Spaces:
Paused
Paused
| import gradio as gr | |
| import PyPDF2 | |
| import requests | |
| import io | |
| import tempfile | |
| import sys | |
| from PIL import Image | |
| def compress_image(image, quality): | |
| img_buffer = io.BytesIO() | |
| image.save(img_buffer, format='JPEG', quality=quality) | |
| img_buffer.seek(0) | |
| return img_buffer | |
| def compress_pdf(input_file, url, strength): | |
| if input_file is None and (url is None or url.strip() == ""): | |
| return None, "Please provide either a file or a URL." | |
| if input_file is not None and url and url.strip() != "": | |
| return None, "Please provide either a file or a URL, not both." | |
| if url and url.strip() != "": | |
| try: | |
| response = requests.get(url) | |
| response.raise_for_status() | |
| pdf_content = io.BytesIO(response.content) | |
| except requests.RequestException as e: | |
| return None, f"Error downloading PDF: {str(e)}" | |
| else: | |
| pdf_content = input_file | |
| try: | |
| reader = PyPDF2.PdfReader(pdf_content) | |
| writer = PyPDF2.PdfWriter() | |
| if strength == "Low": | |
| image_quality = 65 | |
| compression_level = 1 | |
| elif strength == "Medium": | |
| image_quality = 40 | |
| compression_level = 2 | |
| else: # High | |
| image_quality = 20 | |
| compression_level = 3 | |
| for page in reader.pages: | |
| page.compress_content_streams() # Apply content stream compression | |
| # Compress images on the page | |
| for img_index, img in enumerate(page.images): | |
| if img.image is not None: | |
| try: | |
| pil_image = Image.open(io.BytesIO(img.image)) | |
| compressed_image = compress_image(pil_image, image_quality) | |
| page.replace_image(img_index, compressed_image) | |
| except Exception as e: | |
| print(f"Error compressing image: {e}") | |
| writer.add_page(page) | |
| with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file: | |
| writer.write(temp_file) | |
| temp_file_path = temp_file.name | |
| # Apply additional compression using PyPDF2's built-in compression | |
| with open(temp_file_path, 'rb') as file: | |
| reader = PyPDF2.PdfReader(file) | |
| writer = PyPDF2.PdfWriter() | |
| for page in reader.pages: | |
| page.compress_content_streams() | |
| writer.add_page(page) | |
| writer.add_metadata(reader.metadata) | |
| with open(temp_file_path, 'wb') as output_file: | |
| writer.write(output_file) | |
| return temp_file_path, "PDF compressed successfully!" | |
| except Exception as e: | |
| return None, f"Error compressing PDF: {str(e)}" | |
| def process_and_compress(input_file, url, strength): | |
| sys.setrecursionlimit(10000) | |
| output_file, message = compress_pdf(input_file, url, strength) | |
| if output_file: | |
| return output_file, message | |
| else: | |
| return None, message | |
| with gr.Blocks() as demo: | |
| gr.Markdown("# PDF Compressor") | |
| with gr.Row(): | |
| input_file = gr.File(label="Upload PDF") | |
| url_input = gr.Textbox(label="Or enter PDF URL") | |
| strength = gr.Radio(["Low", "Medium", "High"], label="Compression Strength", value="Medium", info="Low: ~25% compression, Medium: ~50% compression, High: ~75% compression") | |
| compress_btn = gr.Button("Compress") | |
| output_file = gr.File(label="Download Compressed PDF") | |
| message = gr.Textbox(label="Message") | |
| compress_btn.click( | |
| process_and_compress, | |
| inputs=[input_file, url_input, strength], | |
| outputs=[output_file, message] | |
| ) | |
| if __name__ == "__main__": | |
| demo.launch(share=True) |