December 2007 - Boost-users - lists.preview.boost.org

[review] Fast track formal review of Forward library begins today
by John Torjo 05 Dec '07

05 Dec '07

Hi all, The formal review of the Boost.Functional/Forward library, proposed by Tobias Schwinger, begins today : Boost.Functional/Forward provides decorator function objects to have an n-ary, generic function object accept both RValues and mutable LValues. An exponential number of overloads is needed with C98, so redundant solutions are highly undesirable. Functional/Forward is a reusable implementation asking to live in a central place in Boost. Download it from here: http://www.boost-consulting.com/vault/index.php?action=downloadfile&filenam… http://www.boost-consulting.com/vault/index.php?&direction=0&order=&directo… The documentation is included in the .zip file, in the usual place: forward/libs/functional/forward/doc/html/index.html What to include in Review Comments ================================== Your comments may be brief or lengthy, but basically the Review Manager needs your evaluation of the library. If you identify problems along the way, please note if they are minor, serious, or showstoppers. Here are some questions you might want to answer in your review: * What is your evaluation of the design? * What is your evaluation of the implementation? * What is your evaluation of the documentation? * What is your evaluation of the potential usefulness of the library? * Did you try to use the library? With what compiler? Did you have any problems? * How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? * Are you knowledgeable about the problem domain? And finally, every review should answer this question: * Do you think the library should be accepted as a Boost library? Be sure to say this explicitly so that your other comments don't obscure your overall opinion. We're looking forward to your evaluation of this Boost candidate! John Torjo - Review Manager - -- http://John.Torjo.com -- C++ expert ... call me only if you want things done right

6 8

Re: [Boost-users] build boost.python tutorial "hello world" without bjam
by damny＠web.de 05 Dec '07

05 Dec '07

thanks for your replay, but nothing happens when i reorder the compile command. like before, the same error messages appear. are there any other ideas? i'm at a loss. daniel > -----Ursprüngliche Nachricht----- > Von: boost-users(a)lists.boost.org > Gesendet: 03.12.07 09:50:12 > An: boost-users(a)lists.boost.org > Betreff: Re: [Boost-users] build boost.python tutorial "hello world" without bjam > > damny(a)web.de wrote: > > hi there, > > > > for some days i try to build the boost.python tutorial "hello world" without bjam on winxp by using mingw. > > so i wrote a *.bat-file: > > > > // --- snip ----------------------------------------------------------------------------- > > @echo off > > SETLOCAL > > SET DIR_MINGW=c:\development_compiler\mingw\ > > SET DIR_BOOST=c:\development_compiler\_boost\ > > SET DIR_PYTHN=c:\development_compiler\python25\ > > SET DIR_PRJCT=%~dp0 > > SET NME_PRJCT=hello > > SET CC=c:\development_compiler\mingw\bin\g++.exe > > SET INCL=-I%DIR_BOOST%include\boost-1_34_1 -I%DIR_PYTHN%include -I%DIR_MINGW%include > > SET LIBS=-L%DIR_BOOST%lib -L%DIR_PYTHN%libs -L%DIR_MINGW%lib > > %CC% -O0 -shared %INCL% %LIBS% -lboost_python-mgw34-s-1_34_1 -lpython25 ^ > > %DIR_PRJCT%%NME_PRJCT%.cpp -o %NME_PRJCT%.pyd > > ENDLOCAL > > // --- snip ----------------------------------------------------------------------------- > > > > this works fine. but nevertheless it produces errors. > > > > // --- snip ----------------------------------------------------------------------------- > > D:\Temp\boost_test>c:\development_compiler\mingw\bin\g++.exe -O0 - shared -Ic:\development_compiler\_boost\include\boost-1_34_1 -Ic:\development_compiler\python25\include -Ic:\development_compiler\mingw\include -Lc:\development_compiler\_boost\lib -Lc:\development_compiler\python25\libs -Lc:\development_compiler\mingw\lib -lboost_python-mgw34-mt-s-1_34_1 -lpython25 D:\Temp\boost_test\hello.cpp -o hello.pyd > > This is just a wild guess, but what happens when you reorder your compile > command, putting the libraries _after_ the source file, like this: > > g++.exe -O0 - shared \ > -Ic:\development_compiler\_boost\include\boost-1_34_1 \ > -Ic:\development_compiler\python25\include \ > -Ic:\development_compiler\mingw\include \ > D:\Temp\boost_test\hello.cpp \ > -Lc:\development_compiler\_boost\lib \ > -Lc:\development_compiler\python25\libs \ > -Lc:\development_compiler\mingw\lib \ > -lboost_python-mgw34-mt-s-1_34_1 -lpython25 \ > -o hello.pyd > > [...] > > Markus > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > _______________________________________________________________________ Jetzt neu! Schützen Sie Ihren PC mit McAfee und WEB.DE. 3 Monate kostenlos testen. http://www.pc-sicherheit.web.de/startseite/?mc=022220

2 1

boost::bind upgrading from 1.33.1 to 1.34.1 problem
by sj＠sjcomp.com 05 Dec '07

05 Dec '07

Hello, I'm upgrading from boost 1.33.1 to 1.34.1 and upgrade does not let me compile the following code: // CRefCounted provides functionality to CRefPtr to have // referenced pointers struct CNode: public CRefCounted { void GetName(){} }; int main(int argc, char **argv){ typedef CRefPtr<CNode> CPNode; typedef std::multimap<std::string, CPNode> TPNodes; TPNodes Children; std::for_each( Children.begin(), Children.end(), boost::bind( &CNode::GetName, // problem in the next line boost::bind(&TPNodes::value_type::second,_1))); return 0; } I get the following messages from MSVC8.0: error C2784: 'T *boost::get_pointer(const boost::shared_ptr<T> &)' : could not deduce template argument for 'const boost::shared_ptr<T> &' from 'const CPNode' error C2784: 'T *boost::get_pointer(const std::auto_ptr<_Ty> &)' : could not deduce template argument for 'const std::auto_ptr<_Ty> &' from 'const CPNode' error C2784: 'T *boost::get_pointer(T *)' : could not deduce template argument for 'T *' from 'const CPNode' If I replace CRefPtr with boost::shared_ptr the code compiles. Hence I assume the problem is with the way I implemented CRefPtr class. Here it is: template<typename T> class CRefPtr { private: T* m_ptr; public: CRefPtr(T* ptr=NULL) :m_ptr(NULL){ *this=ptr; } CRefPtr(const CRefPtr<T>& ptr) :m_ptr(NULL){ *this=ptr; } ~CRefPtr() { if (m_ptr) { m_ptr->DecrementReference(); m_ptr=NULL; } } CRefPtr<T>& operator=(T* ptr) { if(ptr!=m_ptr) { if(m_ptr) { m_ptr->DecrementReference(); } m_ptr=ptr; if(m_ptr) { m_ptr->IncrementReference(); } } return *this; } CRefPtr<T>& operator=(const CRefPtr<T>& ptr){*this=ptr.m_ptr;return *this;} T* operator->() const{return m_ptr;} T& operator*() const{return *m_ptr;} bool operator<(const CRefPtr<T>& obj)const{if(obj.m_ptr<this->m_ptr)return true;else return false;} T* get() const{return m_ptr;} operator T*()const{return m_ptr;} }; What am I missing? Some explanation on why I went this route. I am using CRefPtr instead of shared_ptr because full CNode class has a member CNode* m_Parent. I could not figure out how to use weak pointers, but if I use CRefPtr I can assign a simple pointer returned by GetParent() to a CPNode and it would still work, while I could not achieve this result with shared_ptr. Maybe someone can direct me to a place which explains how to handle such situations properly with shared_ptr. Thanks. -- Regards, Alexander. http://sjcomp.com

3 2

Re: [Boost-users] [asio] how to specify a timeout?
by Andrea Bartoli 05 Dec '07

05 Dec '07

Yes, timeout and read event have two different callback, in read callback you have to m_timer->cancel() to avoid receive timeout also. ----- Messaggio originale ----- Da: Arkadiy Vertleyb <vertleyb(a)hotmail.com> A: boost-users(a)lists.boost.org Inviato: Martedì 4 dicembre 2007, 16:14:40 Oggetto: Re: [Boost-users] [asio] how to specify a timeout? Hello Andrea, "Andrea Bartoli" <bartols2000(a)yahoo.it> wrote > Use an asio::deadline_timer > > Example: > > asio::io_service m_ioService; > asio::deadline_timer * m_timer; > m_timer = new asio::deadline_timer(m_ioService); > > m_timer->expires_from_now( boost::posix_time::seconds(t) ); > m_timer->async_wait( boost::bind(&Foo:handle_timeout, this, > asio::placeholders::error) ); > asio::async_read( *m_socket, > asio::buffer( m_msg.header(), > m_msg.header_length() ), > boost::bind(&Foo::handle_read_header, > this, asio::placeholders::error) ); Thanks for your response. Still, it's not clear to me how to use it. In your example it seems that read operation and timeout are two independent things. How do I make sure only one of these two fires? Is it possile to cancel read operation once timout has been recieved? Thanks, Arkadiy _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users ___________________________________ L'email della prossima generazione? Puoi averla con la nuova Yahoo! Mail: http://it.docs.yahoo.com/nowyoucan.html

2 1

Re: [Boost-users] Building Boost 1.33.0 with bjam --v2 failes underHP-UX 11.23
by Vinokurov, Aleksandr 05 Dec '07

05 Dec '07

"Vladimir Prus" <ghost(a)cs.msu.su> wrote in message news:<fj3pej$14q$1(a)ger.gmane.org>... > Vinokurov, Aleksandr wrote: > > > > > > > Hello all, > > > > I'm trying to build boost 1.33 64-bit libraries under HP-UX 11.23 with the > > help of gcc 4.1.1 thru bjam. I was confused trying to supply -mlp64 > > compiler and linker option to the gcc though bjam options. By default > > calling ``bjam'' uses Boost.Build v.1, and setting ``bjam --v2'' -- will > > do v.2. Uncommenting "using gcc ;" in ./tools/build/v2/ user-config.jam > > file led bjam to "learn" this toolset, but calling ``bjam --v2'' from the > > boost_1_33_0/ folder splashes this: > > > > # bjam --v2 gcc > > ... > > > warning: Python location is not configured > > warning: the Boost.Python library won't be built > > Building Boost.Regex with the optional Unicode/ICU support disabled. > > Please refer to the Boost.Regex documentation for more information > > (and if you don't know what ICU is then you probably don't need it). > > /home/avinokur/boost_1_33_0/tools/build/v2/build/virtual-target.jam:868: > > in virtual-target.register-actual-name from module virtual-target error: > > Duplicate name of actual target: > > <pstage>libboost_date_time-gcc-d-1_33.so.1.33.0 error: previous virtual > > target { > > common%common.copy-libboost_date_time-gcc-d-1_33.so.1.33.0.SHARED_LIB { > > .... > > > Finally the only way to get this 64-bit libs was this hack: ``GXX="g++ > > -mlp64" bjam''. Does anybody knows what the heck is with the bjam --v2 and > > why it is unable to build boost from-the-box? > > Probably an answer with a question is in order -- what's the heck with trying > to build outdated version of Boost, in the first place? Second, Boost.Build > V2 is officially used started with 1.34. Surely, even if had bug in 1.33, > nobody's going to fix that. > > - Volodya > Okay: 1. I have no free time (just now) to spend upgrading my current project from Boost 1.33 to latest one and finding/fixing misplaces and new bugs. 2. If Boost.Build v2 official release was in 1.34, why it was shipped with Boost 1.33 archive? And there are no documentation (looked on the site) on how to force compiler option when using v1 Boost.Build? Sincerely yours, Aleksandr > > > > > > Aleksandr. > > > > > > PS. This is my gcc version, but seems it is not clued anyway: > > > > # gcc -v > > Using built-in specs. > > Target: ia64-hp-hpux11.23 > > Configured with: ../gcc/configure > > Thread model: posix > > gcc version 4.1.1

1 0

Iterator over values of std::map<K, V>
by John Kiopela 04 Dec '07

04 Dec '07

Hi all, I have been browsing through the docs, but I couldn't find anything useful. Does Boost have a std::map<K, V>::iterator adaptor which iterates over the values of the map only. That is, I would like to have an iterator of which operator*() function returns V, instead of std::pair<K, V>. My data is stored in a std::map<K, V> and I would to allow my users to iterate over the values. If it's not here, I'll make one myself. It shouldn't be that hard. Thanks in advance. Kind regards, John

5 8

Building Boost 1.33.0 with bjam --v2 failes under HP-UX 11.23
by Vinokurov, Aleksandr 04 Dec '07

04 Dec '07

Hello all, I'm trying to build boost 1.33 64-bit libraries under HP-UX 11.23 with the help of gcc 4.1.1 thru bjam. I was confused trying to supply -mlp64 compiler and linker option to the gcc though bjam options. By default calling ``bjam'' uses Boost.Build v.1, and setting ``bjam --v2'' -- will do v.2. Uncommenting "using gcc ;" in ./tools/build/v2/ user-config.jam file led bjam to "learn" this toolset, but calling ``bjam --v2'' from the boost_1_33_0/ folder splashes this: # bjam --v2 gcc warning: Python location is not configured warning: the Boost.Python library won't be built Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). /home/avinokur/boost_1_33_0/tools/build/v2/build/virtual-target.jam:868: in virtual-target.register-actual-name from module virtual-target error: Duplicate name of actual target: <pstage>libboost_date_time-gcc-d-1_33.so.1.33.0 error: previous virtual target { common%common.copy-libboost_date_time-gcc-d-1_33.so.1.33.0.SHARED_LIB { gcc%gcc.link.dll-libboost_date_time-gcc-d-1_33.so.1.33.0.SHARED_LIB { gcc%gcc.compile.c++-greg_month.o.OBJ { gregorian/greg_month.cpp.CPP } } { gcc%gcc.compile.c++-greg_weekday.o.OBJ { gregorian/greg_weekday.cpp.CPP } } { gcc%gcc.compile.c++-date_generators.o.OBJ { gregorian/date_generators.cpp.CPP } } } } error: created from ./stage error: another virtual target { common%common.copy-libboost_date_time-gcc-d-1_33.so.1.33.0.SHARED_LIB { common%common.copy-libboost_date_time-gcc-d-1_33.so.1.33.0.SHARED_LIB { gcc%gcc.link.dll-libboost_date_time-gcc-d-1_33.so.1.33.0.SHARED_LIB { gcc%gcc.compile.c++-greg_month.o.OBJ { gregorian/greg_month.cpp.CPP } } { gcc%gcc.compile.c++-greg_weekday.o.OBJ { gregorian/greg_weekday.cpp.CPP } } { gcc%gcc.compile.c++-date_generators.o.OBJ { gregorian/date_generators.cpp.CPP } } } } } error: created from ./stage error: added properties: none error: removed properties: none /home/avinokur/boost_1_33_0/tools/build/v2/build/virtual-target.jam:175: in actualize-no-scanner from module object(file-target)@269 /home/avinokur/boost_1_33_0/tools/build/v2/build/virtual-target.jam:110: in object(file-target)(a)269.actualize from module object(file-target)@269 /home/avinokur/boost_1_33_0/tools/build/v2/build-system.jam:285: in load from module build-system /home/avinokur/boost_1_33_0/tools/build/v2/kernel/modules.jam:259: in import from module modules /home/avinokur/boost_1_33_0/tools/build/v2/kernel/bootstrap.jam:153: in boost-build from module /home/avinokur/boost_1_33_0/boost-build.jam:12: in module scope from module Calling v.1 Boost.Built with "cxxflags=-mlp64" got no luck too: # bjam cxxflags=-mlp64 --------------------------------------------------------------------- *** If you don't need Boost.Python, you can ignore this section *** *** pass --without-python to suppress this message in the future *** skipping Boost.Python library build due to missing or incorrect configuration couldn't find Python.h in "/usr/local/include/python2.2" You can configure the location of your python installation by setting: PYTHON_ROOT - currently "/usr/local" PYTHON_VERSION - The 2-part python Major.Minor version number (e.g. "2.2", NOT "2.2.1") - currently "2.2" The following are automatically configured from PYTHON_ROOT if not otherwise set: PYTHON_LIB_PATH - path to Python library object; currently "/usr/local/lib/python2.2/config" PYTHON_INCLUDES - path to Python #include directories; currently "/usr/local/include/python2.2" --------------------------------------------------------------------- Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). don't know how to make cxxflags=-mlp64 ...found 1 target... ...can't find 1 target... Finally the only way to get this 64-bit libs was this hack: ``GXX="g++ -mlp64" bjam''. Does anybody knows what the heck is with the bjam --v2 and why it is unable to build boost from-the-box? Aleksandr. PS. This is my gcc version, but seems it is not clued anyway: # gcc -v Using built-in specs. Target: ia64-hp-hpux11.23 Configured with: ../gcc/configure Thread model: posix gcc version 4.1.1

3 2

C++ Performance
by nisha kannookadan 04 Dec '07

04 Dec '07

Hi everybody Im a quite new boost user and I wonder, why C++ Program is so slow. I wrote a program, and I have the same program in Matlab, I expected that my C++ version is much faster, but its not. I use matrices, vectors, matrix- and vector proxies. Do some solving and compute calculations with prod, element_div, etc.. Is boost in general slower than matlab, should I use something else? Id be real happy about any help. Regards Nisha K PS: This question was already in the developer forum. _________________________________________________________________ Invite your mail contacts to join your friends list with Windows Live Spaces. It's easy! http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.aspx…

9 22

program_options: Printing description problem
by sj＠sjcomp.com 03 Dec '07

03 Dec '07

Hello, When I try to output options_description to cout (response to --help option, to show all available options) I get an assertion in in xstring header file. It's caused by execution format_paragraph function in option_description.cpp line 381. I narrowed down the problem to the length of the description for an option. This is the code: #include <boost/program_options.hpp> namespace po = boost::program_options; int main(int argc, char **argv){ po::options_description desc("t"); desc.add_options()("h", "1234567890qwertyuiopasdfghjklzxcvbnmQWERTYUIOPASDFGHJKLZ"); po::variables_map vm; po::store(po::parse_command_line(argc, argv, desc), vm); po::notify(vm); if(vm.count("h")){std::cout<<desc<<"\n"; return 1;} return 0; } If I remove last Z in the description of h parameter the program works, otherwise it crashes. If I replace the message with some words, meaning if spaces are present, it still crashes (that's how I originally discovered the problem). What am I doing wrong, or how can I avoid this situation (other than decreasing the length of the descriptions for all of my parameters). I'm using boost 1.33.1 (how can I find out what was done to program_options in 1.34.1?) and Microsoft Visual Studio 2005 Professional. The project is compiled as Multi-threaded DLL Thanks. -- Regards, Alexander. http://sjcomp.com

2 1

Problems with boost::parameter macros &
by StephQ 03 Dec '07

03 Dec '07

I have the following 3 problems: 1) I can not use BOOST_PARAMETER_MEMBER_FUNCTION (and the const version) inside a template class. The following code: #include <boost/parameter.hpp> BOOST_PARAMETER_NAME(mu) BOOST_PARAMETER_NAME(sigma) template<class T1> class mah { private: int t1; int t2; public: BOOST_PARAMETER_MEMBER_FUNCTION( (void), set, tag, (optional (mu, *, 1) (sigma, *, 2) ) ) { t1 = mu; t2 = sigma; } }; int main() { mah<int> chissa; chissa.set( 2, 1 ); } Results in (with MSVC, similar error with gcc 3.4 cygwin): 1>c:\develop\essl_alpha\essl\test\parametertest1.cpp(27) : warning C4346: 'boost::parameter::aux::argument_pack<mah<T1>:: boost_param_params_27set<int>>::type' : dependent name is not a type 1> prefix with 'typename' to indicate a type 1> c:\develop\essl_alpha\essl\test\parametertest1.cpp(32) : see reference to class template instantiation 'mah<T1>' being compiled 1>c:\develop\essl_alpha\essl\test\parametertest1.cpp(27) : error C2923: 'mah<T1>::boost_param_result_27set' : 'boost::parameter::aux::argument_pack<mah<T1>:: boost_param_params_27set<int>>::type' is not a valid template type argument for parameter 'Args' 1>c:\develop\essl_alpha\essl\test\parametertest1.cpp(27) : warning C4346: 'mah<T1>::boost_param_result_27set::type' : dependent name is not a type 1> prefix with 'typename' to indicate a type 1>c:\develop\essl_alpha\essl\test\parametertest1.cpp(27) : error C2146: syntax error : missing ';' before identifier 'set' 1>c:\develop\essl_alpha\essl\test\parametertest1.cpp(27) : error C4430: missing type specifier - int assumed. Note: C++ does not support default-int 1>c:\develop\essl_alpha\essl\test\parametertest1.cpp(27) : warning C4183: 'set': missing return type; assumed to be a member function returning 'int' 2) I would like to create a class template for a "named tuple", something like: named_tuple<int,double,mu,sigma> n; // mu, sigma previously declared, maybe BOOST_PARAMETER_NAME? // should contain two variables of type int, double. Would be really nice if they are named exactly _mu and _sigma, so I need boost PP or something? n.set( _mu=1, _sigma=2.0 ); // set internal variables n.set( _mu=1 ); double x = n.get( _sigma ); n.get( _mu, _sigma ); // This returns a tuple with the internal values. I think a class like this could be really usefull, and it would be really nice to see tuples get this extension. IMHO they are very clear from a user point of view. The problem (apart from problem 1) is that the macro BOOST_PARAMETER_MEMBER_FUNCTION does not play nicely if I just put a template parameter (say N1 of named_tuple<T1,T2,N1,N2>) instead of mu in BOOST_PARAMETER_MEMBER_FUNCTION( (void), set, tag, (optional (mu, *, 1) (sigma, *, 2) ) Again I suspect I need boost PP or use directly ArgumentPack. 3) How should I consider the overloading of a BOOST_PARAMETER_MEMBER_FUNCTION? Is it fine to just write the following? class mah2 : public mah // mah as before but with no templates. { public: template<class A1, class A2> void set( A1 a1, A2 a2 ) // need to supply defaults here... { mah<int,int>::set( a1, a2 ); } }; As usual any help is extremely appreciated. Thank you in advance. Best Regards StephQ

1 0