- Boost-users - lists.preview.boost.org

RE: [Boost-Users] Polymorphic containers (was Re: shared_ptr and inheritance)
by Björn Karlsson 12 Dec '01

12 Dec '01

> From: rz0 [mailto:rz0@yahoo.com] > Hi Bjorn > > Thank you very much for the reply. > > Unfortunately these classes are provided to me and as part of the > requirements, I must be able to work with them. In re-thinking the > problem, it seems that I'm in desperate need of a polymorphic > container - i.e.: > > > Does boost have any polymorphic containers? > > Thanks very much in advance. > > I don't think that polymorphism is the cure - you'll always need to downcast the pointer as the signatures differ. However, there is support in Boost for what you're trying to do - through boost::any. Like I said before, the solution will not be very clean, but it will work. The code snippet below seems to do what you're asking for: #include <iostream> #include <algorithm> #include <vector> #include <Boost\Any.hpp> class A { public: virtual void func1() { std::cout << "A::func1()\n"; }; }; class B : public A { public: void func1() { std::cout << "B::func1()\n"; }; void funcB() { std::cout << "B::funcB()\n"; }; }; class C : public A { public: void func1() { std::cout << "C::func1()\n"; }; void funcC() { std::cout << "C::funcC()\n"; }; }; typedef std::vector<boost::any> AnyVector; void DoStuffWithAny(boost::any& a) { // Try with a B if ( B* pB = boost::any_cast<B>(&a) ) { pB->func1(); pB->funcB(); } // Try with a C else if ( C* pC = boost::any_cast<C>(&a) ) { pC->func1(); pC->funcC(); } // Try with an A else if ( A* pA = boost::any_cast<A>(&a) ) { pA->func1(); } } void DoStuffWithAnyPointer(boost::any& a) { // Try with a B if ( B** ppB = boost::any_cast<B*>(&a) ) { (*ppB)->func1(); (*ppB)->funcB(); } // Try with a C else if ( C** ppC = boost::any_cast<C*>(&a) ) { (*ppC)->func1(); (*ppC)->funcC(); } // Try with an A else if ( A** ppA = boost::any_cast<A*>(&a) ) { (*ppA)->func1(); } } int main(int argc, char* argv[]) { AnyVector aVector; A a; B b; C c; aVector.push_back(a); aVector.push_back(b); aVector.push_back(c); // Do interesting things std::for_each(aVector.begin(), aVector.end(), DoStuffWithAny); AnyVector anotherVector; anotherVector.push_back(new A); anotherVector.push_back(new B); anotherVector.push_back(new C); // Do interesting things std::for_each(anotherVector.begin(), anotherVector.end(), DoStuffWithAnyPointer); return 0; } Another approach would be to provide adapters, but I'm guessing that wouldn't be worth the trouble. HTH, Bjorn Karlsson [Non-text portions of this message have been removed]

2 1

Tokenizer
by Paul Grenyer 11 Dec '01

11 Dec '01

Hi All! I'm using tokenizer in a project. Is there a way to get the number of tokens that have been tokenized? Thanks Paul Paul Grenyer pjgrenyer(a)iee.org

1 0

Boost.Threads Win32 Implementation
by bkarsjens 11 Dec '01

11 Dec '01

The first thing I discovered when trying to get threads working is that the -mthreads option was not being used in jam. I added the following to the gcc-tools.jam file. I'm not totally sure it's right -- it's needed for mingw32, but I don't know if it is necessary for cygwin. if $(NT) { flags gcc CFLAGS <threading>multi : -mthreads ; flags gcc LINKFLAGS <threading>multi : -mthreads ; } Adding that let's gdb actually work on it... It doesn't crash deep in the bowels of thread exiting in some kernel library routine. Although it still crashes in on_thread_exit doing handlers->push_front(func). Actually, it crashes in some STL stuff, but that's I'm not sure why, I'm not that good with gdb yet. After digging through the implementation of threadmon.cpp, I realized that (unless I missed something) on_thread_exit was always called with the same argument, which is the address of cleanup in tss.cpp. Therefore all of the stuff in threadmon.cpp for keeping a list of functions registered with on_thread exit were not really needed, it would always have a list of the same pointer. So why not just keep that pointer and call it on thread exit. I also discovered that DllMain wasn't getting called at all because it wasn't defined as extern "C". This drastically simplifies threadmon.cpp: // threadmon.cpp : Defines the entry point for the DLL application. // #define BOOST_THREADMON_EXPORTS #include "threadmon.hpp" #ifdef BOOST_HAS_WINTHREADS #define WIN32_LEAN_AND_MEAN // Exclude rarely-used stuff from Windows headers #include <windows.h> #ifdef BOOST_MSVC # pragma warning(disable : 4786) #endif typedef void (__cdecl * handler)(void); namespace { handler exit_handler; } #if defined(__BORLANDC__) #define DllMain DllEntryPoint #endif extern "C" BOOL APIENTRY DllMain(HANDLE module, DWORD reason, LPVOID) { switch (reason) { case DLL_PROCESS_ATTACH: case DLL_THREAD_ATTACH: break; case DLL_THREAD_DETACH: case DLL_PROCESS_DETACH: { if (exit_handler) (*exit_handler)(); } break; } return TRUE; } int on_thread_exit(void (__cdecl * func)(void)) { exit_handler = func; return 0; } #endif // BOOST_HAS_WINTHREADS Note that the above was not copied directly, but has had some editing. I think I transcribed it correctly, but maybe I missed a semicolon or something. Also, in tss.cpp, cleanup may be called without any handlers registered... I'm not sure if this was always the case, or if I introduced this case, so the cleanup function needs to do an if(handlers) check before actually trying to call them. And now the tests pass! Wheee! One more thing. I'm still not convinced for the need of the dll. The issue is needing to know when a thread has exited, and I think that could be done by an object that was destructed in a thread-proxy function. Since you are already using a thread_proxy function, I'm wondering if you considered this approach and if so, why you rejected it.

2 3

RE: [Boost-Users] Re: operator<< for boost::any
by Björn Karlsson 11 Dec '01

11 Dec '01

> From: Patrick Guio [mailto:patrick.guio@fys.uio.no] > I have another comment more general about the any class > The declaration > placeholder * content; > in the any class might be public or private depending on the > value of the > value of the define BOOST_NO_MEMBER_TEMPLATE_FRIENDS. > Is it safe? No, it's not safe. It's there for a good reason though - some compilers does not support the (required) friend declaration. Bjorn Karlsson [Non-text portions of this message have been removed]

1 0

Re: operator<< for boost::any
by Patrick Guio 10 Dec '01

10 Dec '01

On Sun, 9 Dec 2001, assaflavieatwork wrote: > If all you want to do is output the value inside the 'any' you can > add a virtual member to the place_holder/holder classes (inside any's > impl) that will output to a ostream&. That is exactly my need > in place_holder: > virtual ostream& write(ostream&) = 0; You mean placeholder? I have added this declaration as a pure virtual function in the queries part. class placeholder { ... public: // queries ... virtual ostream& write(ostream&) = 0; }; > in holder: > ostream& write(ostream& o) { o << held; }; > I have added the declaration in the derived class holder from the abstract base class placeholder template<typename ValueType> class holder : public placeholder { ... public: // queries ... ostream& write(std::ostream& os) { os << held; } public: // representation ValueType held; }; > outside any: > std::ostream& operator << (std::ostream& out, const any& value) { > return value.write(out); > }; I don't understand how this can work since the class any contains a pointer to an object of class placeholder. This object has a member function write but not the class any itself? I have another comment more general about the any class The declaration placeholder * content; in the any class might be public or private depending on the value of the value of the define BOOST_NO_MEMBER_TEMPLATE_FRIENDS. Is it safe? #ifndef BOOST_NO_MEMBER_TEMPLATE_FRIENDS private: // representation template<typename ValueType> friend ValueType * any_cast(any *); #else public: // representation (public so any_cast can be non-friend) #endif placeholder * content; }; > > But that of course means changing the any.hpp file which of course > isn't something everyone is willing to do... Why not including a simple overloading of the << operator? > > Anyway, it won't help you if you want to read the any back from an > istream. You'll have to write some type information to the stream or > else you'll just interpret everything as a string. > I understand that but that is not what I want to do. Sincerely Patrick Guio

2 1

RE: [Boost-Users] Re: shared_ptr and inheritance
by Björn Karlsson 10 Dec '01

10 Dec '01

> From: rz0 [mailto:rz0@yahoo.com] > > > > > >VC++ 6.0 flags the above lines with a conversion error. > > > > > >Is there any way to resolve this? > > > > Can you just define b and c as boost::shared_ptr<A>? > > Hi Jon > > Thank you very much for the response. Unfortunately this will cause > the same problem that I ran into with std::auto_ptr and that is > methods funcB and funcC which are exclusive to B and C are not > allowed to be called. > > Take care > RZ > It seems to me that the problem is in the design - you're trying to use inheritance for the convenience of placeholding (and maybe insulation). That won't work (as you've seen). Even with a compiler that allows the code to compile (those that support member templates will) you will not get the behaviour that you want - funcB and funcC can never be called (from client code). The reason that the code compiles cleanly is that the conversion from B* or C* to A* is valid - but there's no way you can dynamically change the static type of "boost::shared_ptr<A> aptr;". If you really need to get this to work, it is still possible: You could use boost::any, and any_cast through all the possible types. This, however, is not very flexible, and IMO not very readable or maintainable. Regards, Bjorn Karlsson [Non-text portions of this message have been removed]

2 1

shared_ptr and inheritance
by rz0 10 Dec '01

10 Dec '01

Hello I have the following scenario: #include <iostream> #include <boost/smart_ptr.hpp> class A { public: virtual void func1() { std::cout << "A::func1" << std::endl; } }; class B : public A { public: void func1() { std::cout << "B::func1" << std::endl; } void funcB() { std::cout << "B::funcB" << std::endl; } }; class C : public A { public: void func1() { std::cout << "C::func1" << std::endl; } void funcC() { std::cout << "C::funcB" << std::endl; } }; class Holder { boost::shared_ptr<A> aptr; public: void setclass(boost::shared_ptr<A> a) { aptr = a; } void callfunc() { aptr->func1(); } }; int main() { Holder holder; boost::shared_ptr<B> b(new B()); boost::shared_ptr<C> c(new C()); holder.setclass(b); //<------- holder.setclass(c); //<------- holder.callfunc(); return 0; } VC++ 6.0 flags the above lines with a conversion error. Is there any way to resolve this? Thanks very much in advance.

2 3

Can't build Python-lib with VC
by swbuddy2001 08 Dec '01

08 Dec '01

Hi all, i'm trying to build the Boost-Python library (Version 1.26.0) with VS6 using SP6 and i get the following ErrorMsg. Compiling... types.cpp ..\..\..\boost/bind/mem_fn_template.hpp(135) : error C2562: '()' : 'void' function returning a value ..\..\..\boost/bind/mem_fn_template.hpp(133) : see declaration of '()' ..\..\..\boost/bind.hpp(735) : see reference to function template instantiation 'void __thiscall boost::_mfi::cmf1<void,class boost::python::detail::type_object_base,struct _object *>::operator () (class boost::python::detail::type_object_base *const & ,struct _object *) const' being compiled Any suggestions what i can do to get types.cpp compiled? Thanks in advance, Jürgen

2 4

explanation about cyclic_ptr in smart_ptr
by ma_ting_chong 06 Dec '01

06 Dec '01

How do cyclic_ptr work in the smarttest.zip in smart_ptr library?

1 0

Need Digest
by Daryle Walker 05 Dec '01

05 Dec '01

I guess Yahoo Groups had a data burp on 21 Nov 2001, because I didn't get digest 22. Could anyone send it to me? -- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT mac DOT com

1 0