import os os.system('git clone https://github.com/facebookresearch/detectron2.git@v0.6') os.system('pip install -e detectron2') os.system("git clone https://github.com/microsoft/unilm.git") os.system("sed -i 's/from collections import Iterable/from collections.abc import Iterable/' unilm/dit/object_detection/ditod/table_evaluation/data_structure.py") os.system("curl -LJ -o publaynet_dit-b_cascade.pth 'https://layoutlm.blob.core.windows.net/dit/dit-fts/publaynet_dit-b_cascade.pth?sv=2022-11-02&ss=b&srt=o&sp=r&se=2033-06-08T16:48:15Z&st=2023-06-08T08:48:15Z&spr=https&sig=a9VXrihTzbWyVfaIDlIT1Z0FoR1073VB0RLQUMuudD4%3D'") import sys sys.path.append("unilm") sys.path.append("detectron2") import uuid # Import the UUID library import torch import gradio as gr from pdfextract_fun import * from pdfsummary_fun import * from imagesummary_fun import * # Assuming all your defined functions are above and imported correctly into this script @spaces.GPU def process_pdf(pdf_file,state): #base_name = os.path.splitext(os.path.basename(pdf_file.name))[0] unique_id = str(uuid.uuid4()) # Generate a unique identifier output_folder = os.path.join("processed_files", unique_id) # Use UUID for the output folder name, within a parent directory if not os.path.exists(output_folder): os.makedirs(output_folder) # Convert the uploaded PDF file to JPG images convert_pdf_to_jpg(pdf_file.name, output_folder) # Process the images to analyze and extract instances, then rename files sequentially and perform OCR process_jpeg_images(output_folder) #process_jpeg_images(output_folder) rename_files_sequentially(output_folder) ocr_folder(output_folder) image_files = [os.path.join(output_folder, f) for f in os.listdir(output_folder) if f.endswith('.jpg') and ('figure' in f or 'table' in f)] #images = [Image.open(f) for f in image_files] images = [(Image.open(f), os.path.basename(f).split('.')[0]) for f in image_files] # For demonstration, let's just return the path to the output folder # In a real app, you'd want to return images, texts, or links to download the results return images, output_folder def call_pdf_summary(state): ocr_results_folder = os.path.join(state, "ocr_results") summary = pdf_summary(ocr_results_folder) # Assuming pdf_summary accepts an output folder argument return summary def handle_summary_button_click(selected_images): # Check if any image is selected summary = get_image_summary(selected_images) return summary with gr.Blocks(theme=gr.themes.Monochrome()) as app: gr.Markdown("# ChatPaper!") state = gr.State() # Initialize state with gr.Row(): file_input = gr.File(type="filepath", label="Upload a PDF") with gr.Row(): gallery_output = gr.Gallery(label="Extracted Figures and Tables", show_label=True,columns=[3], rows=[1], object_fit="contain", height="auto") with gr.Column(): summary_output = gr.Textbox(label="PDF Summary") summary_button = gr.Button("Generate Summary") with gr.Row(): # Initialize Dropdown without choices; they will be set dynamically image_input = gr.Image(label="Select an Figure or Table for analysis",type='filepath',show_label=True, height="auto") with gr.Column(): image_summary_output = gr.Textbox(label="Figure or Table analysis") image_summary_button = gr.Button("Generate Figure or Table analysis") file_input.change(process_pdf, inputs=[file_input, state], outputs=[gallery_output, state]) summary_button.click(call_pdf_summary, inputs=[state], outputs=[summary_output]) image_summary_button.click(handle_summary_button_click, inputs=image_input, outputs=image_summary_output) # Note: Authentication details removed for security reasons app.launch(share=True) # Launch the app with sharing enabled