[
https://issues.apache.org/jira/browse/FLINK-17445?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17095688#comment-17095688
]
Seth Wiesman commented on FLINK-17445:
--------------------------------------
Hi Brandon,
Very exited to see interest in this feature. Scala support is certainly second
class in the state processor api today.
I would not want to add a dependency on flink-scala to the state proc package,
but create a new module `flink-state-processor-scala`, similar to `cep-scala`
that is a wrapper exposing the correct types and passing type information
around as implicits. If you are interested in working on this I would be happy
to review the PR.
Please feel free to ping me here if you have any questions
> Allow OperatorTransformation to bootstrapWith a Scala DataSet
> -------------------------------------------------------------
>
> Key: FLINK-17445
> URL: https://issues.apache.org/jira/browse/FLINK-17445
> Project: Flink
> Issue Type: Improvement
> Components: API / State Processor
> Affects Versions: 1.10.0, 1.11.0
> Reporter: Brandon Bevans
> Priority: Major
>
> Currently there is no way to use OperatorTransformation.bootstrapWith using a
> Scala DataSet. The functionality does seem to be there since a Scala DataSet
> appears to just be a wrapped Java DataSet. This functionality could be
> achieved with minimal invasiveness by passing through the underlying Java
> DataSet to the call.
>
> Also, the docs suggest that this function should work:
> From
> [https://ci.apache.org/projects/flink/flink-docs-stable/dev/libs/state_processor_api.html]:
> {code:java}
> val currencyDataSet = bEnv.fromCollection(CurrencyRate("USD", 1.0),
> CurrencyRate("EUR", 1.3)) val broadcastTransformation =
> OperatorTransformation .bootstrapWith(currencyDataSet) .transform(new
> CurrencyBootstrapFunction){code}
> I would love to take this issue on.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
