[
https://issues.apache.org/jira/browse/BEAM-562?focusedWorklogId=230313&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-230313
]
ASF GitHub Bot logged work on BEAM-562:
---------------------------------------
Author: ASF GitHub Bot
Created on: 19/Apr/19 23:17
Start Date: 19/Apr/19 23:17
Worklog Time Spent: 10m
Work Description: aaltay commented on issue #7994: [BEAM-562] Add
DoFn.setup and DoFn.teardown to Python SDK
URL: https://github.com/apache/beam/pull/7994#issuecomment-485035810
> That line does not show a cloning of appliedptransform...
> […](#)
> On Fri, Apr 19, 2019 at 3:44 PM Ahmet Altay ***@***.***> wrote: I cannot
reply to this comment in place If self._perform_dofn_pickle_test is false, it
looks like it is not cloned but comes directly from
self._appliedptransform.transform.dofn. Is the whole appliedptransform clone
per bundle? applied transform will be a new instance for each bundle (for
direct runner). (See:
https://github.com/apache/beam/blob/master/sdks/python/apache_beam/runners/direct/executor.py#L366
) — You are receiving this because you were mentioned. Reply to this email
directly, view it on GitHub <[#7994
(comment)](https://github.com/apache/beam/pull/7994#issuecomment-485031290)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAEMO6ONQUU57D4336BHZS3PRJDNRANCNFSM4G36BRSA>
.
You are right, I confused by evaluator. perform_dofn_pickle_test defaults
to True, and there is only one path that is setting to false related to sdf
evaluation.
I think we should clone by default, probably by getting rid of that flag. I
am not sure why we had the flag in the first place, but I am guessing there
will be an underlying issue.
----------------------------------------------------------------
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: 230313)
Time Spent: 6h 10m (was: 6h)
> DoFn Reuse: Add new methods to DoFn
> -----------------------------------
>
> Key: BEAM-562
> URL: https://issues.apache.org/jira/browse/BEAM-562
> Project: Beam
> Issue Type: New Feature
> Components: sdk-py-core
> Reporter: Ahmet Altay
> Priority: Major
> Labels: sdk-consistency, triaged
> Time Spent: 6h 10m
> Remaining Estimate: 0h
>
> Java SDK added setup and teardown methods to the DoFns. This makes DoFns
> reusable and provide performance improvements. Python SDK should add support
> for these new DoFn methods:
> Proposal doc:
> https://docs.google.com/document/d/1LLQqggSePURt3XavKBGV7SZJYQ4NW8yCu63lBchzMRk/edit?ts=5771458f#
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
