Hmmm. While I see your point, it does fit the mold when you consider that
multivalued annotation properties still accept a single-valued notation in
Java itself. The form "@Gimme(Serializable.class)" is equivalent to
"@Gimme({Serializable.class})".
I suppose that I'm a bit more pragmatic about it, and would accept
overloading "=" to mean "contains". I know that my annotation property
started single-valued, then I change to multivalued when I discovered its
limitation. Had aspectj supported multivalued attribute properties as
strictly as you're suggesting, then I would've had to also update my type
expression accordingly, which would've been kind of a pain.
Since Java itself is willing to provide the single-valued shortcut, I'm also
willing to do the same in pointcut matching. You could even argue that by
not overloading "=", you are violating the principle of least surprise.
I propose that "=" support the single-valued case and, in the future, be
overloaded to mean "contains" when full multivalued support is introduced.
-matthew
--
View this message in context:
http://aspectj.2085585.n4.nabble.com/Matching-on-one-of-an-annotation-s-values-in-a-multivalued-property-tp4650839p4650846.html
Sent from the AspectJ - users mailing list archive at Nabble.com.
_______________________________________________
aspectj-users mailing list
aspectj-users@eclipse.org
https://dev.eclipse.org/mailman/listinfo/aspectj-users
