[jira] [Commented] (NIFI-8754) CSVReader with InferSchema ignoring 1st non header row, even with necessary properties set

["Matt Burgess (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/NIFI-8754?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17380678#comment-17380678 ] 

Matt Burgess commented on NIFI-8754:
------------------------------------

Not sure when/if this behavior changed, but IMO this is the expected behavior. With Infer Schema for CSV, it needs the header line to determine the field names, so no matter what Treat First Line as Header is set to, Infer Schema will treat the first line as the header. When using an External Schema, the Treat First Line as Header is honored since you can supply the field names via the schema.

> CSVReader with InferSchema ignoring 1st non header row, even with necessary properties set
> ------------------------------------------------------------------------------------------
>
>                 Key: NIFI-8754
>                 URL: https://issues.apache.org/jira/browse/NIFI-8754
>             Project: Apache NiFi
>          Issue Type: Bug
>    Affects Versions: 1.13.2
>            Reporter: K Sanghavi
>            Priority: Major
>              Labels: bug
>         Attachments: CSVReaderBug.xml, csvreader-config.jpg, csvreader-flow.jpg, csvreader-input.jpg, csvreader-output.jpg
>
>
> When using CSVReader controller service in any Record based processors  of 1.13.2, is not working as expected in InferSchema mode, where it ignores the 1st non header row even with *Treat First Line as Header* property set to *false*. But with External schema reference the same controller service is working as expected. (Attached images for SplitRecord with CSVReader)
> This setup was tested in 1.11.4, and with same configuration for CSVReader with InferSchema was giving expected results.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)