Update app.py

app.py

@@ -3,27 +3,37 @@ import torch
 from transformers import AutoModelForObjectDetection, TableTransformerForObjectDetection
 from PIL import Image
 from pdf_processing import process_pdf
+from io_utils import save_remaining_text_to_txt, save_to_csv
 
-# Load model and process the PDF
+# Process PDF file and save text or data to file
 def process_pdf_file(pdf_file):
     pdf_path = pdf_file.name  # Get the path to the uploaded PDF
     output_folder = "./output"  # Define the output folder
     
-    # Call the existing function to process the PDF
-    process_pdf(pdf_path, output_folder)
+    # Process the PDF file (assuming the function does the necessary work)
+    processed_text = process_pdf(pdf_path, output_folder)  # Modify this based on your function's output
     
-    return f"Processing completed. Output saved to {output_folder}"
+    # Save processed text to a text file
+    txt_file_path = save_remaining_text_to_txt(processed_text, output_folder, 1)  # Assuming page_num = 1 for now
+    
+    # Alternatively, you could generate CSV or any other output
+    data = []  # Replace with your data
+    csv_file_path = save_to_csv(data, output_folder, 'processed_data.csv')
+    
+    # Return file paths as outputs for Gradio
+    return txt_file_path, csv_file_path
 
 # Create Gradio interface
 def create_gradio_interface():
     input_pdf = gr.File(label="Upload PDF", type="filepath")  # Fixed argument here
-    output_text = gr.Textbox(label="Processing Status", interactive=False)
+    output_txt_file = gr.File(label="Download Processed Text", interactive=False)
+    output_csv_file = gr.File(label="Download Processed CSV", interactive=False)
     
     # Define the interface
     interface = gr.Interface(
         fn=process_pdf_file,
         inputs=input_pdf,
-        outputs=output_text,
+        outputs=[output_txt_file, output_csv_file],
         live=True
     )
     return interface