[
https://issues.apache.org/jira/browse/NIFI-6680?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16933369#comment-16933369
]
Joseph Witt commented on NIFI-6680:
-----------------------------------
At least as shown in this example use case the need for mutation of the column
names is an issue of a given system we're interfacing with. So to that end I'd
lean toward having this mutation only occur at that point. If we were to find
this was a common case or something we'd want to do/have understood in the
record interfaces themselves then I'd think what you're showing is a good next
step.
> Nifi PutKudu processor - Convert record field names to lowercase
> ----------------------------------------------------------------
>
> Key: NIFI-6680
> URL: https://issues.apache.org/jira/browse/NIFI-6680
> Project: Apache NiFi
> Issue Type: Wish
> Components: Extensions
> Affects Versions: 1.9.2
> Reporter: Kevin J McCarthy
> Priority: Major
> Original Estimate: 1h
> Time Spent: 0.5h
> Remaining Estimate: 0.5h
>
> Kudu supports case sensitive column names while query tools built over Kudu
> (eg. Impala) do not. If a table is created via Impala, which is the most
> common use case, all column names are converted to lowercase. If an incoming
> record in a FlowFile contains uppercase column names, the columns won't match
> the schema of the destination table unless the schema for the record is
> modified to have lowercase field names.
> Rather than converting the schema of the record, I propose field names are
> converted to lowercase before they are put into Kudu.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
