Deep Doc2Markdown

中文文档 | English

A powerful document-to-markdown converter that extracts text and images from PDF, Word documents, and images, using OCR where needed, to generate clean Markdown output.

Features

• Document Processing:

  • Extract text and images from PDF and Word documents
  • Process standalone images with OCR
  • Use PaddleOCR for high-quality optical character recognition
  • Convert document structure to proper Markdown formatting
  • Smart handling of multi-column layouts

• Modern User Interface:

  • Clean, responsive design with Bootstrap 5
  • Drag & drop file upload interface
  • Real-time conversion status display
  • Tabbed interface for viewing Markdown, original document, and rendered preview
  • Syntax highlighting for Markdown code
  • Content scrolling with custom scrollbars for long documents
  • Smart paragraph detection and formatting for better readability
  • Optimized layout for both desktop and mobile devices
  • Dark mode support

• User Experience:

  • Copy Markdown to clipboard with one click
  • Download converted Markdown as a file
  • Print functionality for rendered Markdown
  • Instant rendering of Markdown to HTML preview
  • Error handling with descriptive messages
  • Auto-refresh UI components

Technologies

• Frontend:

  • HTML5, CSS3, and JavaScript (ES6+)
  • Bootstrap 5 for responsive design
  • Highlight.js for syntax highlighting
  • Marked.js for Markdown rendering
  • Custom CSS for enhanced UI components
  • AnimateCSS for smooth transitions

• Backend:

  • Python 3.8+ with FastAPI
  • PyPDF2 for PDF processing
  • python-docx for Word document processing
  • PaddleOCR for image text recognition
  • Jinja2 for templating
  • Asynchronous processing for improved performance

• Data Processing:

  • Text cleaning and formatting algorithms
  • Smart paragraph detection
  • Intelligent handling of document structure
  • Image extraction and processing
  • OCR confidence scoring and validation

Project Structure

document-to-markdown/
│
├── app/
│   ├── core/
│   │   ├── document_extractor/      # Extract content from documents
│   │   │   ├── pdf_extractor.py     # Extract text and images from PDF
│   │   │   ├── docx_extractor.py    # Extract text and images from Word
│   │   │   └── image_handler.py     # Process standalone images
│   │   │
│   │   ├── ocr/
│   │   │   ├── paddle_ocr.py        # PaddleOCR implementation
│   │   │   └── ocr_processor.py     # OCR processing workflow
│   │   │
│   │   ├── text_processor/          # Text processing module
│   │   │   ├── text_cleaner.py      # Text cleaning
│   │   │   └── text_merger.py       # Merge document and OCR text
│   │   │
│   │   └── markdown_converter/      # Markdown conversion module
│   │       ├── md_formatter.py      # Text to Markdown conversion
│   │       ├── image_formatter.py   # Image processing and Markdown syntax
│   │       └── structure_parser.py  # Parse document structure
│   │
│   ├── media/                       # Store processed images
│   │   └── images/                  
│   │
│   ├── api/                         # API endpoints
│   │   ├── document_api.py          # Document processing API
│   │   └── static_files.py          # Serve static files
│   │
│   ├── models/                      # Data models
│   │   └── document_models.py       # Document data models
│   │
│   └── frontend/                    # Web interface
│       ├── frontend_api.py          # Frontend API
│       └── templates/               # HTML templates
│           └── index.html           # Main app page
│
├── config/                          # Configuration
│   └── config.py                    # App configuration
│
├── requirements.txt                 # Dependencies
└── main.py                          # Main application

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/deep-doc2markdown.git
   cd deep-doc2markdown
   

2. Create a virtual environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   

3. Install dependencies:

   pip install -r requirements.txt
   

4. Install PaddlePaddle:

   Depending on your system, you might need specific installation instructions for PaddlePaddle. Visit the PaddlePaddle installation guide for details.

Usage

1. Start the application:

   python main.py
   

2. Open your browser and navigate to:

   http://127.0.0.1:8000
   

3. Use the web interface to upload your documents and convert them to Markdown.

API Endpoints

• POST /api/upload: Upload a document for conversion
• GET /api/status/{doc_id}: Check the status of a conversion job
• GET /api/markdown/{doc_id}: Get the generated Markdown content

Development

Requirements

• Python 3.8+
• FastAPI
• PyPDF2 for PDF processing
• python-docx for Word document processing
• PaddleOCR for image text recognition
• Jinja2 for templating

Configuration

Edit config/config.py to adjust settings:

• Change the server host and port
• Configure OCR settings
• Modify file storage paths

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgements

• PaddleOCR for the OCR functionality
• FastAPI for the web framework
• PyPDF2 for PDF processing
• Bootstrap for the UI framework
• Highlight.js for code highlighting
• Marked.js for Markdown parsing