convert_doc_pdf.py

from sys import *
from docx import *
from reportlab.lib.pagesizes import *
from reportlab.pdfgen.canvas import *

import os


def doc_pdf(docpath):
    # Ensure the file has a .docx extension
    if not docpath.lower().endswith(".docx"):
        print("Error: The file must be a .docx file.")
        exit(1)

    # Check if the input file exists
    if not os.path.exists(docpath):
        print(f"Error: The file {docpath} was not found.")
        exit(1)

    # Generate the output PDF path
    pdfpath = docpath.rsplit(".", 1)[0] + ".pdf"

    try:
        # Load the .docx file
        doc = Document(docpath)

        # Create a PDF canvas
        c = Canvas(pdfpath, pagesize=letter)
        width, height = letter

        # Set the starting position for text
        y_position = height - 40

        # Loop through the paragraphs in the document
        for para in doc.paragraphs:
            # If there's enough space on the page, add the paragraph text
            if y_position > 40:
                c.drawString(40, y_position, para.text)
                y_position -= 15  # Move down for the next line
            else:
                # If we run out of space, create a new page
                c.showPage()
                y_position = height - 40
                c.drawString(40, y_position, para.text)
                y_position -= 15

        # Save the PDF
        c.save()

        print(f"PDF saved at -> {pdfpath}")

    except Exception as e:
        print(f"Error occurred during conversion: {e}")
        exit(1)


if __name__ == "__main__":
    if len(argv) != 2:
        print("Usage: python doc_pdf.py <path_to_doc_file>")
        exit(1)

    docpath = argv[1]
    doc_pdf(docpath)