rok commented on pull request #10598:
URL: https://github.com/apache/arrow/pull/10598#issuecomment-868449124


   > Pandas actually has a different method (the one you link is an attribute 
that has been there for a long time, so but pandas added an additional method 
so it could have arguments) that takes a locale arguments: 
https://pandas.pydata.org/docs/reference/api/pandas.Series.dt.day_name.html
   
   Oh, nice to know!
    
   > Personally, I would (at least for now) leave all locale-specific handling 
to the bindings / downstream applications (except for `strftime`). As Nic says, 
if we only return integers (and not `labels`) for now, we shouldn't need locale 
handling (I wouldn't let the number depend on the locale, that seems very easy 
to miss).
   
   Yeah, leaving locale out of this does seem like the sane option for now.
    
   > About the actual options in this PR: that seems like a good start, but I 
don't think all those options would be needed in a single Options struct? For 
example, the ambiguous/non-existent handling, those can be put in a specific 
options struct for the kernel that needs it?
   
   I'm ok with single Options struct or multiple.
   If we go for multiple we'd probably split this into 
`TemporalStrftimeOptions` 
([ARROW-13174](https://issues.apache.org/jira/browse/ARROW-13174)), 
`TemporalComponentExtractionOptions`, `TemporalLocalizationOptions` 
([ARROW-13033](https://issues.apache.org/jira/browse/ARROW-13033)). Am I 
missing something?


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to