[
https://issues.apache.org/jira/browse/BEAM-14294?focusedWorklogId=765594&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-765594
]
ASF GitHub Bot logged work on BEAM-14294:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 03/May/22 18:04
Start Date: 03/May/22 18:04
Worklog Time Spent: 10m
Work Description: TheNeuralBit commented on code in PR #17384:
URL: https://github.com/apache/beam/pull/17384#discussion_r864047641
##########
sdks/python/apache_beam/utils/windowed_value.pxd:
##########
@@ -43,6 +43,23 @@ cdef class WindowedValue(object):
cpdef WindowedValue with_value(self, new_value)
+cdef class WindowedBatch(object):
+ cpdef WindowedBatch with_values(self, object new_values)
+
+cdef class HomogeneousWindowedBatch(WindowedBatch):
+ cdef public WindowedValue _wv
+
+ cpdef WindowedBatch with_values(self, object new_values)
+
+cdef class ConcreteWindowedBatch(WindowedBatch):
+ cdef public object values
Review Comment:
yeah the WindowedValue(None) approach could make sense. The downside is it's
not "columnar", but for that matter neither is the current implementation based
on python lists. An optimized version could be yet another future WindowedBatch
implementation.
Issue Time Tracking
-------------------
Worklog Id: (was: 765594)
Time Spent: 5h (was: 4h 50m)
> MVP for SDK worker changes to support process_batch
> ---------------------------------------------------
>
> Key: BEAM-14294
> URL: https://issues.apache.org/jira/browse/BEAM-14294
> Project: Beam
> Issue Type: Sub-task
> Components: sdk-py-core
> Reporter: Brian Hulette
> Assignee: Brian Hulette
> Priority: P2
> Time Spent: 5h
> Remaining Estimate: 0h
>
> The initial MVP may only work in some restricted circumstances (e.g.
> @yields_element on process_batch, or batch-to-batch without a 1:1
> input:output mapping might not be supported). These cases should fail early.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
