Reported by Pavel:

> $ echo ‘call(* A+..foo(..))’ | parse-aspectj -s PointcutExpr
> Call(MethodPattern(AnnoPattern([]),ModPattern([]),NamePattern(“*”),
> WildcardMemberName(Subtype(NamePattern(“A”)),“foo”),
> [FormalWildcard],None))
>
> ‘A+..foo’ is actually not a valid pattern there, according to abc’s
> grammar, and both abc and ajc reject it.

Submitted on 7 June 2006 at 14:38

AJF-27, martin:

In general, I’m sometimes a bit more liberal in the Java and AspectJ
syntax, by not rejecting things syntactically that are really just
semantic restrictions. In particular, in some cases the Java and
AspectJ syntax is more liberal if this makes the definition of the
language more attractive (or to me more precise: not to introduce all
kinds of redundant syntactical sorts, which makes understanding of the
language and the definition of embeddings more complicated).

I’m not sure what to do in this case. I’ll take a look at the
definition in ajc and abc and would like to know your opinion on being
syntactically a bit more liberal in some situations.