[Python] cannot create a chunked_array from dictionary_encoding result

[asfimport]

I've experienced a strange error raise when trying to apply pa.chunked_array directly on the indices of dictionary_encoding (code is below). Making a memory view solves the problem.

import pyarrow as pa
ca = pa.array(['a', 'a', 'b', 'b', 'c'])                                                                                           
fca = ca.dictionary_encode()                                                                                                       
fca.indices                                                                                                                        
<pyarrow.lib.Int32Array object at 0x1250fb888>
[
  0,
  0,
  1,
  1,
  2
]

pa.chunked_array([fca.indices])                                                                                                    
---------------------------------------------------------------------------
ArrowInvalid                              Traceback (most recent call last)
<ipython-input-44-71ca3b877e1c> in <module>
----> 1 pa.chunked_array([fca.indices])

~/Projects/miniconda3/envs/pyarrow/lib/python3.7/site-packages/pyarrow/table.pxi in pyarrow.lib.chunked_array()

~/Projects/miniconda3/envs/pyarrow/lib/python3.7/site-packages/pyarrow/error.pxi in pyarrow.lib.check_status()

ArrowInvalid: Unexpected dictionary values in array of type int32

# with another memory view it's  OK
pa.chunked_array([fca.indices.view(fca.indices.type)])                 
Out[45]: 
<pyarrow.lib.ChunkedArray object at 0x12508dc78>
[
  [
    0,
    0,
    1,
    1,
    2
  ]
]
 

Reporter: Artem KOZHEVNIKOV / @artemru
Assignee: Joris Van den Bossche / @jorisvandenbossche

PRs and other links:

• GitHub Pull Request #5656

_{Note: This issue was originally created as ARROW-6882. Please see the migration documentation for further details.}

[asfimport]

Joris Van den Bossche / @jorisvandenbossche:
Thanks for the report. Labeling as 0.15.1 for now as it seems a regression compared to 0.14

[asfimport]

Joris Van den Bossche / @jorisvandenbossche:
Although, it is only a regression because we now validate the resulting array automatically. On 0.14.1 manually validating the resulting array gives the same error.

And you get the same when actually validating the indices array:

In [23]: fca.indices.validate()   
...
ArrowInvalid: Unexpected dictionary values in array of type int32

[asfimport]

Antoine Pitrou / @pitrou:
Issue resolved by pull request 5656
#5656