revert(llm): remove custom HTTP headers patch now in langchain-nvidia-ai-endpoints v0.3.19

[Pouyanpi]

Description

Reverts commit aafd733 (PR #1461). Custom HTTP headers support is now implemented in langchain-nvidia-ai-endpoints v0.3.19, making this patch redundant (see release notes).

[greptile-apps]

Greptile Summary

• Reverts custom HTTP headers patch (commit aafd733) as functionality now available in langchain-nvidia-ai-endpoints v0.3.19
• Removes custom_headers field, __init__ override, and client method wrapping from ChatNVIDIA patch class
• Deletes 458-line test file test_langchain_nvidia_custom_headers.py containing comprehensive test coverage

Confidence Score: 4/5

• Safe to merge pending dependency version update
• Clean revert that properly removes all custom headers code, but poetry.lock still at v0.3.16 while upstream feature requires v0.3.19 - should verify dependency update
• Verify poetry.lock is updated to langchain-nvidia-ai-endpoints >= 0.3.19 before merging

Important Files Changed

Filename	Overview
nemoguardrails/llm/providers/_langchain_nvidia_ai_endpoints_patch.py	Reverted custom HTTP headers patch - removed custom_headers field, __init__ override, and _wrap_client_methods, restoring to original streaming-only patch
tests/llm_providers/test_langchain_nvidia_custom_headers.py	Deleted entire test file (458 lines) covering custom headers functionality - no longer needed as feature moved upstream

Sequence Diagram

sequenceDiagram
    participant Developer
    participant PR
    participant CodeReview
    Developer->>PR: "Create revert PR #1503"
    Developer->>PR: "Remove custom_headers field from ChatNVIDIA"
    Developer->>PR: "Remove __init__ and _wrap_client_methods"
    Developer->>PR: "Delete test_langchain_nvidia_custom_headers.py"
    Developer->>PR: "Reference langchain-nvidia-ai-endpoints v0.3.19"
    PR->>CodeReview: "Submit for review"
    CodeReview->>Developer: "Verify upstream implementation exists"

Loading

[greptile-apps]

_{2 files reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile
React with 👍 or 👎 to share your feedback on this new summary format}

[tgasser-nv]

Looks good, I guess we need to make sure consumers of the toolkit use the default_headers field rather than custom_headers or it wouldn't work?

For some reason the code-coverage check is failing, I think it's a transient issue though

[codecov]

Codecov Report

❌ Patch coverage is 20.00000% with 4 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
.../providers/_langchain_nvidia_ai_endpoints_patch.py	20.00%	4 Missing ⚠️

📢 Thoughts on this report? Let us know!