Recover specific tests in v5

Overview

In v5 branch we still have some tests fail because of the codebase changes. For every failing test we need update the test to match a new v5 behaviour (e.g. updating fixture). For some cases, it can be a bug in the codebase (ping @roll to check)

List of tests

The tests that need to be fixed that way is marked as:

@pytest.mark.xfail(reason="issue-1205")

We can use search to find them, for example - https://github.com/frictionlessdata/frictionless-py/blob/v5/tests/resource/test_convert.py#L92

[roll]

@shashigharti For metadata conversion tests consider assertions with parsed data instead of textual data e.g.:

@pytest.mark.xfail(reason="issue-1205")
def test_pipeline_to_yaml():
    pipeline = Pipeline.from_descriptor("data/pipeline.json")
    expected = "data/fixtures/convert/pipeline.yaml"  # also we can use one from data folder
    assert yaml.safe_load(pipeline.to_yaml()) == yaml.safe_load(expected)

Otherwise it might be too fragile

[roll]

Currently:

tests/package/test_convert.py
277:@pytest.mark.xfail(reason="issue-1205")
343:@pytest.mark.xfail(reason="issue-1205")

tests/resource/test_convert.py
92:@pytest.mark.xfail(reason="issue-1205")

tests/resource/analyze/test_general.py
165:@pytest.mark.xfail(reason="issue-1205")
183:@pytest.mark.xfail(reason="issue-1205")
203:@pytest.mark.xfail(reason="issue-1205")

tests/pipeline/test_convert.py
24:@pytest.mark.xfail(reason="issue-1205")
46:@pytest.mark.xfail(reason="issue-1205")