[Shivam Sharma (ShivamSharma43)]

Summary

For discriminated subclasses, the generated :vartype of the discriminator property referenced the upper-cased enum member name as if it were a module-level symbol:

:vartype type: str or ~azure.search.documents.knowledgebases.models.MODEL_WEB_SUMMARIZATION

models.MODEL_WEB_SUMMARIZATION does not exist, so the Sphinx cross-reference cannot resolve. With this change the docstring points at the discriminator enum class, which the models module does export:

:vartype type: str or ~azure.search.documents.knowledgebases.models.KnowledgeBaseActivityRecordType

Root cause

EnumValue.docstring_type() in generator/pygen/codegen/models/enum_type.py built the reference as f"{client_namespace}.models.{self.name}", where self.name is the enum member's name, not the enum's. A discriminator property on a subclass is typed as a single EnumValue, so every discriminator subclass got the bad reference. (The type annotation, Literal[KnowledgeBaseActivityRecordType.MODEL_WEB_SUMMARIZATION], was already correct.)

Fix

Use self.enum_type.name so the reference points at the enum class. This matches what EnumType.docstring_type() already produces for the base class. generation-subdir handling is unchanged.

Tests

Added to tests/unit/test_typeddict.py, in the existing EnumValue section:

• test_enum_value_docstring_type_references_enum_class: the docstring type is str or ~<ns>.models.Color and matches the parent EnumType's docstring type.
• test_enum_value_docstring_type_with_generation_subdir: generation-subdir is still included in the path.
• test_discriminator_vartype_references_enum_class: DpgModelSerializer.variable_documentation_string on a discriminator property emits :vartype type: str or ~<ns>.models.ActivityRecordType and not the member name.

All three fail on main (for example, - ...models.ActivityRecordType / + ...models.MODEL_WEB_SUMMARIZATION) and pass with this fix. black --check and pylint are clean on the changed files. A chronus changeset (fix) is included.

Fixes #10575

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).
1 pipeline(s) were filtered out due to trigger conditions.
There may be pipelines that require an authorized user to comment /azp run to run.

[microsoft-github-policy-service]

Hi @Shivam Sharma (@ShivamSharma43). Your PR has had no update for 14 days and it is marked as a stale PR. If it is not updated within 14 additional days, the PR will automatically be closed. The stale label is removed automatically when you update or comment on your PR, or when a maintainer reviews it. You can also remove the label manually.

[Shivam Sharma (ShivamSharma43)]

Merged the latest main (no conflicts; the change itself is unchanged: :vartype for a discriminator subclass now points at the enum member that exists in the generated code). Still relevant on current main. iscai-msft Yuchao Yan (@msyyc) could one of you take a look when you have a moment?

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[pkg-pr-new]

Open in StackBlitz

npm i https://pkg.pr.new/@typespec/http-client-python@12127

commit: dcd98c1

[github-actions]

All changed packages have been documented.

• ✅ @typespec/http-client-python

Show changes

@typespec/http-client-python - fix ✏️

Fix the :vartype docstring of constant enum values (such as discriminator properties on polymorphic subclasses) to reference the enum class instead of a non-existent module-level symbol named after the enum member.

[github-actions]

Python emitter diff

Baseline gh:88e5bd18b07d0cbd3bbe92946aff36beaf707b54 vs this PR.

Diff summary: 4 file(s), +6 / -6

Rendered diff: inline on the run summary, or the emitter-diff-html artifact.

Informational check (eng/emitter-diff); does not block the PR.

[iscai-msft]

looks great, thanks so much Shivam Sharma (@ShivamSharma43)! sorry for the delay in getting back to you. thank you so much for contributing, Libba Lawrence (@l0lawrence) do you know when the next python release is?

[Libba Lawrence (l0lawrence)]

looks great, thanks so much Shivam Sharma (Shivam Sharma (@ShivamSharma43))! sorry for the delay in getting back to you. thank you so much for contributing, Libba Lawrence (Libba Lawrence (@l0lawrence)) do you know when the next python release is?

should be one later this month or early next month