Good point re:adaptTo Julian. I am not sure there is a case where this would happen, but some sort of failure would make sense in the case where the code is missing. Let me investigate the options for exception handling e.g. on compilation vs runtime to see if there are any gotchas, specifically around what happens if the class is injected into OSGI after the JSP is first compiled.
-Dan On Fri, Jun 13, 2014 at 5:31 AM, Julian Sedding <[email protected]> wrote: > Hi Dan > > I agree with your fix to getProperty. That makes a lot of sense. > > I'm not sure about the fix for the adaptTo tag to prevent compilation > failure. Java classes will not compile, if they reference a class that > cannot be found. Why should JSPs behave differently? IMO there is > value in failing early (i.e. compilation), because most likely the > code is broken in any case, if the class cannot be found. > > Is there any real-world code leveraging this "feature" of the adaptTo > tag? Otherwise I would lean towards adjusting the test case. > > In any case, I can live with either solution. > > Regards > Julian > > > On Thu, Jun 12, 2014 at 9:22 PM, Daniel Klco <[email protected]> wrote: > > As discussed in SLING-3664 the new TEI classes are breaking the > integration > > tests and change the behavior of some of the tags, specifically: > > > > - In adaptTo it causes the JSP to fail to compile when specifying an > > invalid class instead of just returning null as is the usual Sling > behavior > > - in getProperty it causes the JSP to fail to compile if neither a > > defaultValue nor a returnClass are specified > > > > If we cannot provide backwards compatibly for these issues, I suggest we > at > > least remove the TEI classes for these tags. > > > > Thanks, > > Dan > > > > > > On Thu, Apr 24, 2014 at 8:36 AM, Julian Sedding <[email protected]> > wrote: > > > >> Updated the patch in SLING-3475 to provide TEI implementations for all > >> tags in the Sling taglib. > >> > >> Regards > >> Julian > >> > >> On Wed, Apr 23, 2014 at 5:03 PM, Konrad Windszus <[email protected]> > wrote: > >> > I definitely agree here. Including the TEI is almost no risk but it > >> helps a lot during development (at least with IntelliJ, because Eclipse > >> still lacks decent EL code completion support) > >> > > >> > On 23 Apr 2014, at 16:52, Julian Sedding <[email protected]> wrote: > >> > > >> >> Hi all > >> >> > >> >> It would be great to get the patch for SLING-3475 applied soon. > >> >> > >> >> The added TagExtraInfo should be low risk, as it should not impact > >> >> run-time behavior at all. > >> >> > >> >> However, enabling my IDE to help prevent small bugs would be very > >> >> helpful. Especially in conjunction with Sling Models, which make the > >> >> <sling:adaptTo/> tag much more useful. > >> >> > >> >> Thanks & regards > >> >> Julian > >> >> > >> >> https://issues.apache.org/jira/browse/SLING-3475 > >> > > >> >
