Biomapper

Fixed
- Improved Poetry configuration for better dependency management
- Made DSPy integration more resilient with graceful fallback for missing components
- Enhanced restart script to properly handle biomapper installation in editable mode
- Fixed environment setup to ensure consistent behavior across development setups

Enhanced
- Added support for handling comment lines in CSV/TSV files:
- Created centralized `load_tabular_file` utility function in utils/io_utils.py
- Automatically skips lines starting with '' during data loading
- Intelligently detects file type (CSV/TSV) and selects appropriate separator
- Maintains memory-aware file size limits (50% of available memory)
- Works consistently across both frontend and backend components

Added
- Comprehensive UI documentation with separate technical and user sections
- Integration of Mermaid diagram support for visual workflows in documentation
- Proper Poetry-based documentation build configuration for ReadTheDocs

Enhanced
- Significantly improved handling of large CSV files:
- Removed arbitrary 5MB file size limit in favor of a dynamic approach
- Added intelligent file size detection that sets maximum file size to 50% of available device memory
- Implemented real-time upload progress tracking with visual progress bar
- Enhanced file upload with XMLHttpRequest for accurate progress reporting
- Added file size display in MB next to selected file name
- Modified backend API server configuration to support large CSV files:
- Increased MAX_UPLOAD_SIZE setting from 10MB to 1GB
- Implemented dynamic calculation based on system memory using psutil.virtual_memory().available / 2
- Added 1GB fallback if memory information can't be determined
- Ensured consistent file size limit calculation between frontend and backend

Fixed
- Resolved "413 Payload Too Large" errors when uploading larger datasets

- Initial public release of Biomapper