[Bug]: coders_property_based_test flaky

[Abacn]

What happened?

e.g. https://ci-beam.apache.org/job/beam_PreCommit_Python_Commit/24753/

following test failed

apache_beam.coders.coders_property_based_test.ProperyTestingCoders.test_row_coder
apache_beam.coders.coders_property_based_test.ProperyTestingCoders.test_row_coder

error message:

ValueError: Type names and field names cannot be a keyword: 'as'
Falsifying example: test_row_coder(
    data=data(...),
    self=<apache_beam.coders.coders_property_based_test.ProperyTestingCoders testMethod=test_row_coder>,
)
Draw 1: [('as', <class 'str'>, False)]

You can reproduce this example by temporarily adding @reproduce_failure('6.54.5', b'AAEBJQE3AAAAAA==') as a decorator on your test case
Stacktrace
self = <apache_beam.coders.coders_property_based_test.ProperyTestingCoders testMethod=test_row_coder>

    @settings(deadline=None, print_blob=True)
>   @given(st.data())

apache_beam/coders/coders_property_based_test.py:119: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
apache_beam/coders/coders_property_based_test.py:135: in test_row_coder
    RowType = typing.NamedTuple(  # type: ignore
/usr/lib/python3.8/typing.py:1704: in __new__
    return _make_nmtuple(typename, fields)
/usr/lib/python3.8/typing.py:1594: in _make_nmtuple
    nm_tpl = collections.namedtuple(name, [n for n, t in types])
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

typename = 'RandomRowType', field_names = ['as']

    def namedtuple(typename, field_names, *, rename=False, defaults=None, module=None):
        """Returns a new subclass of tuple with named fields.
    
        >>> Point = namedtuple('Point', ['x', 'y'])
        >>> Point.__doc__                   # docstring for the new class
        'Point(x, y)'
        >>> p = Point(11, y=22)             # instantiate with positional args or keywords
        >>> p[0] + p[1]                     # indexable like a plain tuple
        33
        >>> x, y = p                        # unpack like a regular tuple
        >>> x, y
        (11, 22)
        >>> p.x + p.y                       # fields also accessible by name
        33
        >>> d = p._asdict()                 # convert to a dictionary
        >>> d['x']
        11
        >>> Point(**d)                      # convert from a dictionary
        Point(x=11, y=22)
        >>> p._replace(x=100)               # _replace() is like str.replace() but targets named fields
        Point(x=100, y=22)
    
        """
    
        # Validate the field names.  At the user's option, either generate an error
        # message or automatically replace the field name with a valid name.
        if isinstance(field_names, str):
            field_names = field_names.replace(',', ' ').split()
        field_names = list(map(str, field_names))
        typename = _sys.intern(str(typename))
    
        if rename:
            seen = set()
            for index, name in enumerate(field_names):
                if (not name.isidentifier()
                    or _iskeyword(name)
                    or name.startswith('_')
                    or name in seen):
                    field_names[index] = f'_{index}'
                seen.add(name)
    
        for name in [typename] + field_names:
            if type(name) is not str:
                raise TypeError('Type names and field names must be strings')
            if not name.isidentifier():
                raise ValueError('Type names and field names must be valid '
                                 f'identifiers: {name!r}')
            if _iskeyword(name):
>               raise ValueError('Type names and field names cannot be a '
                                 f'keyword: {name!r}')
E               ValueError: Type names and field names cannot be a keyword: 'as'
E               Falsifying example: test_row_coder(
E                   data=data(...),
E                   self=<apache_beam.coders.coders_property_based_test.ProperyTestingCoders testMethod=test_row_coder>,
E               )
E               Draw 1: [('as', <class 'str'>, False)]
E               
E               You can reproduce this example by temporarily adding @reproduce_failure('6.54.5', b'AAEBJQE3AAAAAA==') as a decorator on your test case

/usr/lib/python3.8/collections/__init__.py:361: ValueError

Not sure this is transient or flakes. Will track.

Issue Priority

Priority: 2

Issue Component

Component: test-failures

[Abacn]

CC: @pabloem

[Abacn]

Also happens on cron job: https://ci-beam.apache.org/job/beam_PreCommit_Python_Cron/6125/

[kennknowles]

Is there a rollback possible?

[pabloem]

yeah we can roll this back

[pabloem]

Rollback PR: #23265

[Abacn]

.close-issue