All-in-MCP

# CryptoBib Integration This document describes the CryptoBib integration in the all-in-mcp server. ## Overview CryptoBib (https://cryptobib.di.ens.fr/) is a comprehensive bibliography database for cryptography research papers. This integration allows searching through the CryptoBib database and retrieving BibTeX entries. ## Features ### Search Functionality - Search through the entire CryptoBib database - Case-insensitive search across all BibTeX fields - Returns both parsed Paper objects and raw BibTeX entries - Configurable result limits ### BibTeX Support - Parse BibTeX entries into structured Paper objects - Retrieve raw BibTeX entries for direct use - Support for all standard BibTeX fields (author, title, booktitle, journal, year, etc.) - Clean LaTeX formatting from titles and authors ## Available Tools ### search-cryptobib-papers Search CryptoBib bibliography database for cryptography papers. **Parameters:** - `query` (required): Search query string (e.g., 'lattice', 'homomorphic encryption') - `max_results` (optional, default: 10): Maximum number of results to return - `return_bibtex` (optional, default: false): Whether to return raw BibTeX entries - `force_download` (optional, default: false): Force download the newest crypto.bib file **Examples:** ```json { "query": "lattice cryptography", "max_results": 5, "return_bibtex": false } ``` To get specific BibTeX entries, you can use targeted search queries with `return_bibtex=true`. ## Usage Examples ### Basic Search ```python from all_in_mcp.academic_platforms.cryptobib import CryptoBibSearcher searcher = CryptoBibSearcher() # Search for papers papers = searcher.search("homomorphic encryption", max_results=5) for paper in papers: print(f"Title: {paper.title}") print(f"Authors: {', '.join(paper.authors)}") print(f"BibTeX Key: {paper.paper_id}") ``` ### Get Raw BibTeX ```python # Get raw BibTeX entries bibtex_entries = searcher.search_bibtex("zero knowledge", max_results=3) for entry in bibtex_entries: print(entry) ``` ### Get Specific BibTeX Entry ```python # Search for specific entries using targeted queries papers = searcher.search("ACISP:LZXSW24", max_results=1) if papers: # Access the BibTeX from the paper's extra field bibtex = papers[0].extra.get('bibtex', '') print(bibtex) # Or get raw BibTeX entries directly bibtex_entries = searcher.search_bibtex("ACISP:LZXSW24", max_results=1) if bibtex_entries: print(bibtex_entries[0]) ``` ## BibTeX Entry Format CryptoBib uses standard BibTeX format. Example entry: ```bibtex @InProceedings{ACISP:LZXSW24, author = "Fangzhou Liu and Xueqi Zhu and Ruozhou Xu and Danping Shi and Peng Wang", title = "The Offline Quantum Attack Against Modular Addition Variant of {Even}-{Mansour} Cipher", pages = "3--19", editor = acisp24ed, booktitle = acisp24name1, volume = acisp24vol1, address = acisp24addr, month = acisp24month, publisher = acisp24pub, series = mylncs, year = 2024, doi = "10.1007/978-981-97-5025-2_1", } ``` ## Implementation Details ### Search Strategy The implementation uses streaming HTTP requests to search through the large crypto.bib file without downloading it entirely. It: 1. Makes a streaming request to the crypto.bib file 2. Parses BibTeX entries on-the-fly 3. Performs case-insensitive matching against the search query 4. Returns results as they are found up to the specified limit ### Performance Considerations - Uses streaming to avoid memory issues with the large bibliography file - Processes entries incrementally - Stops searching once the maximum number of results is reached - Implements proper brace counting for correct BibTeX entry parsing ### Limitations - No PDF download capability (CryptoBib is a bibliography database only) - No full-text search of paper content - Search is limited to BibTeX field content - Dependent on CryptoBib server availability ## Testing Run the CryptoBib tests: ```bash # Run all tests pytest tests/test_cryptobib.py # Run with verbose output pytest tests/test_cryptobib.py -v # Test specific functionality python test_cryptobib.py ``` ## Error Handling The implementation includes comprehensive error handling for: - Network connectivity issues - Malformed BibTeX entries - Server response errors - Parsing failures All errors are logged and gracefully handled, returning empty results or error messages as appropriate.