fix: remove URLs in pydantic error messages

[ssonal]

Fixes #509

Pydantic includes error URLs in ValidationErrors which means extra tokens used in LLM calls.

Pydantic relies on an environment variable to control this setting and is on by default. https://github.com/pydantic/pydantic-core/blob/e1fc99dd3207157aad77defc20ab6873fd268b5b/python/pydantic_core/_pydantic_core.pyi#L818

This change sets the environment variable dynamically when the package is invoked.

---

🚀 This description was created by Ellipsis for commit 5c2a359

Summary:

This PR sets an environment variable to remove URLs from Pydantic error messages and includes a test to verify the change.

Key points:

• Set PYDANTIC_ERRORS_INCLUDE_URL to 0 in set_env_variable in /instructor/utils.py
• Call set_env_variable in /instructor/__init__.py
• Add test in /tests/test_function_calls.py to verify URLs are not in Pydantic error messages

---

Generated with ❤️ by ellipsis.dev

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to 5c2a359 in 32 seconds

More details

• Looked at 70 lines of code in 3 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. instructor/__init__.py:26

• Draft comment:
  Setting environment variables at runtime can lead to inconsistent behavior across different parts of the application, especially in a multi-threaded or asynchronous context. Consider setting this environment variable outside the application scope, for example through configuration management or directly in the environment where the application runs.
• Reason this comment was not posted:
  Confidence changes required: 80%
  The PR aims to remove URLs from Pydantic error messages by setting an environment variable. The implementation sets the environment variable PYDANTIC_ERRORS_INCLUDE_URL to "0" in the set_env_variable function in utils.py. This function is called in __init__.py to ensure the variable is set when the package is used. The test test_pylance_url_config in test_function_calls.py checks if the URL is not included in the error message, which aligns with the PR's goal. However, setting environment variables at runtime can be problematic because it might not affect already imported modules or those imported in parallel at startup. This could lead to inconsistent behavior where some parts of the application still show URLs in error messages.

Workflow ID: wflow_iQ2aBd5qo3ROeUh0

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[jxnl]

I would prefer this to be a piece of the documentation rather than something that we force users to use.

[ssonal]

This does sound like a sane default to me, especially when retries come into the picture. Would be cumbersome to have to set this env variable each time. But maybe something like a config file makes more sense instead?