[
https://issues.apache.org/jira/browse/HBASE-14375?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15894543#comment-15894543
]
Sean Busbey commented on HBASE-14375:
-------------------------------------
bq. Do you think we can just mark the DefaultSource private as it is?
Yeah I think that's fine. It should get across that we don't want folks
building on top of it. Alternatively you could mark it LimitedPrivate<CONFIG>
with a note that says it's meant to be the implementation for
'org.apache.hadoop.hbase.spark' and not downstream modifiable. I'd consider
this equivalent to using the DataSourceRegister trait and a name in
HBaseSparkConf. (to be clear, like just marking it Private or
LimitedPrivate<CONFIG> better than altering HBaseSparkConf.)
If you prefer modifying HBaseSparkConf or using LimitedPrivate<SPARK> I'd be
fine with that too.
bq. Regarding the API docs, is it our goal that we convert the scala-doc to
java-doc and show them together with the rest of HBase Java API docs?
That's the goal, yes.
bq. Hopefully the annotations can be effectively filtered as they are now.
Otherwise we have to find other ways to do it. For example, going into the
scala classes to update the class/function modifiers? Hopefully that can be
separated into another task.
I agree that it should be a different JIRA. I was just idly commenting about it
since both things were on my mind.
> define public API for spark integration module
> ----------------------------------------------
>
> Key: HBASE-14375
> URL: https://issues.apache.org/jira/browse/HBASE-14375
> Project: HBase
> Issue Type: Task
> Components: spark
> Reporter: Sean Busbey
> Assignee: Jerry He
> Priority: Blocker
> Fix For: 2.0.0
>
> Attachments: HBASE-14375-v1.patch
>
>
> before we can put the spark integration module into a release, we need to
> annotate its public api surface.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
