Re: [patch] testsuite/gdb.c++/local.exp: accept more nested types

[Michael Elizabeth Chastain <mec at shout dot net>]

Hi Daniel,

> I think that I (as C++ maintainer rather than testsuite maintainer)
> need to figure out what the correct output is. 

Well, actually, the manual should specify the output of ptype,
and you should implement the manual, and I should test to the manual.
Without a written spec, we have to thrash this out among ourselves.

> "InnerLocal::NestedInnerLocal nest1" is not actually valid C++ in
> general, I don't believe; what if there is a class
> InnerLocal::InnerLocal::NestedInnerLocal?

Fortunately for us, that is not legal C++.  gcc 2.96-rh and gcc 3.1
both give me a "duplicate nested type" error.

I admit this is a technicality.  But for purposes of human readability,
it means that "InnerLocal::NestedInnerLocal nest1" will behave according
to the expectations of a reasonable human reader.

> This is also uglier; for std:: types there's going to be a lot of
> useless noise.

It's useful noise if the user has their own type with the same name
as a std:: type.

> By this logic, InnerLocal::NestedInnerLocal deserves a KFAIL comment
> for the moment, since the code to do this is not yet present.

I don't think it's a failure at all.

Frankly, if we do decide it's a failure, I'd rather just leave it as
a FAIL for a while.  I don't want to add a bunch more regular expressions
and file a bug report just to make it a different kind of failure.

We're at an impasse here ... does anyone else have an opinion?

Michael C