Re: GCC-3.0: possible regression

[Jan van Dijk <jan at etpmod dot phys dot tue dot nl>]

The example code I presented was a bit misleading: forget about the
char& version below.
	
The real problem is the ambiguity due to having both:

1) char operator[](size_t) const;
2) operator const char*() const;

in a class. gcc-3.0 does not know what to do when writing `s[0]', while
previous versions generated no errors.

FYI: This issue is not merely academic, as the sample code I presented
is a stripped down version of the wxString class in the widely used
wxWindows library (www.wxwindows.org). If indeed gcc-3.0 is right, there
is a serious problem with that class which cannot be easily resolved
without losing backward compatibility.

	Thanks a lot, Jan.

Fergus Henderson wrote:
> 
> On 03-Jul-2001, jan@etpmod.phys.tue.nl <jan@etpmod.phys.tue.nl> wrote:
> > The following code compiles with 2.95.3, but not with 3.0. Before filing
> > a bug report, I would like your opinion what the standard has to say
> > about it (I did not find a previous related report). What's wrong: the
> > code or 3.0?
> >
> > #include <cstddef>
> > struct MyString
> > {
> >         char operator [](size_t) const;
> >         char& operator [](size_t);
> >         operator const char*() const;
> > };
> >
> > main()
> > {
> >         MyString s;
> >         char c = s[0];
> >         return 0;
> > }
> >
> > Compiler output follows:
> >
> > jan@gum03:/real-home/jan > g++ -c test.cpp
> > test.cpp: In function `int main()':
> > test.cpp:12: choosing `char& MyString::operator[](unsigned int)' over
> >    `operator[]'
> > test.cpp:12:   because worst conversion for the former is better than
> > worst
> >    conversion for the latter
> >
> > I understand what's going on, but had expected the const-version of the
> > []-operator to be chosen, like with 2.95.3.
> 
> Why would the const version of operator [] be chosen?
> The non-const version is an equal match on the size_t argument,
> and a better match on the `this' argument, since the object `s'
> is not const.  So I think 3.0 is correct here.
> 
> --
> Fergus Henderson <fjh@cs.mu.oz.au>  |  "I have always known that the pursuit
> The University of Melbourne         |  of excellence is a lethal habit"
> WWW: <http://www.cs.mu.oz.au/~fjh>  |     -- the last words of T. S. Garp.