[
https://issues.apache.org/jira/browse/MYFACES-4764?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18108865#comment-18108865
]
Volodymyr Siedlecki commented on MYFACES-4764:
----------------------------------------------
We'll need to fix the main branch, too.
I believe this is a bug as it is a regression, unless I'm wrong? I tested
Mojarra 4.1, and it also rendered form1.
> @child(N) fails in 4.1.x snapshot
> ---------------------------------
>
> Key: MYFACES-4764
> URL: https://issues.apache.org/jira/browse/MYFACES-4764
> Project: MyFaces Core
> Issue Type: Bug
> Reporter: Volodymyr Siedlecki
> Priority: Major
>
> The @child(N) search keyword resolver in ChildSearchKeywordResolver contains
> a bug introduced in MyFaces 4.1 that causes @child(N) to resolve to the wrong
> component when any UntargetableComponent children precede the target in the
> component tree.
> In ChildSearchKeywordResolver.resolve(), the loop correctly counts only
> non-UntargetableComponent children to find the logical @child(N). However,
> when invoking the callback it uses children.get(childNumber) — the raw list
> index — instead of children.get(i) — the actual position of the matching
> component.
> Working:
> <label for="form1">Body Child(1)</label>
> Failing:
> <label for="j_id_a">Body Child(1)</label>
> Change where this problem was introduced:
> https://github.com/apache/myfaces/commit/01566fc67f7d672892ac7324b66e8fb7a927c8fb#diff-d1db49a29a4071345c052b93fc70743eb79f29b27bb84221089a5e62dd1d369fL18-L69
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
