This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Export bits of the demangler's internal interface
On Wed, Dec 31, 2003 at 03:29:46PM -0500, Ian Lance Taylor wrote:
> Daniel Jacobowitz <drow@mvista.com> writes:
>
> > What do you think of the below? These are the bits of the demangler that
> > I'm currently re-using in GDB. It's not a pretty interface, but I'm more
> > interested in using it than designing a proper interface for it; if you'd
> > rather do this from the other side first then I can put some more thought
> > into it. Right now it's easy enough to update all clients when you make an
> > incompatible change.
>
> My bias would be toward designing an interface which is at least
> slightly better. In particular, I'm not sure the arrays should be
> exposed as arrays.
>
> I gather that the main thing you need is the ability to generate
> arbitrary mangling trees, and the ability to print the demangled
> string corresponding to such a tree.
Yes, that's right. I'll try to find some time for this in the next
week or two, then. Not exposing the arrays is trivial. Any preference
on names for global symbols, since DJ suggested that the d_* names may
not be appropriate?
> > Also, how concerned are you about changing the demangler output? Is that a
> > major compatibility issue? I'd like to print char constants using
> > single-quoted literals, instead of numerically.
>
> Changing the demangler output is not a major compatibility issue.
> Ideally any changes from here on out should be reflected in the
> libiberty testsuite, so that we at least don't change them
> accidentally.
OK. That'll be my next issue after this one then.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer