[ 
https://issues.apache.org/jira/browse/TAP5-2513?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Chris Poulsen updated TAP5-2513:
--------------------------------
    Attachment: 
TAP5_2513_Specifying_a_regexp_validation_rule_where_the_pattern_contains_a_comma_breaks1.patch

Allow validate specification to be null or empty string (again). 

> Specifying a regexp validation rule where the pattern contains a comma breaks
> -----------------------------------------------------------------------------
>
>                 Key: TAP5-2513
>                 URL: https://issues.apache.org/jira/browse/TAP5-2513
>             Project: Tapestry 5
>          Issue Type: Bug
>    Affects Versions: 5.4
>            Reporter: Chris Poulsen
>              Labels: patch
>         Attachments: 
> TAP5_2513_Specifying_a_regexp_validation_rule_where_the_pattern_contains_a_comma_breaks1.patch
>
>
> I tried to do the following: 
> {code}
>     @Component( parameters = { "value=value", 
> "validate=regexp=[0-9a-fA-F]{3,6}" } )
>     private TextField myField;
> {code}
> Tapestry complains about the regexp pattern not being valid (in particular 
> that the "6" is unexpected. This is due to the parser in 
> (FieldValidatorSourceImpl) expecting that comma only exist to separate 
> validator rules.
> Similar issues seem to have already been fixed for the other ways to express 
> regexp constraints (TAP5-520).
> It seems that there are several different ways implemented to extract the 
> validation rules currently, I tried replacing the code in 
> FieldValidatorSourceImpl with the code used in 
> (ValidateAnnotationConstraintGenerator / FieldValidatorDefaultSourceImpl )  - 
> It seems to produce the correct results at runtime, but will probably require 
> some unit test adjustments, as the char-by-char parser have been replaced (it 
> looks like a decent error message is still produced when an invalid pattern 
> is specified, but it is another place that returns the error).



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

Reply via email to