[jira] [Commented] (ARROW-15779) [Python] Create python bindings for Substrait consumer

["Weston Pace (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/ARROW-15779?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17501471#comment-17501471 ] 

Weston Pace commented on ARROW-15779:
-------------------------------------

[~jorisvandenbossche][~amol-] Any preference on where we stick this method?  I think it will just be something like...

```
run_query(plan: bytes): RecordBatchReader
```
So do we want {{pyarrow.compute.run_query}}, {{pyarrow.dataset.run_query}}, or {{pyarrow.engine.run_query}}?

> [Python] Create python bindings for Substrait consumer
> ------------------------------------------------------
>
>                 Key: ARROW-15779
>                 URL: https://issues.apache.org/jira/browse/ARROW-15779
>             Project: Apache Arrow
>          Issue Type: Improvement
>          Components: Python
>            Reporter: Weston Pace
>            Priority: Major
>
> We will want to figure out how to expose the Substrait consumer to python.  This could be a single method that accepts a buffer of bytes and returns an iterator of record batches but we might also want a helper method that returns a table.  I'm thinking this would go in the compute namespace.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)