Spaces:
Sleeping
Sleeping
| import gradio as gr | |
| from pdf2docx import Converter | |
| from docx import Document | |
| from fpdf import FPDF | |
| import os | |
| import tempfile | |
| import pytesseract | |
| from pdf2image import convert_from_path | |
| from reportlab.lib.pagesizes import A4 | |
| from reportlab.pdfgen import canvas | |
| from io import BytesIO | |
| # Ensure Tesseract is installed and its path is set | |
| pytesseract.pytesseract.tesseract_cmd = r'/usr/bin/tesseract' # Update this path to your Tesseract installation | |
| title_and_description = """ | |
| # PDF to Word and Word to PDF Converter | |
| This tool allows you to convert PDF files to Word documents and Word documents to PDF files. | |
| Note: Scanned PDFs (image-based PDFs) are supported using OCR. | |
| """ | |
| def pdf_to_word(pdf_file): | |
| """ | |
| Converts a PDF file to a Word document. | |
| Handles both text-based and image-based (scanned) PDFs. | |
| """ | |
| try: | |
| # Check if Poppler is installed | |
| try: | |
| pages = convert_from_path(pdf_file.name, 500) # Convert PDF to images | |
| except Exception as e: | |
| return f"Error: Unable to process PDF. Is Poppler installed and in PATH? ({e})" | |
| # Create a temporary directory to store intermediate files | |
| with tempfile.TemporaryDirectory() as temp_dir: | |
| docx_filename = os.path.join(temp_dir, os.path.basename(pdf_file.name).replace('.pdf', '.docx')) | |
| # If the PDF is image-based, use OCR to extract text | |
| doc = Document() | |
| for page in pages: | |
| text = pytesseract.image_to_string(page) # Extract text from the image | |
| doc.add_paragraph(text) # Add the extracted text to the Word document | |
| # Save the Word document | |
| doc.save(docx_filename) | |
| # Return the path to the converted file | |
| return docx_filename | |
| except Exception as e: | |
| return f"Error: {e}" | |
| def word_to_pdf(docx_file): | |
| """ | |
| Converts a Word document to a PDF file. | |
| Handles images carefully using reportlab. | |
| """ | |
| try: | |
| # Create a temporary directory to store the output file | |
| with tempfile.TemporaryDirectory() as temp_dir: | |
| pdf_filename = os.path.join(temp_dir, "output.pdf") | |
| # Create a PDF using reportlab | |
| packet = BytesIO() | |
| can = canvas.Canvas(packet, pagesize=A4) | |
| can.setFont("Helvetica", 12) | |
| # Read the Word document | |
| doc = Document(docx_file.name) | |
| y = 800 # Starting y-coordinate for text | |
| for para in doc.paragraphs: | |
| text = para.text.strip() | |
| if not text: | |
| continue | |
| # Add text to the PDF | |
| can.drawString(100, y, text) | |
| y -= 15 # Move down for the next line | |
| # Handle page breaks | |
| if y < 50: | |
| can.showPage() | |
| y = 800 | |
| # Save the PDF | |
| can.save() | |
| packet.seek(0) | |
| with open(pdf_filename, "wb") as f: | |
| f.write(packet.read()) | |
| # Return the path to the converted file | |
| return pdf_filename | |
| except Exception as e: | |
| return f"Error: {e}" | |
| with gr.Blocks() as app: | |
| gr.Markdown(title_and_description) | |
| with gr.Row(): | |
| with gr.Column(): | |
| with gr.Accordion("PDF to Word"): | |
| pdf_input = gr.File(label="Upload PDF", file_types=[".pdf"]) | |
| convert_pdf_to_word = gr.Button("Convert to Word") | |
| word_output = gr.File(label="Download Word file", type="filepath", file_types=[".docx"]) | |
| convert_pdf_to_word.click(pdf_to_word, inputs=[pdf_input], outputs=[word_output]) | |
| with gr.Column(): | |
| with gr.Accordion("Word to PDF"): | |
| word_input = gr.File(label="Upload Word", file_types=[".docx"]) | |
| convert_word_to_pdf = gr.Button("Convert to PDF") | |
| pdf_output = gr.File(label="Download PDF file", type="filepath", file_types=[".pdf"]) | |
| convert_word_to_pdf.click(word_to_pdf, inputs=[word_input], outputs=[pdf_output]) | |
| app.launch() |