Taking discussion to dev_at_javaserverfaces.java.net.
>>>>> On Sun, 14 Dec 2014 17:11:20 +0100, Frank Caputo <info_at_frankcaputo.de> said:
FC> Ive been looking at this some more and I think, my second patch
FC> makes mojarra work exactly as you intended with
FC> JAVASERVERFACES_SPEC_PUBLIC-719. I changed the failing test to load
FC> a facelet from the webapp root, which still works.
FC> The FaceletWebappResourceHelper is useless for other resources than
FC> facelets, because getNonCompressedInputStream throws an
FC> exception. So I made it, that FaceletWebappResourceHelper is queried
FC> only when the other 2 ResourceHelpers dont find the desired
FC> resource.
FC> I disabled FaceletWebappResourceHelper for contracts (because
FC> WebappResourceHelper already does the trick).
FC> I disabled FaceletWebappResourceHelper for locale prefixes, because
FC> facelets are not supported with locale prefixes (though I like the
FC> idea of localized facelets).
FC> I disabled FaceletWebappResourceHelper with libraries, because
FC> facelets in libraries are only possible with composite components,
FC> which are loaded by the other 2 ResourceHelpers.
FC> Can you please check the second patch again?
FC> IMO loading resources from the webapp root should still be an
FC> exception exclusively for facelets, because otherwise the layout of
FC> a JSF application would be really complicated.
Yes, I agree. I seem to recall you found a bug in the pseudo-code in
spec PDF section 2.6.1.4 that is related to this. Can you remind me of
that? I looked in the EG archive, but didn't find the message.
In the meantime, I'm running the tests with your patch applied. I'll
followup with the results.
Ed
--
| edward.burns_at_oracle.com | office: +1 407 458 0017
| 41 days til DevNexus 2015
| 51 days til JavaLand 2015
| 61 days til CONFESS 2015