omission of "virtual" in overridden method declarations in derived classes
in [C++]
|
Prev: omission of "virtual" in overridden method declarations in derived classes
Next: Version 1.11 of C++ Middleware Writer now on line
From: James Hopkin on 23 Feb 2010 03:02 On Feb 23, 11:22 am, "Marco Nef" <maill...(a)shima.ch> wrote: > >> Or is the virtuality inherited implicitly from the root of the > >> inheritance tree even if it is omitted in intermediate classes? > > > Yes. C++ does not require the programmer to repeat the 'virtual' keyword > > for a virtual method in a derived class. > > > Some time ago I asked to change this in the new standard, so that virtual > must be repeated in derived classes. But the reply in this group was that > lots of old code would not compile anymore. That was a bad answer, as > compilers could have a flag to ignore such an additional security feature in > old code... > I think you have this the wrong way round. Changing the Standard to make a huge amount of existing correct code invalid is a 'bad answer'. A 'good answer', assuming for a moment that this requirement is a good thing, would be for compilers to offer this feature as an extension. In fact, Visual C++ does something similar, where you declare overriding functions in derived classes both 'virtual' and 'override' (contextual keyword). In my experience, errors of the kind you mention are common in code- bases with deep and sprawling hierarchies. Component-based designs, for example, suffer less, if at all. (A similar problem is that specifying the correct base class to call is error prone - Visual C++ offers the __super keyword for this case). James -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: Chris Uzdavinis on 23 Feb 2010 03:01 On Feb 23, 5:22 am, "Marco Nef" <maill...(a)shima.ch> wrote: > >> Or is the virtuality inherited implicitly from the root of the > >> inheritance tree even if it is omitted in intermediate classes? > > > Yes. C++ does not require the programmer to repeat the 'virtual' keyword > > for a virtual method in a derived class. > > Which is one of the worst "features" of C++ in large projects, as it causes > lots of errors (how do you know that a method is virtual if it is not > written in the declaration of the class you are working with?). It should > also be that base methods can only be called in the direct base class, or > the programmer has to explicitly write a cast to skip a hierachy level. > > Some time ago I asked to change this in the new standard, so that virtual > must be repeated in derived classes. But the reply in this group was that > lots of old code would not compile anymore. That was a bad answer, as > compilers could have a flag to ignore such an additional security feature in > old code... The new standard has something that will address this issue, using attributes to designate that a function overrides (or hides) a base class name. These attributes are [[override]] and [[hiding]], respectively. If your derived class claims to override or hide, but doesn't actually do so, it is an error. Further, you can opt-in to having a class require such attributes by adding an attribute to the class, with the attribute [[base_check]]. Then, if a class overrides or hides a base class name and doesn't indicate that it does, that too is also a compile-time error. See section 7.6.5 in the draft for details. Currently: http://www.open-std.org/JTC1/sc22/wg21/docs/papers/2010/n3035.pdf -- Chris [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: Anders Dalvander on 23 Feb 2010 15:54
On Feb 23, 8:57 pm, Edward Diener <eldiener_no_spam_h...(a)tropicsoft.com> wrote: > It is called documentation. That's how you know. Sometimes the code is the documentation. Anyway, templates can mess things up. See example below, and answer the following question: Is Derived::DoIt a virtual member function? template <class TBase> class Derived : public TBase { public: void DoIt() { std::cout << "Hello World" << std::endl; } }; If changing the rules to the suggestion of Marco, and others, the answer is easy: No. (Otherwise a compile time error would occur.) But according to the current rules the answer is: It depends. Personally I haven't decided what I think is is best. Regards, Anders Dalvander -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] |