import gradio as gr import tempfile import os import subprocess import uuid def process_files(pdf_file, word_file): # Each upload returns a path (str) with type="filepath" # Create a unique temp directory for each run (prevents parallel collision) temp_dir = tempfile.mkdtemp(prefix="hf_redtext_") # Copy user-uploaded files into temp directory with standard names pdf_path = os.path.join(temp_dir, "input.pdf") word_path = os.path.join(temp_dir, "input.docx") os.rename(pdf_file, pdf_path) os.rename(word_file, word_path) # Step 1: Extract PDF data to txt pdf_txt_path = os.path.join(temp_dir, "pdf_data.txt") subprocess.run( ["python", "extract_pdf_data.py", pdf_path, pdf_txt_path], check=True ) # Step 2: Extract red text from Word to JSON word_json_path = os.path.join(temp_dir, "word_data.json") subprocess.run( ["python", "extract_red_text.py", word_path, word_json_path], check=True ) # Step 3: Update docx JSON with PDF txt, output updated JSON updated_json_path = os.path.join(temp_dir, "updated_word_data.json") subprocess.run( ["python", "update_docx_with_pdf.py", word_json_path, pdf_txt_path, updated_json_path], check=True ) # Step 4: Compare word file with updated JSON and update docx final_docx_path = os.path.join(temp_dir, "updated.docx") subprocess.run( ["python", "updated_word.py", word_path, updated_json_path, final_docx_path], check=True ) # Return final updated 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()