if (!hasThis()) {
throw new IllegalStateException("no this");
}
- initializeThisAnnotationVars(); // FIXME ASC Why bother with this if we always return one?
+ initializeThisAnnotationVars(); // FIXME asc Why bother with this if we always return one?
// Even if we can't find one, we have to return one as we might have this annotation at runtime
Var v = (Var) thisAnnotationVars.get(forAnnotationType);
if (v==null)
if (!hasTarget()) {
throw new IllegalStateException("no target");
}
- initializeTargetAnnotationVars(); // FIXME ASC why bother with this if we always return one?
+ initializeTargetAnnotationVars(); // FIXME asc why bother with this if we always return one?
Var v =(Var) targetAnnotationVars.get(forAnnotationType);
// Even if we can't find one, we have to return one as we might have this annotation at runtime
if (v==null)
argAnnotationVars = new Map[numArgs];
for (int i = 0; i < argAnnotationVars.length; i++) {
argAnnotationVars[i] = new HashMap();
- //FIXME ASC Fill this in
- // populate
+ //FIXME asc just delete this logic - we always build the Var on demand, as we don't know at weave time
+ // what the full set of annotations could be (due to static/dynamic type differences...)
}
}
public void initializeKindedAnnotationVars() {
// Then create one BcelVar entry in the map for each annotation, keyed by
// annotation type (TypeX).
- // FIXME Refactor these once all shadow kinds added - there is lots of commonality
+ // FIXME asc Refactor these once all shadow kinds added - there is lots of commonality
ResolvedTypeX[] annotations = null;
TypeX relevantType = null;