This is the mail archive of the gcc-help@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: Overloaded method resolution order

From: Nathan Sidwell <nathan at codesourcery dot com>
To: Stuart Brooks <stuartb at cat dot co dot za>
Cc: gcc-help at gcc dot gnu dot org
Date: Thu, 13 Jan 2005 16:33:41 +0000
Subject: Re: Overloaded method resolution order
Organization: CodeSourcery LLC
References: <OF548326CE.CC57752F-ON88256F87.00115E54-88256F87.00122173@peoplesoft.com> <m3brbvqkiq.fsf@gossamer.airs.com> <01a601c4f8a0$2e7bd620$5e22a8c0@STU95>

Stuart Brooks wrote:

According to Stroustrup (7.4), certain criteria are tried in order, so as to
determine which matches the closest. I would have thought that, at least in
the case of s2 (above), the "const char* getString() const" method would
have been selected. I had also hoped that since the second one was
protected, it would use the first one instead in any case. Am I missing
something here?


yes.  G++'s behaviour is correct.  The non-const version
is selected because the object is non-constant.  therefore an exact
match is a better conversion than the qualification conversion (as the
latter contains the former).  Access is not considered until after overload
resolution.  That separation allows you to relax the access on
a member and not break an already working program.

nathan

--
Nathan Sidwell    ::   http://www.codesourcery.com   ::     CodeSourcery LLC
nathan@codesourcery.com    ::     http://www.planetfall.pwp.blueyonder.co.uk

References:
- gcc 2.96 -wchar-stdc++ option in 3.2.3?
  - From: Mai-San_Chan
- Re: gcc 2.96 -wchar-stdc++ option in 3.2.3?
  - From: Ian Lance Taylor
- Overloaded method resolution order
  - From: Stuart Brooks

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]