[ 
https://issues.apache.org/jira/browse/NIFI-1123?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15009117#comment-15009117
 ] 

ASF GitHub Bot commented on NIFI-1123:
--------------------------------------

Github user apiri commented on the pull request:

    https://github.com/apache/nifi/pull/116#issuecomment-157450706
  
    Was combing through the actual PR.  Not sure I am onboard with the 
additional property and misunderstood what the intent was.  There was a similar 
effort with the Character Set Validator where we validated that the specified 
charset was valid if specified, otherwise just that a string was returned via 
EL should it be an expression.
    
    Regex Validator, by default evaluates the EL and then applies validation to 
the resulting "regex".  Would it be reasonable to have a separate validator 
that matches the semantics of what you are looking for (akin to that of how 
ConvertCharacterSet functions)?  Should be okay from the standpoint of 
compatibility and non-breaking changes.
    
    Additionally, playing with this a bit, I did a blanket delete (.*) which 
leads to some bad effects in terms of the framework as the UUID is removed and 
puts us in a bad state.


> Extend the "Delete Attributes Expression" to support Expression Language
> ------------------------------------------------------------------------
>
>                 Key: NIFI-1123
>                 URL: https://issues.apache.org/jira/browse/NIFI-1123
>             Project: Apache NiFi
>          Issue Type: Improvement
>          Components: Core Framework
>            Reporter: Joe Skora
>            Assignee: Joe Skora
>            Priority: Minor
>              Labels: easyfix, features, patch
>
> Allow the "Delete Attributes Expression" to accept Expression Language to 
> dynamically produce the regular expression to identify attributes to be 
> deleted per discussion on 
> [NIFI-641|https://issues.apache.org/jira/browse/NIFI-641].



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

Reply via email to