miniocr 0.0.3

A small OCR package

MiniOCR

A powerful and easy-to-use Python package for performing Optical Character Recognition (OCR) on images, PDF documents, and PowerPoint presentations using OpenAI's Vision API.

Features

• 🖼️ Multi-format support: Process images (PNG, JPG, JPEG, GIF, BMP, TIFF, WebP), PDF files, and PPTX presentations
• ⚡ Parallel processing: Concurrent processing of multiple pages/slides for improved performance
• 🌐 Cross-platform: Works on Windows, macOS, and Linux
• 📄 Visual OCR for PPTX: Converts PowerPoint slides to images for accurate visual content extraction
• 🔄 Async support: Built with asyncio for efficient processing
• 📝 Markdown output: Converts documents to clean, structured markdown format

Installation

Prerequisites

For PDF processing, you'll need to install Poppler:

macOS:

brew install poppler

Ubuntu/Debian:

sudo apt-get install poppler-utils

Windows: Download and install from Poppler for Windows

For PowerPoint (.pptx) processing, you'll need to install LibreOffice:

macOS:

brew install --cask libreoffice

Ubuntu/Debian:

sudo apt-get install libreoffice

Windows: Download and install from LibreOffice official website

Install MiniOCR

pip install miniocr

Or install from source:

git clone https://github.com/w95/miniocr.git
cd miniocr
pip install -e .

Quick Start

Setup

First, you'll need an OpenAI API key. Set it as an environment variable:

export OPENAI_API_KEY="your-api-key-here"

Or pass it directly when initializing the class.

Basic Usage

import asyncio
from miniocr import MiniOCR

async def main():
    # Initialize with API key (or use environment variable)
    ocr = MiniOCR(api_key="your-api-key-here")
    
    # Process an image
    result = await ocr.ocr("path/to/image.jpg")
    print(result["content"])
    
    # Process a PDF
    result = await ocr.ocr("path/to/document.pdf")
    print(f"Processed {result['pages']} pages")
    print(result["content"])
    
    # Process a PowerPoint presentation
    result = await ocr.ocr("path/to/presentation.pptx")
    print(result["content"])

if __name__ == "__main__":
    asyncio.run(main())

Advanced Usage

import asyncio
from miniocr import MiniOCR

async def advanced_example():
    ocr = MiniOCR()
    
    # Process with custom settings
    result = await ocr.ocr(
        file_path="document.pdf",
        model="gpt-4o",  # Use different OpenAI model
        concurrency=10,  # Process up to 10 pages simultaneously
        output_dir="./output",  # Save markdown to file
        cleanup=True  # Clean up temporary files
    )
    
    print(f"File: {result['file_name']}")
    print(f"Pages processed: {result['pages']}")
    print(f"Content length: {len(result['content'])} characters")

asyncio.run(advanced_example())

Processing URLs

import asyncio
from miniocr import MiniOCR

async def process_url():
    ocr = MiniOCR()
    
    # Process a file from URL
    result = await ocr.ocr("https://example.com/document.pdf")
    print(result["content"])

asyncio.run(process_url())

API Reference

MiniOCR Class

__init__(api_key: str = None)

Initialize the MiniOCR instance.

Parameters:

• api_key (str, optional): OpenAI API key. If not provided, will use OPENAI_API_KEY environment variable.

async ocr(file_path, model="gpt-4o-mini", concurrency=5, output_dir=None, cleanup=True)

Process a file and extract text using OCR.

Parameters:

• file_path (str): Path or URL to the file to process
• model (str): OpenAI model to use (default: "gpt-4o-mini")
• concurrency (int): Number of concurrent API requests (default: 5)
• output_dir (str, optional): Directory to save markdown output
• cleanup (bool): Whether to clean up temporary files (default: True)

Returns:

• dict: Dictionary containing:
  • content (str): Extracted text in markdown format
  • pages (int): Number of pages/slides processed
  • file_name (str): Name of the processed file

Supported file types:

• Images: .png, .jpg, .jpeg, .gif, .bmp, .tiff, .webp
• Documents: .pdf
• Presentations: .pptx

Configuration

Environment Variables

• OPENAI_API_KEY: Your OpenAI API key (required)

Model Options

MiniOCR supports various OpenAI models:

• gpt-4o-mini (default, cost-effective)
• gpt-4o (higher accuracy)
• gpt-4-turbo

Output Format

MiniOCR converts documents to clean markdown with the following features:

• Tables: Converted to HTML format for better structure
• Checkboxes: Represented as ☐ (unchecked) and ☑ (checked)
• Special elements: Logos, watermarks, and page numbers are wrapped in brackets
• Charts and infographics: Interpreted and converted to markdown tables when applicable

Error Handling

import asyncio
from miniocr import MiniOCR

async def handle_errors():
    ocr = MiniOCR()
    
    try:
        result = await ocr.ocr("nonexistent.pdf")
    except ValueError as e:
        print(f"Unsupported file type: {e}")
    except Exception as e:
        print(f"Processing error: {e}")

asyncio.run(handle_errors())

Performance Tips

1. Adjust concurrency: Increase concurrency parameter for faster processing of multi-page documents
2. Use appropriate models: gpt-4o-mini for cost-effectiveness, gpt-4o for higher accuracy
3. Process in batches: For large numbers of files, process them in batches to avoid rate limits
4. Local processing: Keep files local when possible to avoid download overhead

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Testing

Run the test suite:

pytest tests/

License

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

Changelog

v0.0.1

• Initial release
• Support for images, PDF, and PPTX files
• Async processing with concurrency control
• Cross-platform compatibility

Support

If you encounter any issues or have questions, please open an issue on GitHub.

Acknowledgments

• Built with OpenAI's Vision API
• Uses pdf2image for PDF processing
• Uses python-pptx for PowerPoint processing