Another good example. Fortunately it is also straightforward to get
working. I checked in another small change to scope searching to make
sure that splitable names (ie; b::Foo) can start in namespaces in
scope, not just classes in the lineage of the expandable tag.
Did that make sense?
Well, I added your example into the test harness.
Thanks
Eric
>>> Hannu Koivisto <azure@...> seems to think that:
>Greetings,
>
>--- foo.cpp ---
>namespace a {
>namespace b {
>
>class Foo
>{
>public:
> int write();
>};
>
>} // namespace b
>} // namespace a
>
>
>namespace a {
>namespace c {
>
>class Bar
>{
>public:
> int baz();
>
>private:
> b::Foo &foo;
>};
>
>int Bar::baz()
>{
> return foo.w<complete>
>}
>
>} // namespace c
>} // namespace a
>--- foo.cpp ---
>
>Semantic cannot complete foo.w to foo.write in Bar::baz().
>
|