import os import streamlit as st from PyPDF2 import PdfMerger from google.cloud import vision from google.oauth2 import service_account import fitz from PIL import Image from io import BytesIO import tempfile import json from streamlit_sortables import sort_items import subprocess from tqdm import tqdm from datetime import datetime import uuid #os.environ["PATH"] += os.pathsep + "/usr/bin" + os.pathsep + "/usr/local/bin" #Removed redundant path setting print(f"Current PATH: {os.environ['PATH']}") # Check if /usr/bin/pdfinfo exists if os.path.exists("/usr/bin/pdfinfo"): print("pdfinfo exists at /usr/bin/pdfinfo") # Check the file permissions permissions = os.stat("/usr/bin/pdfinfo").st_mode print(f"File permissions: {oct(permissions)}") else: print("pdfinfo does not exist at /usr/bin/pdfinfo") # Load Google Cloud Vision credentials from secret credentials_json = os.getenv("GOOGLE_CREDENTIALS_JSON") if credentials_json: credentials_dict = json.loads(credentials_json) credentials = service_account.Credentials.from_service_account_info(credentials_dict) client = vision.ImageAnnotatorClient(credentials=credentials) else: client = None # Function to extract text using Google Cloud Vision def extract_text_with_google_vision(image_bytes): """Extracts text using Google Cloud Vision.""" image = vision.Image(content=image_bytes) response = client.document_text_detection(image=image) if response.error.message: raise Exception(f"Google Cloud Vision API Error: {response.error.message}") return response.full_text_annotation.text if response.full_text_annotation else "" # Function to process PDF for transcription def process_pdf(file): """Converts PDF pages to images and extracts text.""" text = "" with tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) as temp_pdf: temp_pdf.write(file.read()) temp_pdf_path = temp_pdf.name try: doc = fitz.open(temp_pdf_path) for i in tqdm(range(len(doc)), desc="Processing pages"): page = doc.load_page(i) pix = page.get_pixmap() image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples) image_bytes = BytesIO() image.save(image_bytes, format="PNG") image_bytes.seek(0) try: page_text = extract_text_with_google_vision(image_bytes.getvalue()) text += f"--- Page {i + 1} ---\n{page_text}\n\n" except Exception as e: st.error(f"Error on page {i + 1}: {e}") finally: os.remove(temp_pdf_path) with tempfile.NamedTemporaryFile(suffix=".txt", delete=False) as temp_txt: temp_txt.write(text.encode("utf-8")) temp_txt_path = temp_txt.name return temp_txt_path # Function to generate thumbnail from PDF def get_pdf_thumbnail(uploaded_file, page_num, rotation=0): """Generates a thumbnail image of a PDF page.""" uploaded_file.seek(0) doc = fitz.open(stream=uploaded_file.read(), filetype="pdf") page = doc.load_page(page_num) page.set_rotation(rotation) pix = page.get_pixmap() img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples) img.thumbnail((400, 560)) return img # Function to merge PDFs def merge_pdfs(reordered_pages, uploaded_file): """Merges multiple PDFs into one.""" merger = PdfMerger() uploaded_file.seek(0) doc = fitz.open(stream=uploaded_file.read(), filetype="pdf") temp_files = [] for page_data in reordered_pages: page = doc.load_page(page_data['page_index']) page.set_rotation(page_data['rotation']) temp_pdf_page = tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) page.get_pixmap().save(temp_pdf_page.name) temp_files.append(temp_pdf_page.name) merger.append(temp_pdf_page.name) output_filename = "combined_document.pdf" with open(output_filename, "wb") as output_file: merger.write(output_file) for file in temp_files: os.remove(file) return output_filename def download_file(output_file, file_name, mime_type): with open(output_file, "rb") as f: st.download_button( label="Download File", data=f, file_name=file_name, mime=mime_type, ) def generate_user_filename(filename, file_type): """Generates a filename based on user input and file type.""" return f"{filename}{file_type}" def generate_unique_filename(original_filename, suffix, file_type): """Generates a unique filename based on date, time, and original filename.""" now = datetime.now() timestamp = now.strftime("%Y%m%d_%H%M%S") name, ext = os.path.splitext(original_filename) return f"{name}_{timestamp}{suffix}{file_type}" def rotate_pdf(pdf_file, rotation_angle): """Rotates all pages of a PDF by a specified angle.""" with tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) as temp_pdf: temp_pdf.write(pdf_file.read()) temp_pdf_path = temp_pdf.name doc = fitz.open(temp_pdf_path) for page in doc: page.set_rotation(rotation_angle) os.remove(temp_pdf_path) return doc def display_pdf_preview(doc, page_num=0): """Displays a preview of a PDF page.""" page = doc.load_page(page_num) pix = page.get_pixmap() img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples) st.image(img, use_container_width=True) # Sidebar Navigation st.sidebar.title("Tool Selector") selection = st.sidebar.radio("Choose a tool:", ["PDF Combiner", "PDF Transcriber", "PDF Rotator", "PDF Document Separator"]) # PDF Document Separator Tool if selection == "PDF Document Separator": st.title("PDF Document Separator") st.write("Upload a multi-document PDF and separate the documents out.") uploaded_file = st.file_uploader("Upload a PDF", type="pdf") if uploaded_file: # Initialize session state variables if not already defined if "uploaded_pdf" not in st.session_state: st.session_state.uploaded_pdf = None if "target_pages" not in st.session_state: st.session_state.target_pages = [] # List of dictionaries of "page_index", "rotation", "uuid", "image" if "available_pages" not in st.session_state: st.session_state.available_pages = [] if "pdf_filename" not in st.session_state: st.session_state.pdf_filename = "combined_document" if "transcribe_filename" not in st.session_state: st.session_state.transcribe_filename = "transcribed_text" st.session_state.uploaded_pdf = uploaded_file def reset_state(): """Resets the session state for the PDF Document Separator tool.""" st.session_state.uploaded_pdf = None st.session_state.target_pages = [] st.session_state.available_pages = [] if not st.session_state.available_pages: uploaded_file.seek(0) doc = fitz.open(stream=uploaded_file.read(), filetype="pdf") for page_num in range(len(doc)): image = get_pdf_thumbnail(uploaded_file, page_num) st.session_state.available_pages.append({"page_index":page_num, "rotation":0, "uuid":str(uuid.uuid4()), "image":image}) def move_page_to_target(page_uuid): """Moves a page from available_pages to target_pages.""" page_to_move = next((page for page in st.session_state.available_pages if page["uuid"] == page_uuid), None) if page_to_move: st.session_state.target_pages.append(page_to_move) st.session_state.available_pages = [page for page in st.session_state.available_pages if page["uuid"] != page_uuid] def remove_page_from_available(page_uuid): """Removes a page from available_pages.""" st.session_state.available_pages = [ page for page in st.session_state.available_pages if page["uuid"] != page_uuid ] left_col, right_col = st.columns(2) with left_col: st.header("Document Overview") st.write("Drag pages to the right to start creating the new document.") for page in st.session_state.available_pages: col1, col2 = st.columns([1,2]) with col1: st.image(page["image"], caption=f"Page {page['page_index'] + 1}", use_container_width=True) with col2: if st.button(f"Add Page", key=f"add_{page['uuid']}"): move_page_to_target(page["uuid"]) if st.button("Remove Page", key=f"remove_{page['uuid']}"): remove_page_from_available(page["uuid"]) with right_col: st.header("Target Document Builder") st.write("Reorder the pages, change rotation, and create the final document.") if st.session_state.target_pages: # Use sort_items for the right-hand column to manage reordering target_page_uuids = [page["uuid"] for page in st.session_state.target_pages] reordered_page_uuids = sort_items(target_page_uuids) reordered_target_pages = [next(page for page in st.session_state.target_pages if page['uuid'] == uuid) for uuid in reordered_page_uuids] st.session_state.target_pages = reordered_target_pages for page in st.session_state.target_pages: with st.expander(f"Page {page['page_index'] + 1} Options", expanded=True): col1, col2 = st.columns([1,1]) with col1: st.image(page["image"], use_container_width=True) with col2: rotation = st.selectbox(f"Rotation", [0, 90, 180, 270], key=page["uuid"], index = [0,90,180,270].index(page['rotation'])) if rotation != page['rotation']: page['rotation'] = rotation page['image'] = get_pdf_thumbnail(st.session_state.uploaded_pdf, page['page_index'], rotation) st.text_input("Enter output PDF filename", key='pdf_filename_input', value = st.session_state.pdf_filename) if st.button("Create PDF"): st.session_state.pdf_filename = st.session_state.pdf_filename_input with st.spinner("Creating PDF..."): output_file = merge_pdfs(st.session_state.target_pages, st.session_state.uploaded_pdf) st.success("PDF created successfully!") download_file(output_file, generate_user_filename(st.session_state.pdf_filename, ".pdf"), "application/pdf") os.remove(output_file) st.text_input("Enter output transcription filename", key='transcribe_filename_input', value = st.session_state.transcribe_filename) if st.button("Transcribe PDF"): st.session_state.transcribe_filename = st.session_state.transcribe_filename_input with st.spinner("Processing..."): output_file = process_pdf(st.session_state.uploaded_pdf) if output_file: st.success("Text extraction complete!") st.text_area("Extracted Text", open(output_file, encoding="utf-8").read(), height=400) output_file_name = generate_unique_filename(st.session_state.transcribe_filename, "(T)", ".txt") with open(output_file, "rb") as f: st.download_button("Download Extracted Text", f, file_name=output_file_name, mime="text/plain") os.remove(output_file) else: st.error("Could not process the PDF, please try again") if st.button("Reset State"): reset_state() # The rest of your original app.py code should go here # PDF Combiner with Preview and Reordering elif selection == "PDF Combiner": st.title("PDF Combiner with Preview & Reordering") st.write("Upload individual PDF pages, visualize them, reorder, and merge into a single PDF.") uploaded_files = st.file_uploader("Upload PDF pages", type="pdf", accept_multiple_files=True) if uploaded_files: # Generate thumbnails and filenames for each uploaded PDF thumbnails = [] filenames = [] for file in uploaded_files: thumbnails.append(get_pdf_thumbnail(file, 0)) filenames.append(file.name) # Display thumbnails with filenames for reordering st.write("**Drag and drop to reorder the PDFs:**") reordered_filenames = sort_items(filenames) # Map the filenames back to the corresponding files reordered_files = [uploaded_files[filenames.index(name)] for name in reordered_filenames] # Display the thumbnails in the new order st.write("**Preview of selected order:**") cols = st.columns(len(reordered_files)) for idx, file in enumerate(reordered_files): with cols[idx]: st.image(get_pdf_thumbnail(file, 0), caption=file.name, use_container_width=True) # Merge PDFs in the specified order if st.button("Merge PDFs"): output_file = merge_pdfs(reordered_files) st.success("PDF pages combined successfully!") download_file(output_file, generate_unique_filename("combined_document", "", ".pdf"), "application/pdf") os.remove(output_file) # PDF Transcriber Tool elif selection == "PDF Transcriber": st.title("PDF Transcriber Tool") st.write("Upload a scanned PDF to transcribe the text.") if not client: st.error("Google Cloud credentials are not set. Please configure the secret.") uploaded_file = st.file_uploader("Upload a PDF", type=["pdf"]) if uploaded_file and st.button("Transcribe PDF"): with st.spinner("Processing..."): output_file = process_pdf(uploaded_file) if output_file: st.success("Text extraction complete!") st.text_area("Extracted Text", open(output_file, encoding="utf-8").read(), height=400) output_file_name = generate_unique_filename(uploaded_file.name, "(T)", ".txt") with open(output_file, "rb") as f: st.download_button("Download Extracted Text", f, file_name=output_file_name, mime="text/plain") os.remove(output_file) else: st.error("Could not process the PDF, please try again") # PDF Rotator Tool elif selection == "PDF Rotator": st.title("PDF Rotator") st.write("Upload a PDF and rotate all pages by a specified angle.") pdf_file = st.file_uploader("Upload PDF", type=["pdf"]) rotation_angle = st.selectbox("Select rotation angle:", [90, 180, 270]) if pdf_file: if 'rotated_doc' not in st.session_state: st.session_state.rotated_doc = None if st.button("Rotate PDF"): with st.spinner("Rotating PDF..."): st.session_state.rotated_doc = rotate_pdf(pdf_file, rotation_angle) st.success("PDF rotated successfully!") if st.session_state.rotated_doc: display_pdf_preview(st.session_state.rotated_doc) if st.button("Download Rotated PDF"): with tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) as temp_pdf: st.session_state.rotated_doc.save(temp_pdf.name) download_file(temp_pdf.name, generate_unique_filename(pdf_file.name, "(R)", ".pdf"), "application/pdf") os.remove(temp_pdf.name)