[python] pylint runs on tests/ samples/ folders even when not generated by emitter

[l0lawrence]

[github-actions]

🎯 Issue Triage

The Python client emitter is running pylint on tests/ and samples/ directories even when those folders were not generated by the emitter, causing spurious lint failures on pre-existing or user-managed code.

Category: Bug
Area: emitter:client:python

⚠️ No reproduction provided. A minimal repro (e.g., a TypeSpec project with pre-existing tests/ or samples/ folders) would help diagnose this issue.

🔍 Analysis

• The Python emitter's post-generation pylint step appears to scan directories by name (tests/, samples/) rather than limiting its scope to files it actually generated.
• This can cause failures when users have existing test or sample code that doesn't conform to pylint rules enforced by the emitter, even though the emitter never touched those files.
• The fix likely involves tracking which files/directories were emitted and restricting the pylint invocation to only those paths.

📝 Suggested next steps

• Confirm which version introduced this behavior
• Identify the pylint invocation in the emitter pipeline and add a guard to only lint emitter-generated paths
• Add a test case with a pre-existing tests/ folder to prevent regression

Generated by Agentic Triage