[ARROW-12431] [Python] pa.array mask inverted when type is binary and value to be converted is numpy array

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 5.0.0
• Component/s: Python
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/28220
• Assignee: Alessandro Molina
• Reporter: Daniel Nugent
• Created: 17/Apr/21 07:33
• Updated: 11/Jan/23 08:26
• Resolved: 15/Jun/21 09:57

Description

Python 3.9.2 | packaged by conda-forge | (default, Feb 21 2021, 05:02:46)                                   
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> import pyarrow as pa
>>>
>>> pa.array(np.array([b'\x00']),type=pa.binary(1), mask = np.array([False]))
<pyarrow.lib.FixedSizeBinaryArray object at 0x7fa080ca3640>
[
  null
]
>>> pa.array(np.array([b'\x00']),type=pa.binary(1), mask = np.array([True]))
<pyarrow.lib.FixedSizeBinaryArray object at 0x7fa080ca3700>
[
  00
]
>>> pa.array([b'\x00'],type=pa.binary(1), mask = np.array([False]))
<pyarrow.lib.FixedSizeBinaryArray object at 0x7fa083cc9520>
[
  00
]
>>> pa.__version__
'3.0.0'
>>> np.__version__
'1.20.1'

Happens both with FixedSizeBinary and variable sized binary (I was working with FixedSizeBinary). Does not happen for integers (presumably other types, didn't exhaustively check)?

Issue Links

• links to: GitHub Pull Request #10199