[ARROW-8567] [Python] pa.array() sometimes ignore "safe=False"

• Type: Bug
• Status: In Progress
• Priority: Minor
• Resolution: Unresolved
• Affects Version/s: 0.17.0
• Component/s: Python
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/24733
• Assignee: Unassigned
• Reporter: Antoine Pitrou
• Created: 23/Apr/20 13:06
• Updated: 11/Jan/23 08:01

Description

Generally, pa.array(data).cast(sometype, safe=...) is equivalent to
pa.array(data, sometype, safe=...). Consider the following:

>>> pa.array([Decimal('12.34')]).cast(pa.int32(), safe=False)                                                                                                               
<pyarrow.lib.Int32Array object at 0x7efbbd1556e0>
[
  12
]
>>> pa.array([Decimal('12.34')], pa.int32(), safe=False)                                                                                                                    
<pyarrow.lib.Int32Array object at 0x7efbbd13dad0>
[
  12
]

However, that is not always the case:

>>> pa.array([Decimal('1234')]).cast(pa.int8(), safe=False)                                                                                                                 
<pyarrow.lib.Int8Array object at 0x7efbbd18e4b0>
[
  -46
]
>>> pa.array([Decimal('1234')], pa.int8(), safe=False)                                                                                                                      
Traceback (most recent call last):
  ...
ArrowInvalid: Value 1234 too large to fit in C integer type

I don't think this is very important: first because you can call cast() directly, second because the results are unusable anyway.

Issue Links

• relates to: ARROW-16491 [Python] Table.from_pandas is doing unsafe cast for float array to int array
• links to: GitHub Pull Request #13936