feat: add arxiv tool. resolves #252

[zhjuzi]

feat: Add ArxivTool for Academic Paper Search and Retrieval

Description

Added a new ArxivTool that enables searching and retrieving academic papers from arXiv. This tool enhances the system's ability to access and process academic research papers, supporting both search by keywords and full paper text retrieval.

Features

1. ArxivTool Implementation (arxiv_tool.py):

   • Paper search functionality with keyword support
   • Full paper text retrieval using paper ID
   • Smart query processing with length limitation
   • Comprehensive paper metadata extraction (title, authors, date, summary)
   • Retry mechanism for handling network issues
   • Clean error handling and informative error messages
   • 
   • 

2. Test Implementation (test_arxiv_tool.py):

   • Unit tests for paper search functionality
   • Unit tests for paper detail retrieval
   • Unit tests for invalid mode handling
   • Integration tests with actual arXiv API

3. Utility Implementation (utils/retry.py):

   • Generic retry decorator for handling transient failures
   • Configurable retry count and delay
   • Detailed error reporting with retry attempts

Technical Details

• Built on top of the base Tool class
• Uses arxiv library for API interaction
• Uses pypdf for PDF text extraction
• Implements retry mechanism for network resilience
• Configurable search parameters (results count, sort criteria)

Test Results

• All unit tests pass successfully
• Test coverage for all main functionalities
• Real API integration tested and working
• 

Dependencies

• arxiv
• pypdf
• pydantic

Related Issue

Implements #issue252

Checklist

• [√] Read and understood the Contributor Guidelines
• [√] Code follows project's coding style (PEP8 and Google Python Style)
• [√] Added comprehensive test cases
• [√] All tests pass
• [√] Added necessary documentation
• [√] Added type hints and docstrings
• [√] Implemented retry mechanism for network resilience
• [√] Proper error handling and user feedback