What
Add focused tests for CUDA tensor validation behavior in qdp-python.
Current coverage exercises validation indirectly through feature tests, but the validation contract itself is still spread across multiple files and scenarios. This follow-up should add a clearer test layer for the Python-facing CUDA validation rules.
Why
Recent batch float32 CUDA angle binding work exposed how easy it is for test expectations to drift from actual validation behavior.
A dedicated validation-oriented test set would make it easier to catch regressions in:
- supported and unsupported dtypes
- contiguous vs non-contiguous tensors
- empty tensors
- engine/tensor device mismatch
- unsupported encoding methods
This should reduce the chance of stale expectations surviving in downstream tests.
How
- add focused qdp-python tests for CUDA validation behavior
- cover dtype acceptance and rejection across supported encoding methods
- cover non-contiguous, empty, wrong-device, and unsupported-method cases
- keep the scope test-focused; avoid unrelated refactoring unless a small validation bug is uncovered
What
Add focused tests for CUDA tensor validation behavior in qdp-python.
Current coverage exercises validation indirectly through feature tests, but the validation contract itself is still spread across multiple files and scenarios. This follow-up should add a clearer test layer for the Python-facing CUDA validation rules.
Why
Recent batch
float32CUDA angle binding work exposed how easy it is for test expectations to drift from actual validation behavior.A dedicated validation-oriented test set would make it easier to catch regressions in:
This should reduce the chance of stale expectations surviving in downstream tests.
How