Fix: Python cancellation test

[mrkaye97]

Description

Broke the DAG outputs in the REST API a couple of commits ago, so this works around that. going to fix that breakage forward in #3689

Type of change

• Bug fix (non-breaking change which fixes an issue)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
hatchet-docs	Ready	Preview, Comment	Apr 24, 2026 2:26pm

[Copilot]

Pull request overview

Updates the Python cancellation integration test to use the run-details API and the SDK’s unified RunStatus enum so the test can reliably observe cancellation without relying on REST task status/output fields.

Changes:

• Replace REST V1TaskStatus checks with RunStatus.
• Switch polling from hatchet.runs.aio_get(...) to hatchet.runs.aio_get_details(...).
• Treat QUEUED as an in-progress state while waiting for cancellation.

Comments suppressed due to low confidence (1)

sdks/python/examples/cancellation/test_cancellation.py:24

• The timeout failure (assert False, "Workflow run did not cancel in time") doesn’t include the last observed run status/details, which makes CI failures hard to diagnose. Consider capturing the final run.status (and optionally run.model_dump_json(...)) and including it in the failure message in the else branch.

    else:
        assert False, "Workflow run did not cancel in time"