[ 
https://issues.apache.org/jira/browse/BEAM-10009?focusedWorklogId=472970&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-472970
 ]

ASF GitHub Bot logged work on BEAM-10009:
-----------------------------------------

                Author: ASF GitHub Bot
            Created on: 20/Aug/20 16:37
            Start Date: 20/Aug/20 16:37
    Worklog Time Spent: 10m 
      Work Description: TheNeuralBit commented on pull request #12573:
URL: https://github.com/apache/beam/pull/12573#issuecomment-677773121


   @robertwb, @robinyqiu - In this PR I've mapped python's Timestamp 
(microsecond precision) to the millis_instant logical type lossily so it could 
be mapped to DATETIME on the Java side and used in SQL.
   
   Based on discussion in https://github.com/apache/beam/pull/12493, now I'm 
thinking the better answer would be to add a micros_instant logical type so 
Python's data can go across the xlang boundary faithfully, and it's up to SQL 
to throw an error when the timestamp precision is too high, or silently drop 
precision if the user chooses.
   
   There's an open question about what to do if the output of the SqlTransform 
includes a millis_instant logical type. Would that still get mapped to a python 
Timestamp?


----------------------------------------------------------------
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]


Issue Time Tracking
-------------------

    Worklog Id:     (was: 472970)
    Time Spent: 0.5h  (was: 20m)

> Support for date times in Python schemas
> ----------------------------------------
>
>                 Key: BEAM-10009
>                 URL: https://issues.apache.org/jira/browse/BEAM-10009
>             Project: Beam
>          Issue Type: Improvement
>          Components: sdk-py-core
>            Reporter: Brian Hulette
>            Priority: P3
>          Time Spent: 0.5h
>  Remaining Estimate: 0h
>
> When DATETIME is converted to a MillisInstant logical type, add support for 
> it to RowCoder and python schemas.
> Probably we should map MillisInstant to datetime with  
> [datetime.datetime.from_timestamp|https://docs.python.org/3/library/datetime.html#datetime.datetime.fromtimestamp]



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to