app.py

import gradio as gr
import tempfile
import os
import shutil
import subprocess

def process_files(pdf_file, word_file):
    # Create a unique temporary directory for this run
    temp_dir = tempfile.mkdtemp(prefix="hf_redtext_")

    # Define standard filenames for use in the pipeline
    pdf_path = os.path.join(temp_dir, "input.pdf")
    word_path = os.path.join(temp_dir, "input.docx")
    pdf_txt_path = os.path.join(temp_dir, "pdf_data.txt")
    word_json_path = os.path.join(temp_dir, "word_data.json")
    updated_json_path = os.path.join(temp_dir, "updated_word_data.json")
    final_docx_path = os.path.join(temp_dir, "updated.docx")

    # Copy the uploaded files to the temp directory
    shutil.copy(pdf_file, pdf_path)
    shutil.copy(word_file, word_path)

    # Step 1: Extract text from the PDF
    subprocess.run(["python", "extract_pdf_data.py", pdf_path, pdf_txt_path], check=True)

    # Step 2: Extract red text from the Word document
    subprocess.run(["python", "extract_red_text.py", word_path, word_json_path], check=True)

    # Step 3: Update the Word JSON using the PDF text (calls OpenAI)
    subprocess.run(["python", "update_docx_with_pdf.py", word_json_path, pdf_txt_path, updated_json_path], check=True)

    # Step 4: Apply the updated JSON to the Word doc to create the final output
    subprocess.run(["python", "updated_word.py", word_path, updated_json_path, final_docx_path], check=True)

    # Return the final .docx file
    return final_docx_path

iface = gr.Interface(
    fn=process_files,
    inputs=[
        gr.File(label="Upload PDF File", type="filepath"),
        gr.File(label="Upload Word File", type="filepath")
    ],
    outputs=gr.File(label="Download Updated Word File"),
    title="Red Text Replacer",
    description="Upload a PDF and Word document. Red-colored text in the Word doc will be replaced by matching content from the PDF."
)

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