Raúl Cumplido created ARROW-17725:
-------------------------------------
Summary: [CI][Python] Python minimal nightly builds fail importing
pyarrow.parquet
Key: ARROW-17725
URL: https://issues.apache.org/jira/browse/ARROW-17725
Project: Apache Arrow
Issue Type: Bug
Components: Continuous Integration, Python
Reporter: Raúl Cumplido
Fix For: 10.0.0
Both example-python-minimal builds failed on the last nightly builds:
*
[example-python-minimal-build-fedora-conda|https://github.com/ursacomputing/crossbow/actions/runs/3049355177/jobs/4915336459]
*
[example-python-minimal-build-ubuntu-venv|https://github.com/ursacomputing/crossbow/actions/runs/3049345474/jobs/4915318371]
The error happens when collecting tests because it failed to import
pyarrow.parquet as pq:
{code:java}
============================= test session starts ==============================
platform linux -- Python 3.10.6, pytest-7.1.3, pluggy-1.0.0
rootdir: /arrow/python, configfile: setup.cfg
plugins: hypothesis-6.47.1, lazy-fixture-0.6.3
collected 4213 items / 2 errors / 3 skipped====================================
ERRORS ====================================
_________________ ERROR collecting pyarrow/tests/test_hdfs.py __________________
pyarrow/tests/test_hdfs.py:32: in <module>
from pyarrow.tests.parquet.test_dataset import (
<frozen importlib._bootstrap>:1027: in _find_and_load
???
<frozen importlib._bootstrap>:1006: in _find_and_load_unlocked
???
<frozen importlib._bootstrap>:688: in _load_unlocked
???
/miniconda-for-arrow/envs/pyarrow-3.10/lib/python3.10/site-packages/_pytest/assertion/rewrite.py:168:
in exec_module
exec(co, module.__dict__)
pyarrow/tests/parquet/test_dataset.py:566: in <module>
@pytest.mark.parametrize("read", (pq.read_table, pq.read_pandas))
E AttributeError: 'NoneType' object has no attribute 'read_table'
____________ ERROR collecting pyarrow/tests/parquet/test_dataset.py ____________
pyarrow/tests/parquet/test_dataset.py:566: in <module>
@pytest.mark.parametrize("read", (pq.read_table, pq.read_pandas))
E AttributeError: 'NoneType' object has no attribute 'read_table'
=========================== short test summary info ============================
ERROR pyarrow/tests/test_hdfs.py - AttributeError: 'NoneType' object has no a...
ERROR pyarrow/tests/parquet/test_dataset.py - AttributeError: 'NoneType' obje...
!!!!!!!!!!!!!!!!!!! Interrupted: 2 errors during collection !!!!!!!!!!!!!!!!!!!!
========================= 3 skipped, 2 errors in 5.82s
========================= {code}
The changes on the repo since the last success can be seen here:
[https://github.com/apache/arrow/compare/4ae26d1bb7537ca19d3b925cf4e7bf5f6a1c0d07...f57e1bad92d536cf8188e6d5bcc414cef7a048e8]
I've been able to reproduce locally with:
{code:java}
$ cd arrow/python/examples/minimal_build
$ docker-compose build minimal-fedora-conda
$ docker-compose run --rm minimal-fedora-conda{code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
