[
https://issues.apache.org/jira/browse/BVAL-66?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12873047#action_12873047
]
Roman Stumm commented on BVAL-66:
---------------------------------
I tried to merge the changes into the current code base, but I had to do some
changes, because of other changes in the affected methods.
All tests are still passing, but I would feel better, if some additional new
tests for the patch would have also been provided..
> Method validation parameter processing problems
> -----------------------------------------------
>
> Key: BVAL-66
> URL: https://issues.apache.org/jira/browse/BVAL-66
> Project: BeanValidation
> Issue Type: Improvement
> Affects Versions: 0.1-incubating, 0.2-incubating
> Reporter: Roman Stumm
> Assignee: Roman Stumm
> Fix For: 0.2-incubating
>
>
> Although still experimental and although only a few tests exist for
> Method-level-validation, some people are already using it and are reporting
> problems or providing patches.
> For the old code-base (agimatec-validation, no further development going on)
> I got a patch that fixes some misbehavior in the
> MethodValidatorMetaBeanFactory.
> To provide a better way for the agimatec-validation users to migrate to
> apache-bval, I suggest to merge the changes into the current code base.
> see original post at:
> http://code.google.com/p/agimatec-validation/issues/detail?id=26
> 1.
> when processAnnotation is called in MethodValidatorMetaBeanFactory
> is called the type of the class owning the method is used to process
> parameter constraints rather than the type of the parameter. This causes
> something like a @Min annotation on a parameter to generate an exception
> because no appriopriate validator class can be found for the class
> declaring the method.
> 2.
> the @Valid annotation causes 'setCascaded' to be called on the
> MethodDescriptor but not on the ParameterDescriptor where it needs to be
> set. Because of this validation doesn't cascade to the parameter object.
> 3.
> the @Valid is treated as being mutually exclusive from other
> constraint annotations on a parameter. It is common to use @Valid with
> @NotNull since @Valid doesn't make assumptions about the nullability of a
> value.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
