On Wed, Nov 26, 2008 at 9:14 AM, <[EMAIL PROTECTED]>wrote:
> Hello,
>
>
>
> Today we noticed that the t:checkbox component used inside a
> t:selectManyCheckbox generates checkboxes all with the same ID. This is a
> probably since we use h:outputLabel with the "for" attribute, mainly for
> accessibility purposes, but the ids don't match. We have something like this
> in the page.
>
>
>
> <t:selectManyCheckbox
>
> id=*"articlesChecked"*
>
> value="#{mbArticles.articlesChecked}"
>
> layout=*"spread"*
>
> onclick=*"checkState(this,false,'articlesList')"*>
>
>
>
>
>
> Then a bit further down we have this in a dataList
>
>
>
>
>
> <h:outputLabel for=*"articleCheckbox"*>
>
> <t:checkbox index="#{count}" for=*":articlesForm:articlesChecked"*
> id=*"articleCheckbox"*/>
>
>
>
> So the checkboxes generated are like
>
>
>
> <input type="checkbox" name="articlesChecked" id="articlesChecked" />
>
>
>
> I know the HTML spec pretty well and for this case only the "name"
> attribute really needs to be repeated, the id can be unique and should be,
> it's actually invalid to have duplicate ids in the DOM (then again so is an
> id starting with the "_" character. I'll never understand that decision).
>
>
>
> I extended the HtmlCheckboxRenderer and basically changed it such that it
> will output the name attribute as the clientId of the UISelectMany component
> and the ID as the clientId of the HtmlCheckBox component. I override
> encodeEnd and kept it identical except I have my own renderSingleCheckBox
> method, which is also identical except it calls my version of
> renderCheckBox, which is nearly identical except for this part:
>
>
>
>
>
> *if* (renderId) {
>
> *if* (StringUtils.*isNotEmpty*(checkboxClientId)) {
>
> writer.writeAttribute(HTML.*ID_ATTR*, checkboxClientId, *
> null*);
>
> }
>
> }
>
>
>
> checkboxClientId is an extra parameter passed in.
>
>
>
> So far this solution seems to work. The ids are correctly generated and the
> functionality hasn't broken (meaning I can select a few checkboxes and
> they're picked up on the other side).
>
>
>
> So my questions here are:
>
>
>
> Is there any possible side affect that I haven't though about? I probably
> haven't tested every possible scenario of use here.
>
>
>
> Has this been addressed already in a later version of tomahawk? I'm using
> 1.1.6 here. If not then it should be updated in a future release of
> Tomahawk.
>
>
>
Hi
Yes, see https://issues.apache.org/jira/browse/TOMAHAWK-1188 for details. I
hope this helps.
regards
Leonardo Uribe
> Matt
>
>
>
>
>
