I have to see the code first i think.
Because if the button is not in the hierachy of the transparant container
you could argue that the button needs to be visible. Because what is
invisible?"
johan
On 10/24/07, Gerolf Seitz <[EMAIL PROTECTED]> wrote:
>
> On 10/24/07, Johan Compagner <[EMAIL PROTECTED]> wrote:
> >
> > is there a test case that fails? (the result should be not rendered but
> it
> > does render a bit?)
>
>
> no, unfortunately, there's not (yet).
>
> then i can look into it a bit more, i haven't seen much of the transparent
> > resolver code.
> >
> But are you saying that somebody is setting the transparent resolver to
> none
> > visible?
>
>
> exactly.
>
> And what component is then inserted to be rerendered for the ajax request?
>
>
> the panel, which is the parent of the (invisible) transparent resolver,
> which contains the markup for the (visible) button
>
> Why does isVisibleInHierarchy not work? The transparent resolver is still
> a
> > parent right?
>
>
> nope, the transparent resolver is not in the hierarchy of the visible
> button, which is probably the root problem of it all ;)
>
> johan
>
>
> Gerolf
>
>
> On 10/24/07, Gerolf Seitz <[EMAIL PROTECTED]> wrote:
> > >
> > > the transparent resolver is set to non visible in the onClick handler
> of
> > > an
> > > AjaxButton.
> > > ParentResolver is not involved at all.
> > >
> > > the problem is, that the transparent resolver is not found in the
> > > hierarchy
> > > of the button,
> > > and thus it's not possible to use Component#isVisibleInHierarchy in
> > > Component.java:2245
> > >
> > > changing the code you mentioned (rather in MarkupContainer than in
> > > ParentResolver) doesn't help at all,
> > > because this code is in renderComponentTagBody, which is called in
> > > MarkupContainer#onComponentTagBody.
> > > BUT, MarkupContainer#onComponentTagBody is not called, because the
> > > transparent is not visible (check in Component.java:2245).
> > >
> > > tbh, i'm out of ideas on this one :(
> > >
> > >
> > > On 10/24/07, Johan Compagner <[EMAIL PROTECTED]> wrote:
> > > >
> > > > When do you set a transparent resolver to none visible?
> > > > I also think if 1 parent (transparent resolver or not) is invisible
> > then
> > > > the
> > > > childs shouldn't be rendered.
> > > >
> > > > Because why else would you set it to not visible?
> > > >
> > > > But i haven't looked deeply into the resolver stuff But isn't it
> > fixable
> > > > here:
> > > >
> > > >
> > > > MarkupContainer parent = container;
> > > > while ((parent != null) && (parent.isTransparentResolver()))
> > > > {
> > > > // Try to find the component with the parent component.
> > > > parent = parent.getParent();
> > > > if (parent != null)
> > > > {
> > > > Component component = parent.get(tag.getId());
> > > > if (component != null)
> > > > {
> > > > component.render (markupStream);
> > > > return true;
> > > > }
> > > > }
> > > > }
> > > > in the ParentResolver class?
> > > > So an extra check in that loop if the parent is visible?
> > > >
> > > > johan
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > On 10/23/07, Gerolf Seitz < [EMAIL PROTECTED]> wrote:
> > > > >
> > > > > anybody with more insight into transparentResolvers (matej,
> igor?):
> > > > >
> > > > > what is the desired behavior for invisible transparentResolvers?
> > > > > it's really undefined right now, as the resolver is not rendered,
> > but
> > > > the
> > > > > "children" _are_ (though without markup).
> > > > >
> > > > > if "children" of an invisible transparentResolver should be
> > rendered,
> > > i
> > > > > may
> > > > > have an easy fix for this in Component#render(MarkupStream):
> > > > > instead of
> > > > > if (isRenderAllowed() && isVisible()) ....
> > > > > we could do
> > > > > if (isRenderAllowed() &&
> > > > > (isVisible() || (this instanceof MarkupContainer &&
> > > > > ((MarkupContainer)this).isTransparentResolver())))
> > > > >
> > > > > nevertheless i think that the expected behavior of setting a
> > > > > transparentResolver to invisible is that it's "children" are not
> > > > rendered.
> > > >
> > > >
> > > >
> > > > Gerolf
> > > > >
> > > > > On 10/23/07, Gerolf Seitz (JIRA) <[EMAIL PROTECTED]> wrote:
> > > > > >
> > > > > > invisible TransparentResolver skips markup of visible children
> and
> > > > thus
> > > > > > resulting in an exception in Page#checkRendering (component not
> > > found
> > > > in
> > > > > > markup)
> > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> --------------------------------------------------------------------------------------------------------------------------------------------------------
> > > >
> > > > > >
> > > > > > Key: WICKET-1095
> > > > > > URL:
> > > > https://issues.apache.org/jira/browse/WICKET-1095
> > > > > > Project: Wicket
> > > > > > Issue Type: Bug
> > > > > > Affects Versions: 1.3.0-beta4
> > > > > > Reporter: Gerolf Seitz
> > > > > > Fix For: 1.3.0-beta5
> > > > > >
> > > > > >
> > > > > > the code involved can be found at
> > > > > > http://papernapkin.org/pastebin/view/13287
> > > > > >
> > > > > > if a transparent (= isTransparentResolver() {return true;})
> > > > > > markupContainer is set to invisible, the component's markup is
> > > skipped
> > > > (
> > > > > > markupStream.skipComponent() in Component.java line 2319).
> > > > > > if the parent containing the transparent markupContainer is
> added
> > to
> > > > the
> > > > > > ajax request target and there is a component (in this case a
> > button)
> > > > > > "inside" the transparent markupcontainer's markup but is
> actually
> > a
> > > > > child of
> > > > > > the transparent markupcontainer's parent, the button is never
> > > actually
> > > > > > rendered (due to skipping the markup of the transparent
> > > > > webmarkupcontainer)
> > > > > > but is still visibleInHierarchy (see Page.java line 1005),
> because
> > > the
> > > > > > invisible transparent markupcontainer is not a parent of the
> > button.
> > > > > >
> > > > > > the ultimate question is: what is the desired behavior for an
> > > > invisible
> > > > > > transparentResolver? render it's "children" or not?
> > > > > >
> > > > > > --
> > > > > > This message is automatically generated by JIRA.
> > > > > > -
> > > > > > You can reply to this email to add a comment to the issue
> online.
> > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
