[
https://issues.apache.org/jira/browse/NIFI-305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14301882#comment-14301882
]
Mark Payne commented on NIFI-305:
---------------------------------
Joe: I agree it's a bit of a pain to have the implementing class add
boilerplate code to provide them, but I think if we allowed some ability to
order them it would be more confusing. We could revisit this later, but this is
the approach that we've generally taken with other abstract processors, so I'd
be inclined to just keep with it at this point personally.
> Refactor base class from MergeContent
> -------------------------------------
>
> Key: NIFI-305
> URL: https://issues.apache.org/jira/browse/NIFI-305
> Project: Apache NiFi
> Issue Type: Improvement
> Components: Core Framework
> Affects Versions: 0.0.1
> Reporter: Joseph Gresock
> Priority: Minor
> Attachments:
> 0001-NIFI-305-Slight-refactorings-to-provide-more-flexibi.patch
>
>
> The binning logic in MergeContent is extremely useful, and could be pulled
> out into an abstract superclass. This would allow other processors to
> perform binning logic without being tied to a merged flow file. For example,
> a processor may want to submit a batch request to a database like Solr, or to
> a REST endpoint.
> The vast majority of the code in MergeContent would remain in the concrete
> class, but there are several points, such as the BinManager usage, that could
> be abstracted for easy extensibility.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
