C++Talk.NET C++ language newsgroups   Archives   FAQ   Search   Memberlist   Usergroups   Register   Profile   Log in to check your private messages   Log in  Use of 'const' with pImpl objects          C++Talk.NET Forum Index -> C++ Language (Moderated) View previous topic :: View next topic   Author Message Old Wolf Guest Posted: Mon Apr 18, 2005 9:03 am    Post subject: Use of 'const' with pImpl objects Example: struct FooImpl; struct Foo { void bar(); void baz() const; Foo(); ~Foo(); private: FooImpl *pImpl; } The 'const' qualifier on Foo::baz does not actually have any effect, because Foo::baz (and indeed, all of the member functions) don't modify pImpl; they only modify *pImpl. It would be possible to make all member functions const, or no member functions const, and have no compiler errors. Should I eschew it completely here? Or should I continue to apply 'const' to the same functions that a non-pImpl class would have had them, as a means of self-documentation? [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Mark Van Peteghem Guest Posted: Mon Apr 18, 2005 11:05 am    Post subject: Re: Use of 'const' with pImpl objects Old Wolf wrote: Quote: Example: struct FooImpl; struct Foo { void bar(); void baz() const; Foo(); ~Foo(); private: FooImpl *pImpl; } The 'const' qualifier on Foo::baz does not actually have any effect, because Foo::baz (and indeed, all of the member functions) don't modify pImpl; they only modify *pImpl. It would be possible to make all member functions const, or no member functions const, and have no compiler errors. Should I eschew it completely here? Or should I continue to apply 'const' to the same functions that a non-pImpl class would have had them, as a means of self-documentation? You could use it as self-documentation, but also as documentation to other users of your class. Many classes do this, e.g. std::vector has many const methods that could actually change the contents of the vector (because these are in a dynamically allocated array), but they don't. Many of its methods even have both a const and a non-const version (e.g. operator[](size_type)), where the non-const version allows modification of the vector, while the const version doesn't. This changes the notion of const-ness, the contents of the vector become really part of the object. -- Mark dot Van dot Peteghem at q-mentum dot com http://www.q-mentum.com -- easier and more powerful unit testing [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Ahti Legonkov Guest Posted: Mon Apr 18, 2005 11:07 am    Post subject: Re: Use of 'const' with pImpl objects Old Wolf wrote: Quote: It would be possible to make all member functions const, or no member functions const, and have no compiler errors. Should I eschew it completely here? Or should I continue to apply 'const' to the same functions that a non-pImpl class would have had them, as a means of self-documentation? You should use const the same way you would with non-pImpl. But you can get some help from compiler using a smart pointer. Something like this: #include <boost/scoped_ptr.hpp> template <typename T> class fancy_ptr : boost::scoped_ptr<T> { typedef boost::scoped_ptr<T> base_t; public: fancy_ptr(T* p) : base_t(p) { } ~fancy_ptr() { } // make pointee const if pointer is const T const* operator->() const { return base_t::operator->(); } T const& operator*() const { return base_t::operator*(); } T* operator->() { return base_t::operator->(); } T& operator*() { return base_t::operator*(); } }; class Blah { struct Impl; fancy_ptr<Impl> pImpl; public: void bar(); void baz() const; Blah(); ~Blah(); }; struct Blah::Impl { int i; }; Blah::Blah() : pImpl(new Impl()) { } Blah::~Blah() { } void Blah::bar() { pImpl->i = 9; } // OK void Blah::baz() const { pImpl->i = 10; } // ERROR - pImpl is const :) :) -- Ahti Legonkov leg zero at hot dot ee [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Alan Griffiths Guest Posted: Tue Apr 19, 2005 10:02 pm    Post subject: Re: Use of 'const' with pImpl objects "Old Wolf" <oldwolf (AT) inspire (DOT) net.nz> wrote Quote: It would be possible to make all member functions const, or no member functions const, and have no compiler errors. Should I eschew it completely here? Or should I continue to apply 'const' to the same functions that a non-pImpl class would have had them, as a means of self-documentation? Const serves a duel role - primarily it separates out the const and not-const interfaces available to client code (which is unaffected), and also provides some checking of implementation code. The latter role can be reinstated by using a const qualified pointer instead of a raw pointer. In this case grin_ptr<> seems a good fit for your requirement: http://www.octopull.demon.co.uk/arglib/TheGrin.html -- Alan Griffiths http://www.octopull.demon.co.uk/ [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top kanze@gabi-soft.fr Guest Posted: Wed Apr 20, 2005 9:12 am    Post subject: Re: Use of 'const' with pImpl objects Old Wolf wrote: Quote: Example: struct FooImpl; struct Foo { void bar(); void baz() const; Foo(); ~Foo(); private: FooImpl *pImpl; } The 'const' qualifier on Foo::baz does not actually have any effect, because Foo::baz (and indeed, all of the member functions) don't modify pImpl; they only modify *pImpl. It would be possible to make all member functions const, or no member functions const, and have no compiler errors. Should I eschew it completely here? Or should I continue to apply 'const' to the same functions that a non-pImpl class would have had them, as a means of self-documentation? This is the old logical const (aka Humpty-Dumpty const) vs. bit-wise const discussion. I believe that a consensus exists for logical const, although some ten years back, the discussion could be animated. The use of const in the interface isn't just documentation. Client code which only has a const object, or a reference to const, cannot call the non-const functions without a const_cast. The fact that it is you, and not the compiler, who defines exactly what const means doesn't mean that it doesn't mean anything. It remains an integral part of the type system. -- James Kanze GABI Software Conseils en informatique orientée objet/ Beratung in objektorientierter Datenverarbeitung 9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34 [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Display posts from previous: All Posts1 Day7 Days2 Weeks1 Month3 Months6 Months1 Year Oldest FirstNewest First         C++Talk.NET Forum Index -> C++ Language (Moderated) All times are GMT Page 1 of 1   Jump to: Select a forum C/C++----------------C LanguageC Language (Moderated)C++ Language (Moderated)C++ language (comp.lang.c++)C++ language, library and standards Non-English C/C++ Newsgroups----------------C++ (German)C++ (French)  You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum Powered by phpBB © 2001, 2006 phpBB Group SEO toolkit © 2004-2006 webmedic.