| Thanks for your great open source work on the papercopilot project! | |
| I recently used the keyword search feature on the papercopilot website and noticed that it places a significant burden on the server, resulting in slow response times, unresponsiveness, and lag. Out of concern for these performance issues and a desire to improve the user experience, I have created this pull request. | |
| ## PR Title | |
| Add Enhanced Local Paper Search Tool with Improved UX and Multi-keyword Support | |
| ## PR Description | |
| This PR adds an improved local paper search tool that provides a more efficient way to search and filter conference papers. The tool includes both a web interface (using Streamlit) and command-line functionality, with enhanced user experience focused on typical search behaviors. | |
| ### Key Features | |
| - Fast local paper searching with improved keyword handling | |
| - Multiple keyword search with AND/OR operators | |
| - Default filtering of rejected/withdrawn papers (aligned with user expectations) | |
| - Advanced filter options for including rejected papers when needed | |
| - More intuitive metrics display in the UI | |
| - Multi-field search support | |
| - Results download in JSON format | |
| - User-friendly web interface with better feedback | |
| - Command-line support with additional options | |
| - Directory and multi-conference search capabilities | |
| ### Changes | |
| - Added `app.py` for Streamlit interface | |
| - Updated `extract.py` with enhanced keyword parsing and filtering options | |
| - Removed explicit dependence on config.py by inlining configuration | |
| - Improved code organization with helper functions for better maintainability | |
| - Added caching for performance optimization | |
| - Updated UI to provide better guidance and feedback to users | |
| This tool helps reduce server load by allowing users to perform searches locally while maintaining all the functionality of the web version. It also introduces new features like multi-keyword search modes and smarter default filtering that better aligns with user expectations. | |
| The changes have been thoroughly tested and adhere to the existing coding style and best practices. I believe these enhancements will greatly benefit users and contribute to the overall success of the papercopilot project. | |
| Please let me know if you have any questions or feedback regarding this PR. I'm open to discussion and further improvements. | |
| Thank you for considering this contribution! |