Re: Patch to fix PR9861

[Andrew Haley <aph at redhat dot com>]

TJ Laurenzo writes:
 > On 9/26/05, Andrew Haley <aph@redhat.com> wrote:
 > > TJ Laurenzo writes:
 > >  > On 9/26/05, Andrew Haley <aph@redhat.com> wrote:
 > >  > > I tried your patch, and while it correctly demangles, it doesn't allow
 > >  > > me to put a breakpoint on 'f.main(java.lang.String[])' However,
 > >  > > putting a breakpoint on 'void f.main(java.lang.String[])' does work,
 > >  > > as you might expect.
 > >  > >
 > >  > > This has the rather nasty effect of breaking tab completion in gdb.
 > >  > > For example, you can't type "break 'f<tab>" and get a nice list of the
 > >  > > methods in f to choose from.
 > >  > >
 > >  > > It ought to be possible to fix this in gdb, though.  I'll investigate
 > >  > > some more.
 > >  >
 > >  > That just occurred to me about 10 seconds before your email hit my
 > >  > mailbox.  There are some C++ cases where this can happen too (function
 > >  > templates and operators, if I recall).  They are probably corner cases
 > >  > that GDB doesn't handle specially, though.
 > >  >
 > >  > What if the return type demangling didn't print the return type first
 > >  > (the return type could be in parentheses or something after the
 > >  > argument list or it could end with " returns void")?  That would also
 > >  > be a way to solve the problem.
 > >
 > > It would.  Nice idea.
 > >
 > Would you agree with changing the display for Java formatted names to
 > print it more like a Pascal declaration:
 >     f.main(String[]):void

At first glance that looks OK.  It would be interesting to see if that
works with gdb.

 > I don't think we would need to touch the existing C++ format, even
 > though it sounds like it could benefit from the same things for C++
 > templates, et al.

Could be.

Andrew.