[
https://issues.apache.org/jira/browse/NIFI-8146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17268123#comment-17268123
]
ASF subversion and git services commented on NIFI-8146:
-------------------------------------------------------
Commit f02079c2faa008c9d2a5fc0cebafcb398c22bf0c in nifi's branch
refs/heads/main from Mark Payne
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=f02079c ]
NIFI-8146: Added RecordPath for Statement Type and to specify where to find the
data within the incoming Record. Also performed some minor code refactoring to
avoid lots of large lambdas, as it results in code that is more readable and
more maintainable. Updated some properties to use dependsOn() notation now that
this is available.
NIFI-8146: Updated PutDatabaseRecord to avoid using the functional style
framework that had previously been used. Doing so resulted in code that was
difficult to understand and maintain. Added additional unit tests and improved
MockRecordParser so that it could throw configurable types of exceptions
NIFI-8146: Fixed checkstyle issues
Signed-off-by: Matthew Burgess <[email protected]>
This closes #4763
> Allow RecordPath to be used for specifying operation type and data fields
> when using PutDatabaseRecord
> ------------------------------------------------------------------------------------------------------
>
> Key: NIFI-8146
> URL: https://issues.apache.org/jira/browse/NIFI-8146
> Project: Apache NiFi
> Issue Type: New Feature
> Components: Extensions
> Reporter: Mark Payne
> Assignee: Mark Payne
> Priority: Major
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> PutDatbaseRecord requires that the Statement Type be defined as a property or
> a FlowFile attribute. This means that if a FlowFile has many records, it must
> be split apart into individual Records if there is more than 1 type of
> statement needed per FlowFile.
> It also assumes that the data to be inserted/updated/deleted/etc is the full
> record. However, it's common to have some wrapper around the actual data, as
> is the case with a tool like Debezium, which includes an Operation Type, a
> 'before' snapshot and an 'after' snapshot. To accommodate this, we should
> allow Record-friendly methods for specifying the path to the data and the
> operation type.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
