[jira] [Updated] (STORM-2295) KafkaSpoutStreamsNamedTopics changing the sequence of fields name while emitting data

["Sachin Pasalkar (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/STORM-2295?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sachin Pasalkar updated STORM-2295:
-----------------------------------
    Affects Version/s:     (was: 1.0.0)
                       1.x

> KafkaSpoutStreamsNamedTopics changing the sequence of fields name while emitting data
> -------------------------------------------------------------------------------------
>
>                 Key: STORM-2295
>                 URL: https://issues.apache.org/jira/browse/STORM-2295
>             Project: Apache Storm
>          Issue Type: Bug
>          Components: storm-kafka-client
>    Affects Versions: 1.x
>            Reporter: Sachin Pasalkar
>            Priority: Blocker
>
> If you look at below code *allFields* variable is HashSet. To which we have added the o/p of *kafkaSpoutStream.getOutputFields().toList()*. That sort data on hash basis rather than keeping same sequence.
> {code:java}
> 	@Override
> 	public Fields getOutputFields() {
> 		final Set<String> allFields = new HashSet<>();
> 		for (KafkaSpoutStream kafkaSpoutStream : topicToStream.values()) {
> 			allFields.addAll(kafkaSpoutStream.getOutputFields().toList());
> 		}
> 		return new Fields(new ArrayList<>(allFields));
> 	}
> {code}
> Changes needed is below
> {code:java}
> final Set<String> allFields = new LinkedHashSet<>();
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)