("Error starting Word application", exc_info=ex)
            raise
        return self

    def __exit__(self, exc_type, exc_val, exc_tb):
        logger.debug("Closing Word application...")
        try:
            if self.word:
                self.word.Quit()
            pythoncom.CoUninitialize()
            logger.debug("Word application closed")
        except Exception as ex:
            logger.error("Error closing Word application", exc_info=ex)
            raise

    def process_document(self, docx_file):
        """Handles both TOC update and PDF conversion in a single Word session"""
        logger.debug(f"Opening document: {docx_file}")
        start_time = time.time()
        
        if not os.path.exists(docx_file):
            logger.error(f"The file does not exist: {docx_file}")
            raise Exception(f"The file does not exist: {docx_file}")

        try:
            # Log Word application state
            logger.debug(f"Word application object exists: {self.word is not None}")
            
            doc = self.word.Documents.Open(docx_file)
            # Check if document was opened successfully
            if doc is None:
                logger.error("Failed to open document - Word returned None")
                raise Exception("Failed to open document in Word")
                
            logger.debug(f"Document opened in {time.time() - start_time:.2f} seconds")
            logger.debug(f"Document object exists: {doc is not None}")
            
            # Add diagnostics for the document
            try:
                logger.info(f"Document name: {doc.Name}")
                logger.info(f"Document path: {doc.Path}")
                logger.info(f"Document pages: {doc.ComputeStatistics(2)}")  # 2 is for page count
                logger.info(f"Document paragraphs count: {len(doc.Paragraphs)}")
            except Exception as doc_ex:
                logger.error(f"Error accessing document properties: {str(doc_ex)}")
                raise Exception("Failed to access document properties")

            # Check for Table of Contents
            try:
                toc_count = doc.TablesOfContents.Count
                logger.info(f"Number of Tables of Contents found: {toc_count}")
            except Exception as toc_ex:
                logger.error(f"Error checking TOC count: {str(toc_ex)}")
                toc_count = 0

            # Only try to update TOC if it exists
            if toc_count > 0:
                logger.debug("Updating Table of Contents...")
                toc_start = time.time()
                doc.TablesOfContents(1).Update()
                logger.debug(f"TOC updated in {time.time() - toc_start:.2f} seconds")
                doc.Save()
            else:
                logger.warning("No Table of Contents found in document - skipping TOC update")

            # Convert to PDF
            pdf_file = docx_file.replace('.docx', '.pdf')
            logger.debug(f"Converting to PDF: {pdf_file}")
            pdf_start = time.time()
            doc.SaveAs(pdf_file, FileFormat=17)
            logger.debug(f"PDF conversion completed in {time.time() - pdf_start:.2f} seconds")
            doc.Close()
            logger.debug("Document closed")
            return pdf_file

        except Exception as ex:
            logger.error(f"Failed to process document. Exception: {str(ex)}", exc_info=True)
            if 'doc' in locals() and doc:
                try:
                    doc.Close(SaveChanges=False)
                    logger.debug("Document closed after error")
                except:
                    pass
            raise Exception("Failed to process document. Check the file and try again.")

@app.route('/status', methods=['GET'])
def status():
    return {'status': 'API is running', 'version': '1.0'}, 200

@app.route('/convert', methods=['POST'])
def convert():
    logger.debug("Received conversion request")
    if 'file' not in request.files:
        logger.warning("No file provided in request")
        return 'No file provided', 400
        
    file = request.files['file']
    if not file.filename.endswith('.docx'):
        logger.warning(f"Invalid file format: {file.filename}")
        return 'Invalid file format. Please upload a .docx file', 400

    temp_dir = None
    temp_docx = None
    pdf_path = None

    try:
        logger.debug("Creating temporary directory...")
        temp_dir = tempfile.mkdtemp()
        temp_docx = os.path.join(temp_dir, 'input.docx')
        logger.debug(f"Saving uploaded file to {temp_docx}")
        file.save(temp_docx)
        
        # Add file checks and logging
        logger.info(f"File exists: {os.path.exists(temp_docx)}")
        logger.info(f"Absolute file path: {os.path.abspath(temp_docx)}")
        
        # Check file permissions
        file_stat = os.stat(temp_docx)
        permissions = stat.filemode(file_stat.st_mode)
        logger.info(f"File permissions: {permissions}")
        
        # Get current user
        current_user = getpass.getuser()
        logger.info(f"Current user: {current_user}")
        
        logger.debug("Starting document processing...")
        process_start = time.time()
        with WordHandler() as word_handler:
            pdf_path = word_handler.process_document(temp_docx)
        logger.debug(f"Document processing completed in {time.time() - process_start:.2f} seconds")
        
        response = send_file(
            pdf_path,
            mimetype='application/pdf',
            as_attachment=True,
            download_name='converted.pdf'
        )
        
        @response.call_on_close
        def cleanup():
            logger.debug("Starting cleanup...")
            try:
                if temp_docx and os.path.exists(temp_docx):
                    os.unlink(temp_docx)
                if temp_dir and os.path.exists(temp_dir):
                    os.rmdir(temp_dir)
                if pdf_path and os.path.exists(pdf_path):
                    os.unlink(pdf_path)
                logger.debug("Cleanup completed")
            except Exception as e:
                logger.error(f"Error during cleanup", exc_info=e)
                
        return response
        
    except Exception as e:
        logger.error("Error during conversion request", exc_info=e)
        return str(e), 500

if __name__ == "__main__":
    app.run(host='0.0.0.0', port=5001, debug=True)

And here are the logs :

WARNING:werkzeug: * Debugger is active!
INFO:werkzeug: * Debugger PIN: 366-636-785
DEBUG:__main__:Received conversion request
DEBUG:__main__:Creating temporary directory...
DEBUG:__main__:Saving uploaded file to C:\Users\User\AppData\Local\Temp\tmp5dwqrhb1\input.docx
INFO:__main__:File exists: True
INFO:__main__:Absolute file path: C:\Users\User\AppData\Local\Temp\tmp5dwqrhb1\input.docx
INFO:__main__:File permissions: -rw-rw-rw-
INFO:__main__:Current user: User
DEBUG:__main__:Starting document processing...
DEBUG:__main__:WordHandler initialized
DEBUG:__main__:Starting Word application...
DEBUG:__main__:Word application started in 0.35 seconds
DEBUG:__main__:Opening document: C:\Users\User\AppData\Local\Temp\tmp5dwqrhb1\input.docx
DEBUG:__main__:Word application object exists: True
ERROR:__main__:Failed to open document - Word returned None
ERROR:__main__:Failed to process document. Exception: Failed to open document in Word
Traceback (most recent call last):
  File "C:\Users\User\Desktop\docx2pdf\app.py", line 64, in process_document
    raise Exception("Failed to open document in Word")
Exception: Failed to open document in Word
DEBUG:__main__:Closing Word application...
DEBUG:__main__:Word application closed
ERROR:__main__:Error during conversion request
Traceback (most recent call last):
  File "C:\Users\User\Desktop\docx2pdf\app.py", line 64, in process_document
    raise Exception("Failed to open document in Word")
Exception: Failed to open document in Word

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "C:\Users\User\Desktop\docx2pdf\app.py", line 160, in convert
    pdf_path = word_handler.process_document(temp_docx)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\User\Desktop\docx2pdf\app.py", line 115, in process_document
    raise Exception("Failed to process document. Check the file and try again.")
Exception: Failed to process document. Check the file and try again.
INFO:werkzeug:172.16.120.156 - - [09/Jan/2025 12:51:05] "[35m[1mPOST /convert HTTP/1.1[0m" 500 -

Thanks in advance.

本文标签：