I often have cases like below when I do not fail a test based on a specific exception attributes.
In case of pytest-rerunfailures the logic can look like:
@pytest.mark.flaky(
only_rerun=ClientResponseError,
condition=lambda error: error.status in {429, 500},
)
async def test_request():
await request("www.example.com")
@pytest.mark.flaky(
only_rerun=CustomDatabaseError,
condition=lambda error: error["code"] != 123456,
)
async def test_database():
await database.set("foo", "bar")
def _callback(error: Exception) -> bool:
if isinstance(error, CustomDatabaseError): ...
if isinstance(error, AnotherError): ...
raise
@pytest.mark.flaky(
only_rerun=[CustomDatabaseError, AnotherError],
condition=_callback,
)
async def test_database():
await database.set("foo", "bar")There are also some 3rd-party packages that raise custom exceptions and I must handle an exception in different ways based on some attributes.
Question:
Is this functionality out of scope of the current plugin and I should write a separate plugin?
I often have cases like below when I do not fail a test based on a specific exception attributes.
In case of
pytest-rerunfailuresthe logic can look like:There are also some 3rd-party packages that raise custom exceptions and I must handle an exception in different ways based on some attributes.
Question:
Is this functionality out of scope of the current plugin and I should write a separate plugin?