ImportError: cannot import name _orc

[linxabc]

Meeting error import orc in RedHat 7, could you give some insights on how to solve it. thanks!

Traceback (most recent call last):
File "train.py", line 14, in
from pyarrow import orc
File "~/virtualenv-2.7/lib/python2.7/site-packages/pyarrow/orc.py", line 21, in
from pyarrow import _orc
ImportError: cannot import name _orc

[xhochy]

How have you installed pyarrow?

[linxabc]

installed via pip.
there's no error if just import pyarrow

[wesm]

It doesn't look like the ORC extension is being built in the manylinux1 wheels. I opened a JIRA, let's track the issue there https://issues.apache.org/jira/browse/ARROW-2672

[wesm]

@ishare you can obtain this functionality in the meantime by installing with conda

[leferrad]

Why was this issue closed? From version 0.15, I still have the same error, so I cannot use newer versions of pyarrow until ORC is available for usage.
Here the very simple snippet to reproduce error (tested with pyarrow==0.16.0):

from pyarrow import orc 
# Throws:
#     23 from pyarrow import types
#     24 from pyarrow.lib import Schema
# ---> 25 import pyarrow._orc as _orc
#      26 
#      27 
# 
# ModuleNotFoundError: No module named 'pyarrow._orc'

[xhochy]

The issue is tracked in JIRA, please comment there (obviously the issue is clear, all we need is someone to implement it).

[wesm]

If anyone comes across this while searching the internet, the latest JIRA issue about having pyarrow.orc in the wheels is https://issues.apache.org/jira/browse/ARROW-7811. We had to disable it a while ago due to a packaging issue and no one has volunteered to fix it yet