All-in-MCP

# IACR ePrint Archive Integration This document describes the integration with the IACR ePrint Archive for academic paper search, download, and reading functionality. ## Overview The IACR (International Association for Cryptologic Research) ePrint Archive is a repository for cryptography and information security research papers. Our integration provides: - Search functionality across all papers - PDF download capabilities - Text extraction from papers - Metadata parsing ## Features ### Search Capabilities The search function queries the IACR archive and returns: - Paper titles - Author information - Publication dates - Abstracts - Direct URLs to papers - PDF download links - Keywords and categories ### Download Functionality - Direct PDF downloads from IACR servers - Automatic file naming based on paper ID - Configurable save directories - Error handling for unavailable papers ### Text Extraction - Full text extraction from downloaded PDFs - Formatted output with metadata - Handling of various PDF formats - Error recovery for corrupted files ## Technical Implementation ### API Endpoints The IACR integration uses: - **Search API**: `https://eprint.iacr.org/search` - **Paper URLs**: `https://eprint.iacr.org/YYYY/NNNN` - **PDF URLs**: `https://eprint.iacr.org/YYYY/NNNN.pdf` ### Data Structure Papers are represented using the standardized `Paper` class: ```python @dataclass class Paper: paper_id: str # Format: "YYYY/NNNN" title: str # Full paper title authors: List[str] # List of author names abstract: str # Paper abstract url: str # Paper page URL pdf_url: str # Direct PDF URL published_date: datetime # Publication date source: str # Always "iacr" categories: List[str] # Subject categories keywords: List[str] # Extracted keywords doi: str # DOI if available ``` ### Error Handling The implementation handles: - Network connectivity issues - Invalid paper IDs - PDF download failures - Text extraction errors - Rate limiting ### Rate Limiting To be respectful to IACR servers: - Search requests are throttled - PDF downloads include delays - Maximum concurrent requests limited - Retry logic with exponential backoff ## Usage Examples ### Basic Search ```python from all_in_mcp.academic_platforms.iacr import IACRSearcher searcher = IACRSearcher() papers = searcher.search("zero knowledge", max_results=10) for paper in papers: print(f"{paper.title} ({paper.paper_id})") ``` ### Download and Read ```python # Download a specific paper pdf_path = searcher.download_pdf("2023/1234", "./downloads") # Extract text content text_content = searcher.read_paper("2023/1234", "./downloads") print(text_content[:500]) # First 500 characters ``` ## Configuration ### Environment Variables - `IACR_DOWNLOAD_DIR`: Default download directory - `IACR_MAX_RETRIES`: Maximum retry attempts - `IACR_TIMEOUT`: Request timeout in seconds ### Default Settings ```python DEFAULT_CONFIG = { "max_retries": 3, "timeout": 30, "download_dir": "./downloads", "user_agent": "All-in-MCP Academic Search" } ``` ## Limitations ### Current Limitations - Search is limited to title and abstract matching - No advanced search operators - PDF extraction may fail for scanned documents - Large papers may take time to process ### Future Enhancements - Advanced search with filters - Author-based search - Citation network analysis - Full-text search capabilities - Batch download operations ## Troubleshooting ### Common Issues **Search returns no results**: - Check query spelling - Try broader search terms - Verify network connectivity **PDF download fails**: - Check paper ID format (YYYY/NNNN) - Verify paper exists on IACR - Check file system permissions **Text extraction incomplete**: - PDF may be image-based - Try re-downloading the PDF - Check for PDF corruption ### Debug Mode Enable debug logging: ```python import logging logging.basicConfig(level=logging.DEBUG) ``` This will show detailed information about: - HTTP requests and responses - PDF processing steps - Error details and stack traces