app.py

import gradio as gr
import fitz  # PyMuPDF
import os
import re
from datetime import datetime

def pdf_to_kindle_md(pdf_file, image_quality):
    # Create output directory
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    output_dir = f"output_{timestamp}"
    os.makedirs(output_dir, exist_ok=True)
    
    doc = fitz.open(pdf_file.name)
    md_content = []
    image_count = 1
    
    for page_num, page in enumerate(doc):
        # Extract text with formatting
        blocks = page.get_text("dict")["blocks"]
        for b in blocks:
            for l in b.get("lines", []):
                for s in l.get("spans", []):
                    text = s["text"]
                    font_flags = s["flags"]
                    
                    # Detect bold text (using font flags)
                    is_bold = font_flags & 2 ** 4  # 16 is the bold flag
                    
                    # Detect headings (heuristic based on font size)
                    font_size = s["size"]
                    if font_size >= 20:
                        md_content.append(f"\n# {text}\n")
                    elif font_size >= 16:
                        md_content.append(f"\n## {text}\n")
                    elif font_size >= 14:
                        md_content.append(f"\n### {text}\n")
                    else:
                        if is_bold:
                            md_content.append(f"**{text}** ")
                        else:
                            md_content.append(f"{text} ")
                
                md_content.append("\n")
        
        # Extract images
        image_list = page.get_images()
        for img_index, img in enumerate(image_list):
            xref = img[0]
            base_image = doc.extract_image(xref)
            image_bytes = base_image["image"]
            image_ext = base_image["ext"]
            image_path = os.path.join(output_dir, f"image_{image_count}.{image_ext}")
            with open(image_path, "wb") as image_file:
                image_file.write(image_bytes)
            md_content.append(f"\n![Image {image_count}]({image_path})\n\n")
            image_count += 1
        
        md_content.append("\n\n")
    
    # Post-processing
    final_md = "".join(md_content)
    # Clean up excessive newlines
    final_md = re.sub(r'\n{3,}', '\n\n', final_md)
    # Improve list formatting
    final_md = re.sub(r'(\d+)\. ', r'\n\1. ', final_md)
    # Remove trailing spaces
    final_md = re.sub(r' +(\n)', r'\1', final_md)
    
    # Save markdown file
    md_path = os.path.join(output_dir, "converted.md")
    with open(md_path, "w", encoding="utf-8") as md_file:
        md_file.write(final_md)
    
    return final_md, output_dir

def create_interface():
    with gr.Blocks() as app:
        gr.Markdown("# PDF to Kindle-Friendly Markdown Converter")
        
        with gr.Row():
            with gr.Column():
                pdf_input = gr.File(label="Upload PDF File", type="file")
                image_quality = gr.Slider(1, 10, value=7, label="Image Quality (1-10)")
                convert_btn = gr.Button("Convert to Markdown")
            
            with gr.Column():
                md_output = gr.Textbox(label="Markdown Preview", interactive=False, lines=20)
                download_group = gr.File(label="Download Files")
        
        convert_btn.click(
            fn=pdf_to_kindle_md,
            inputs=[pdf_input, image_quality],
            outputs=[md_output, download_group]
        )
        
        gr.Examples(
            examples=[["sample_report.pdf", 7]],
            inputs=[pdf_input, image_quality]
        )
    
    return app

if __name__ == "__main__":
    app = create_interface()
    app.launch()