refactor: Enhance tools with enums and improved error handling

[ogabrielluiz]

This pull request includes several updates to various search tool components in the backend. The main changes involve adding exception handling, updating schema definitions, and refactoring some methods for better code organization and clarity.

Exception Handling Improvements:

• Added ToolException handling in duck_duck_go_search_run.py, glean_search_api.py, serp_api.py, tavily_search.py, and yahoo_finance.py to provide more informative error messages and improve error handling. [1] [2] [3] [4] [5]

Schema and Input Refactoring:

• Introduced new schema classes (SerpAPISchema, TavilySearchSchema, GleanSearchAPISchema) to better define input parameters and improve code readability. [1] [2] [3]
• Updated TavilySearchToolComponent to use enums for search_depth and topic parameters, ensuring valid values and improving code clarity. [1] [2]

Method Refactoring:

• Refactored _build_wrapper methods in serp_api.py and glean_search_api.py to accept parameters and improve flexibility. [1] [2]
• Updated run_model methods to include type hints and handle type assignments more gracefully. [1] [2] [3]

Miscellaneous Changes:

• Added type: ignore comments to suppress type checking warnings where necessary. [1] [2] [3]
• Included additional imports for Enum and ToolException where needed. [1] [2]

[codspeed-hq]

CodSpeed Performance Report

Merging #4493 will degrade performances by 12.66%

_{Comparing improve-tools (71ca5ea) with main (bbaec2b)}

Summary

⚡ 1 improvements
❌ 1 regressions
✅ 13 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Benchmarks breakdown

	Benchmark	main	improve-tools	Change
❌	test_successful_run_with_input_type_any	454.8 ms	520.7 ms	-12.66%
⚡	test_successful_run_with_input_type_text	532.6 ms	457.7 ms	+16.37%

[erichare]

So much cleaner. I tested the Glean Search API one as I was involved in that, works great and i love that the class structure is easier to follow now