- Calling virtual function from base class constructor - 4 Updates
- Fancy font in Usenet post - 3 Updates
- LLVM 5.0.0 released - 1 Update
- Jesus Christ in His proper place - 1 Update
| Juha Nieminen <nospam@thanks.invalid>: Sep 08 06:43PM I have quite a design dilemma, which I will introduce with a quiz question: What does the following print? //------------------------------------------------------------- #include <iostream> class Base { public: Base { func(); } virtual void func() const { std::cout << "base\n"; } }; class Derived: public Base { public: virtual void func() const override { std::cout << "derived\n"; } }; int main() { Derived derived; } //------------------------------------------------------------- I understand why it does what it does. However, this poses a problem. I have a base class that implements certain functionality, and is often instantiated directly. Many details of this functionality can be specialized by derived classes, by implementing certain virtual functions of the base class. These virtual functions specify how the base class should behave in diverse situations. These virtual functions may, for instance, return values which the base class uses (such as colors, textures and other such stuff). In some cases the base class needs to know these values in its constructor, as well as elsewhere (during the existence of the instance). A very simplified (and simplistic) example of how it might work (the actual situation is more complicated): //------------------------------------------------------------- class Base { public: Base(some_parameters) { doStuffWithValues(getValues()); } void anotherFunction(some_parameters) { doStuffWithValues(getValues()); } private: void doStuffWithValues(Values); protected: virtual Values getValues() const; }; //------------------------------------------------------------- You might immediately see the problem: It's getting the incorrect values in its constructor, because it's getting its own values, rather than the ones from the derived class. I wonder if there's a nice and clean solution to this (which, preferably, doesn't require changing the signature of the base class constructor). |
| Bo Persson <bop@gmb.dk>: Sep 08 08:54PM +0200 On 2017-09-08 20:43, Juha Nieminen wrote: > I wonder if there's a nice and clean solution to this (which, > preferably, doesn't require changing the signature of the > base class constructor). The nice and clean solution IS to pass the values needed as parameters to the constructor. The base class might then even store the values it is passed and you don't have to make getValues virtual. Bo Persson |
| scott@slp53.sl.home (Scott Lurndal): Sep 08 06:54PM >I wonder if there's a nice and clean solution to this (which, >preferably, doesn't require changing the signature of the >base class constructor). One must ask _where_ does the derived class get the values from? If from the derived constructor, then they could have been passed to the base class constructor as well. You might consider using an 'init' paradigm - where the base and derived classes have init members instead of constructors, where the derived init member explictly calls the base init member (and the creator of either object calls the init function explicitly after creating the object). As the object is fully realized by the time the init member is called, the base class init function can use the virtual functions without issue. |
| "Alf P. Steinbach" <alf.p.steinbach+usenet@gmail.com>: Sep 08 11:39PM +0200 On 9/8/2017 8:43 PM, Juha Nieminen wrote: > I wonder if there's a nice and clean solution to this (which, > preferably, doesn't require changing the signature of the > base class constructor). There are many type safe solutions, and I once convinced Marshall to put them in the FAQ, <url: https://isocpp.org/wiki/faq/strange-inheritance#calling-virtuals-from-ctor-idiom> Also see this old blog posting of mine, which provides more holistic concrete examples instead of just a listing of techniques, <url: https://alfps.wordpress.com/2010/05/12/how-to-avoid-post-construction-by-using-parts-factories/> Bjarne discusses some aspects of why the 1990's two-phase construction solution is bad, in <url: http://www.stroustrup.com/3rd_safe0.html> (search the PDF for "init"), but he fails to address why it was so commonly used, namely to solve your problem in a child's way, sort of. Summing up: • Don't use two-phase construction. • Somehow (constructor args, templating) pass the requisite info to the base class constructor. Cheers & hth., - Alf |
| Jorgen Grahn <grahn+nntp@snipabacken.se>: Sep 08 05:39AM On Thu, 2017-09-07, Öö Tiib wrote: >> But with some trickery you can do almost anything.!! > Do that trickery using https://groups.google.com/ like fir, (who brought > it up) requested. And do it on alt.test. Not here. /Jorgen -- // Jorgen Grahn <grahn@ Oo o. . . \X/ snipabacken.se> O o . |
| fir <profesor.fir@gmail.com>: Sep 08 07:13AM -0700 W dniu czwartek, 7 września 2017 22:01:49 UTC+2 użytkownik Good Guy napisał: > With over 500 million devices now > running Windows 10, customer satisfaction is higher than any > previous version of windows. here it worked "of course it doesnt.." was in curier and "With over 500 milion " was yellow on blue |
| "Chris M. Thomasson" <invalid@invalid.invalid>: Sep 08 12:39PM -0700 On 9/8/2017 7:13 AM, fir wrote: >> But with some trickery you can do almost anything.!! >> I didn't see the original post but it looks like he/she might be on >> my killer-switch!! [...] > here it worked "of course it doesnt.." was in curier and "With over 500 milion " was yellow on blue Not for me. Well, I do view my messages in plaintext. ;^) |
| Lynn McGuire <lynnmcguire5@gmail.com>: Sep 08 12:01PM -0500 LLVM 5.0.0 released http://lists.llvm.org/pipermail/llvm-announce/2017-September/000075.html "This release is the result of the community's work over the past six months, including: C++17 support, co-routines, improved optimizations, new compiler warnings, many bug fixes, and more." Hat tip to: http://www.osnews.com/story/29998/LLVM_5_0_0_released Lynn |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Sep 08 06:19AM -0700 A good follow-up statement / question and reply: https://groups.google.com/d/msg/comp.lang.c/-dC-tMZeryM/t4lkvcU9AQAJ Thank you, Rick C. Hodgin |
| You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to comp.lang.c+++unsubscribe@googlegroups.com. |
No comments:
Post a Comment