[
https://issues.apache.org/jira/browse/BEAM-5628?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16647969#comment-16647969
]
Simon commented on BEAM-5628:
-----------------------------
This error can be traced back to the _create_generator function (io/vcfio.py:
line 318), where it is mentioned that PyVCF has explicit str() calls when
parsing INFO fields, which fails with UTF-8 decoded strings. For this reason,
the line is encoded back to UTF-8 in the python2 version.
Because removing the encoding step results in hanging of some tests, there is a
chance this relates to 5623.
Does anyone have additional insights?
> Several VcfIO tests fail in Python 3 with TypeError: cannot use a string
> pattern on a bytes-like object
> --------------------------------------------------------------------------------------------------------
>
> Key: BEAM-5628
> URL: https://issues.apache.org/jira/browse/BEAM-5628
> Project: Beam
> Issue Type: Sub-task
> Components: sdk-py-core
> Reporter: Valentyn Tymofieiev
> Assignee: Simon
> Priority: Major
>
> ERROR: test_read_after_splitting (apache_beam.io.vcfio_test.VcfSourceTest)
> "
> ----------------------------------------------------------------------
> Traceback (most recent call last):
> File
> ""/usr/local/google/home/valentyn/projects/beam/clean_head/beam/sdks/python/apache_beam/io/vcfio_test.py"",
> line 336, in test_read_after_splitting
> ] split_records.extend(source_test_utils.read_from_source(*source_info))
> ] File
> ""/usr/local/google/home/valentyn/projects/beam/clean_head/beam/sdks/python/apache_beam/io/source_test_utils.py"",
> line 101, in read_from_source
> for value in reader:
> File
> ""/usr/local/google/home/valentyn/projects/beam/clean_head/beam/sdks/python/apache_beam/io/vcfio.py"",
> line 264, in read_records
> for line in record_iterator:
> File
> ""/usr/local/google/home/valentyn/projects/beam/clean_head/beam/sdks/python/apache_beam/io/vcfio.py"",
> line 330, in __next__
> record = next(self._vcf_reader)
> File
> ""/usr/local/google/home/valentyn/projects/beam/clean_head/beam/sdks/python/target/.tox/py3/lib/python3.5/site-packages/vcf/parser.py"",
> line 543, in __next__
> row = self._row_pattern.split(line.rstrip())
> TypeError: cannot use a string pattern on a bytes-like object
> "
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
