Yeah -- converting this to a warning seems right. Makes it a smoother, more symmetric path to be able to use action="..." for anything -- method in a class, base class, merged Struts config file, etc.
+1 from me. :) Eddie On 1/13/06, Carlin Rogers <[EMAIL PROTECTED]> wrote: > Hi Rich, > > Thanks for the feedback. I think if there are no objections, I'll file a bug > and make the change. > > Kind regards, > Carlin > > On 1/13/06, Rich Feit <[EMAIL PROTECTED]> wrote: > > > > Hey Carlin, > > > > The action attribute was intended to be stricter, since it *must* be an > > action in the current page flow. But you're right, with Struts merge we > > can't know whether a valid action is being merged in. I think relaxing > > to a warning is sufficient. > > > > Rich > > > > Carlin Rogers wrote: > > > > >Hey Rich, > > > > > >I have a compiler question. A forward that references a non-existent > > action > > >using the *action* attribute is currently an error. Referencing a > > >non-existent action or page with the *path* attribute results in a > > compiler > > >warning. > > > > > >For example, I tried a forward annotation with the action attribute set > > to a > > >struts action from a merged struts config. I don't think the compiler > > takes > > >into account the Struts merge for a valid action. When I looked through > > the > > >code, it appears that for a valid action we check the method names on the > > >given page flow, the named shared flows, and finally the deprecated > > >Global.app. So no check on a struts merge. > > > > > >Do you have insight as to why it is an error and thoughts about relaxing > > >this to a warning just like the path attribute? > > > > > >Thanks, > > >Carlin > > > > > > > > > > > > >
