Carlos Vara (JIRA) wrote:
[ https://issues.apache.org/jira/browse/BVAL-29?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12862369#action_12862369 ]
Carlos Vara commented on BVAL-29:
---------------------------------
Thanks for reviewing Donald. I will add the header and include new file
additions in the patch in future fixes.
Also, I usually separate the patches per project as I have them checked out
individually. If you prefer them to be from the root of the svn just tell me.
Defer node creation in implementation of ConstraintViolationBuilder API
-----------------------------------------------------------------------
Key: BVAL-29
URL: https://issues.apache.org/jira/browse/BVAL-29
Project: BeanValidation
Issue Type: Bug
Components: jsr303
Affects Versions: 0.1-incubating
Reporter: Carlos Vara
Assignee: Donald Woods
Fix For: 0.1-incubating
Attachments: bval-jsr303.patch, ConstraintValidatorContextTest.java
Call to inIterable() should modify the previous sibling node instead of the
"current" node, and in case a null named non iterable node is left in the leaf
of the path, it should never be added.
This behavior is not documented in the JSR-303 spec document, but
ConstraintValidatorContext Javadoc gives information on how it should work.
Patch includes some tests that serve also as examples of this API usage.
1 more test passes.
Hi Carlos,
I am satisfied with all of your patches in either form. I am glad, that
you have the motivation, time and understanding to submit those
high-quality patches. I always review them anyway and I, for my part,
have no problems with adding missing stuff.
So, what I wanted so say: Thanks for your work! Do you think, that
most/all the TCK tests, that fail can be fixed with minor changes like
those, that you have fixes so far? I am afraid, that especially those
about XML meta data, would cause more work before they pass....
Regards, Roman
