Description
Today, the python test suite covers basic happy paths for connectivity, stream/topic creation, polling, TLS, and a few error cases, but it does not exercise the Python binding layer broadly or systematically. There is also no property-based testing, and thus invariants around payload conversion, polling strategy construction, auto-commit configuration, and connection-string parsing are not being stress-tested across a wide input space.
We should strengthen the Python SDK test suite so it can catch binding regressions, input-handling bugs, and edge cases earlier.
Component
Python SDK
Proposed solution
- Add property-based testing support to foreign/python test dependencies.
- Add hypothesis to the testing extras in foreign/python/pyproject.toml.
- Add broader tests for Python-exposed API behavior.
- Cover valid and invalid inputs, boundary values, Unicode payloads, binary payloads, empty payloads, large payloads, and malformed connection strings.
- Add property-based tests for binding invariants.
- Expand the existing suite instead of only adding more happy-path tests.
This work will be done in multiple PR's instead of a single one since the change is large.
Alternatives considered
No response
Description
Today, the python test suite covers basic happy paths for connectivity, stream/topic creation, polling, TLS, and a few error cases, but it does not exercise the Python binding layer broadly or systematically. There is also no property-based testing, and thus invariants around payload conversion, polling strategy construction, auto-commit configuration, and connection-string parsing are not being stress-tested across a wide input space.
We should strengthen the Python SDK test suite so it can catch binding regressions, input-handling bugs, and edge cases earlier.
Component
Python SDK
Proposed solution
This work will be done in multiple PR's instead of a single one since the change is large.
Alternatives considered
No response