Skip to content

Python: Current Python CI ignore most of the unit tests #7135

@JonasJ-ap

Description

@JonasJ-ap

Apache Iceberg version

main (development)

Query engine

None

Please describe the bug 🐞

Currently, the two tasks: Test S3 and Test Adlfs in Python CI only runs some tests in test_fsspec.py, ignoring other unit tests:

Screenshot of a recent Python CI log:

Screen Shot 2023-03-18 at 01 24 06
Screen Shot 2023-03-18 at 01 24 25

I also checked the Python CI (integration test) and it only runs tests in test_integration.py. So it seems most of the unit tests are ignored now.

I also noticed that there are two unit tests failing in the current master branch when running locally:

_____________________________________________________________________________________________________________________________________________ test_expr_is_nan_to_pyarrow ______________________________________________________________________________________________________________________________________________

bound_double_reference = BoundReference(field=NestedField(field_id=1, name='foo', field_type=DoubleType(), required=False), accessor=Accessor(position=0,inner=None))

    def test_expr_is_nan_to_pyarrow(bound_double_reference: BoundReference[str]) -> None:
>       assert (
            repr(expression_to_pyarrow(BoundIsNaN(bound_double_reference)))
            == "<pyarrow.compute.Expression (is_null(foo, {nan_is_null=true}) and is_valid(foo))>"
        )
E       AssertionError: assert '<pyarrow.com... is_nan(foo)>' == '<pyarrow.com..._valid(foo))>'
E         - <pyarrow.compute.Expression (is_null(foo, {nan_is_null=true}) and is_valid(foo))>
E         + <pyarrow.compute.Expression is_nan(foo)>

tests/io/test_pyarrow.py:453: AssertionError
_____________________________________________________________________________________________________________________________________________ test_expr_not_nan_to_pyarrow _____________________________________________________________________________________________________________________________________________

bound_double_reference = BoundReference(field=NestedField(field_id=1, name='foo', field_type=DoubleType(), required=False), accessor=Accessor(position=0,inner=None))

    def test_expr_not_nan_to_pyarrow(bound_double_reference: BoundReference[str]) -> None:
>       assert (
            repr(expression_to_pyarrow(BoundNotNaN(bound_double_reference)))
            == "<pyarrow.compute.Expression invert((is_null(foo, {nan_is_null=true}) and is_valid(foo)))>"
        )
E       AssertionError: assert '<pyarrow.com...is_nan(foo))>' == '<pyarrow.com...valid(foo)))>'
E         - <pyarrow.compute.Expression invert((is_null(foo, {nan_is_null=true}) and is_valid(foo)))>
E         + <pyarrow.compute.Expression invert(is_nan(foo))>

tests/io/test_pyarrow.py:460: AssertionError

Seems related to the changes made in #6398 . But before making any change, I want to post this issue to make sure I did not misunderstood anything.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions