All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- LangChain Tools module (
src/tools/)ReceiptOCRTool- LangChain-compatible OCR toolReceiptExtractionTool- Information extraction toolReceiptClassificationTool- Receipt classification tool
- LangChain Agent module (
src/agents/)create_receipt_agent()- Factory function for receipt processing agentReceiptAgentExecutor- High-level agent wrapper
- AGENTS.md - AI-Native repository guide
- README_EN.md - English version of README
- CONTRIBUTING.md - Contribution guidelines
- CODE_OF_CONDUCT.md - Community code of conduct
- SECURITY.md - Security policy
.pre-commit-config.yaml- Pre-commit hooks configuration
- Added language switcher links in README.md
0.2.0 - 2026-03-11
- Official
langchain-paddleocrpackage integration uvpackage manager support- LangChain compatibility layer (
langchain_compat.py) - New environment variable
PADDLEOCR_ACCESS_TOKEN - Seal extraction utility module (
utils/seal_extractor.py)
- Migrated from custom OCR implementation to
PaddleOCRVLLoader - Improved dependency management with
pyproject.toml - Updated documentation for new API configuration
- Import compatibility issues with
langchain-paddleocr
0.1.0 - 2024-01-01
- Initial release
- PaddleOCR-VL-1.5 OCR recognition
- ERINE (Baidu LLM) integration for information extraction
- Seal/stamp recognition feature
- Support for multiple receipt types:
- VAT Special Invoice (增值税专用发票)
- VAT Normal Invoice (增值税普通发票)
- Train Tickets (火车票)
- Taxi Receipts (出租车票)
- Gradio Web UI
- CLI interface
- Sample invoices for testing
| Version | Date | Highlights |
|---|---|---|
| 0.2.0 | 2026-03-11 | Official langchain-paddleocr integration, uv support |
| 0.1.0 | 2024-01-01 | Initial release with OCR, ERINE, seal recognition |