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

It's there a pitfall in boost::regex?
by Phipps Xue 17 Apr '03

17 Apr '03

Good morning all, I used regex library which is shipped with boost 1.30 in my project. But I found in the case I list below, the regex_match() will crash. Is there a pitfall in it? ===================== CODE LIST BEGIN ========================= #include "stdafx.h" #include <afxconv.h> #include <boost/regex.hpp> using namespace boost; void do_regex_match(const string& s) { USES_CONVERSION; CString _c1,_c2; _c1 = "\\_gl_glcf]a:(D_N~yOA_gg_aeN_o/>jm_gf'0(m_--\\6L_gi_AE_gd5<&0A~_E'_Y-'-,uO7_so_glTc G_gl_C,_o/*6hC=@_u"; // Failed _c2 = "\\_gl_glcf]a:(D_N~yOA_gg_aeN_o/jm_gf'0(m_--\\6L_gi_AE_gd5<&0A~_E'_Y-'-,uO7_so_glTc G_gl_C,_o/*6hC=@_u"; // Passed wstring ws1(L""),ws2(L""); ws1=A2W(_c1); ws2=A2W(_c2); wstring wsMatchString(L""); wsMatchString=A2W(".*(要)?.*天气.*"); boost::wregex e(wsMatchString/*,boost::regbase::normal|boost::regbase::icase*/); boost::regex_match(ws2,e);// Pass boost::regex_match(ws1,e);// Crash } ======================== CODE LIST END ========================= As you can see in the code, if any change is made to _c1, the test will pass or it will crash. I will list my compiler and system information below. Compiler: Microsoft Visual C++ 6.0 Professionall with SP5 OS: Microsoft Windows 2000 Professional with SP3 RAM: 512M CPU: PIII 700 Any help will be appreciated. Phipps Xue Software Engineer, MCP/MCSD Email: xuejm(a)sina.com

2 1

Powers
by basicxml2011 17 Apr '03

17 Apr '03

Is their any library or function that can raise a number to a power?

1 0

RE: [Boost-Users] Re: command line parsing
by Ben Hutchings 17 Apr '03

17 Apr '03

Vladimir Prus wrote: > cl_corba_at wrote: <snip> > >> Frankly, I've missed the WinMain case, and not sure what to do. > >> The problem is that single string might erase important > >> information --- what happens if you have command line argument > >> with embedded space. > > > > Embedded spaces should be placed in quotation marks which should be > > removed by the tokenizer. Arguments could be seperated by one ore more > > space. > > Do you mean that if I type > > program "a b c" "C:/Program files" > > then the program will receive this string, with quotes there? > (The linux shell will strip quotes completely). That is correct. Windows programs receive given a single string of arguments to parse whereas Unix programs receive a vector which can be passed to main unchanged by the startup code. <snip> > Is this solution is OK? (The only problem is that adjuacent spaces are > mishandled, but that's fixable). A full solution will need to be a little bit more complicated. There isn't a specification of how Windows command-lines should be generated or interpreted from an array of strings, but it seems to be sensible to interpret them in the same way as Microsoft's run-time library does it: * Outside a quoted section, a double-quote begins a quoted section. * Outside a quoted section, a string of one or more spaces is a separator if there are non-space characters both before and after it; otherwise it's just padding. * Inside a quoted section, a double-quote preceded by an odd number of backslashes (call this n) represents (n-1)/2 backslashes followed by a double-quote. * Inside a quoted section, a double-quote preceded by an even number of backslashes (possibly zero; call this n) represents n/2 backslashes and ends the quoted section. * All other characters represent themselves. Note that double-quotes are not separators and backslashes are not usually escape characters. Also note that a quoted section does not have to be terminated. tokenizer is probably not up to this job. You could probably use regular expressions, but a custom state machine might be the best solution.

2 1

Converting boost::shared_ptr< const T > to boost::shared_ptr< T >
by Alex Tibbles 17 Apr '03

17 Apr '03

Apologies, news editor != text editor (cntrl-enter != autocomplete). I have persistent objects in my appplication and their factory dishes out only constant references (more precisely only copies of a shared_ptr< const T >). In order to make the persistent objects writable, an underlying record needs to be locked in the RDBMS. A friend template instantiation (a WritablePointer< T >) controls this locking, and effectively gives non-const access to the persistent object. The WritablePointer's constructor takes a shared_ptr< const T >. What I want to be able to do is like this (with non-smart pointers): template< typename T > class WritablePointer { public: WritablePointer(const T* const p) : p_(const_cast< T* const >(p)) { p_->lock(); } // ... private: T *p_; }; I have: template< typename T > class WritablePointer { public: typedef typename boost::shared_ptr< T > SmartPointerType; typedef typename boost::shared_ptr< const T > ConstSmartPointerType; WritablePointer(const ConstSmartPointerType& p) : p_(static_cast< const SmartPointerType& >(p)) { p_->lock(); } private: SmartPointerType p_; }; This compiles under gcc 3.2.1 but with a warning: /usr/include/boost/shared_ptr.hpp:153: warning: invalid conversion from `const T* const' to `T*' /usr/include/boost/shared_ptr.hpp: In constructor [I have changed the concrete type from my application to T to simplify]. What I want to do is cast away the const-ness in exactly one place - the WritablePointer ctor. But because I get the warnings it's harder to spot truly illicit conversions. (I copied this technique from an article I read (on Dr. Dobb's or CUJ probably) which used volatile and a mutex to protect shared data between threads, and a smart pointer much like mine to provide write access. I'd be grateful for any suggestions. Alex

2 2

boost:format Problems.
by Bernhard Glück 16 Apr '03

16 Apr '03

Hi ! Today my company switched over to MSVC 1310 ( Visual Studio .NET 2003 ) We make heavy use of some boost libraries. And now we have a real showstopper, boost::format does nto compile anymore without errors. Are there any fixes for that ? we use boost_1_30_0 built using the VC7 toolset ( from MSVC 7.1 ) Here is the code passage that causes errors ( but many others do the same ) // Generate a unique dummy name. boost::format uniqueName("Unknown%1%"); int idNumber = sNames.size() + sIdCounter; // Just an integer. uniqueName % idNumber; Here are the compiler errors: d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2065: 'disambiguater' : undeclared identifier d:\Sdk\boost\boost\format\feed_args.hpp(140) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put_head<char,std::char_traits<char>,T>(std::basic_ostream<_Elem,_Traits> &,const T &)' being compiled with [ T=int, _Elem=char, _Traits=std::char_traits<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(237) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put<Ch,Tr,T>(T,const boost::io::detail::format_item<Ch,Tr> &,std::basic_string<_Elem,_Traits,_Ax> &,std::basic_ostringstream<_Elem,_Traits,_Alloc> &)' being compiled with [ Ch=boost::basic_format<char>::CharT, Tr=boost::basic_format<char>::Traits, T=const int &, _Elem=char, _Traits=std::char_traits<char>, _Ax=std::allocator<char>, _Alloc=std::allocator<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(246) : see reference to function template instantiation 'void boost::io::detail::distribute<Ch,Tr,T>(boost::basic_format<Ch> &,T)' being compiled with [ Ch=boost::basic_format<char>::CharT, Tr=boost::basic_format<char>::Traits, T=const int & ] d:\Sdk\boost\boost\format\format_class.hpp(63) : see reference to function template instantiation 'boost::basic_format<Ch> &boost::io::detail::feed<boost::basic_format<Ch>::CharT,boost::basic_format<Ch>::Traits,const T&>(boost::basic_format<Ch> &,const int)' being compiled with [ Ch=char, T=int ] d:\Work\FakedReality\Source\Core\Engine\Identifier.cpp(87) : see reference to function template instantiation 'boost::basic_format<Ch> &boost::basic_format<Ch>::operator %<int>(const T &)' being compiled with [ Ch=char, T=int ] d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2275: 'Ch' : illegal use of this type as an expression d:\Sdk\boost\boost\format\feed_args.hpp(140) : see declaration of 'Ch' d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2039: 'put_head' : is not a member of 'operator``global namespace''' d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2780: 'void boost::io::detail::`anonymous-namespace'::put_head(std::basic_ostream<Ch,Tr> &,const T &)' : expects 2 arguments - 3 provided d:\Sdk\boost\boost\format\feed_args.hpp(78) : see declaration of 'boost::io::detail::`anonymous-namespace'::put_head' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2275: 'Ch' : illegal use of this type as an expression d:\Sdk\boost\boost\format\feed_args.hpp(155) : see declaration of 'Ch' d:\Sdk\boost\boost\format\feed_args.hpp(155) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put_last<char,std::char_traits<char>,T>(std::basic_ostream<_Elem,_Traits> &,const T &)' being compiled with [ T=int, _Elem=char, _Traits=std::char_traits<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2039: 'put_last' : is not a member of 'operator``global namespace''' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2780: 'void boost::io::detail::`anonymous-namespace'::put_last(std::basic_ostream<Ch,Tr> &,const T &)' : expects 2 arguments - 3 provided d:\Sdk\boost\boost\format\feed_args.hpp(82) : see declaration of 'boost::io::detail::`anonymous-namespace'::put_last' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C3861: 'disambiguater': identifier not found, even with argument-dependent lookup Framework.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled Actor.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled Log.cpp d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2065: 'disambiguater' : undeclared identifier d:\Sdk\boost\boost\format\feed_args.hpp(140) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put_head<char,std::char_traits<char>,T>(std::basic_ostream<_Elem,_Traits> &,const T &)' being compiled with [ T=int, _Elem=char, _Traits=std::char_traits<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(237) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put<Ch,Tr,T>(T,const boost::io::detail::format_item<Ch,Tr> &,std::basic_string<_Elem,_Traits,_Ax> &,std::basic_ostringstream<_Elem,_Traits,_Alloc> &)' being compiled with [ Ch=boost::basic_format<char>::CharT, Tr=boost::basic_format<char>::Traits, T=const int &, _Elem=char, _Traits=std::char_traits<char>, _Ax=std::allocator<char>, _Alloc=std::allocator<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(246) : see reference to function template instantiation 'void boost::io::detail::distribute<Ch,Tr,T>(boost::basic_format<Ch> &,T)' being compiled with [ Ch=boost::basic_format<char>::CharT, Tr=boost::basic_format<char>::Traits, T=const int & ] d:\Sdk\boost\boost\format\format_class.hpp(63) : see reference to function template instantiation 'boost::basic_format<Ch> &boost::io::detail::feed<boost::basic_format<Ch>::CharT,boost::basic_format<Ch>::Traits,const T&>(boost::basic_format<Ch> &,const int)' being compiled with [ Ch=char, T=int ] d:\Work\FakedReality\Source\Core\Debug\Log.cpp(114) : see reference to function template instantiation 'boost::basic_format<Ch> &boost::basic_format<Ch>::operator %<int>(const T &)' being compiled with [ Ch=char, T=int ] d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2275: 'Ch' : illegal use of this type as an expression d:\Sdk\boost\boost\format\feed_args.hpp(140) : see declaration of 'Ch' d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2039: 'put_head' : is not a member of 'operator``global namespace''' d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2780: 'void boost::io::detail::`anonymous-namespace'::put_head(std::basic_ostream<Ch,Tr> &,const T &)' : expects 2 arguments - 3 provided d:\Sdk\boost\boost\format\feed_args.hpp(78) : see declaration of 'boost::io::detail::`anonymous-namespace'::put_head' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2275: 'Ch' : illegal use of this type as an expression d:\Sdk\boost\boost\format\feed_args.hpp(155) : see declaration of 'Ch' d:\Sdk\boost\boost\format\feed_args.hpp(155) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put_last<char,std::char_traits<char>,T>(std::basic_ostream<_Elem,_Traits> &,const T &)' being compiled with [ T=int, _Elem=char, _Traits=std::char_traits<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2039: 'put_last' : is not a member of 'operator``global namespace''' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2780: 'void boost::io::detail::`anonymous-namespace'::put_last(std::basic_ostream<Ch,Tr> &,const T &)' : expects 2 arguments - 3 provided d:\Sdk\boost\boost\format\feed_args.hpp(82) : see declaration of 'boost::io::detail::`anonymous-namespace'::put_last' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C3861: 'disambiguater': identifier not found, even with argument-dependent lookup State.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled d:\Sdk\boost\boost\iterator_adaptors.hpp(1409) : see reference to class template instantiation 'boost::filter_iterator_generator<Predicate,Iterator,Value,Reference,Pointer,Category,Distance>' being compiled d:\Work\FakedReality\Source\Core\Scripting\State.cpp(303) : error C3861: 'lua_dostring': identifier not found, even with argument-dependent lookup Script.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled IScriptExport.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled Type.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled IConverter.cpp d:\Sdk\boost\luabind\object.hpp(153) : error C2027: use of undefined type 'luabind::object' d:\Sdk\boost\luabind\object.hpp(90) : see declaration of 'luabind::object' d:\Sdk\boost\luabind\object.hpp(177) : see reference to class template instantiation 'luabind::detail::proxy_caller<Tuple>' being compiled Version.cpp d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2065: 'disambiguater' : undeclared identifier d:\Sdk\boost\boost\format\feed_args.hpp(140) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put_head<char,std::char_traits<char>,T>(std::basic_ostream<_Elem,_Traits> &,const T &)' being compiled with [ T=int, _Elem=char, _Traits=std::char_traits<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(237) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put<Ch,Tr,T>(T,const boost::io::detail::format_item<Ch,Tr> &,std::basic_string<_Elem,_Traits,_Ax> &,std::basic_ostringstream<_Elem,_Traits,_Alloc> &)' being compiled with [ Ch=boost::basic_format<char>::CharT, Tr=boost::basic_format<char>::Traits, T=const int &, _Elem=char, _Traits=std::char_traits<char>, _Ax=std::allocator<char>, _Alloc=std::allocator<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(246) : see reference to function template instantiation 'void boost::io::detail::distribute<Ch,Tr,T>(boost::basic_format<Ch> &,T)' being compiled with [ Ch=boost::basic_format<char>::CharT, Tr=boost::basic_format<char>::Traits, T=const int & ] d:\Sdk\boost\boost\format\format_class.hpp(63) : see reference to function template instantiation 'boost::basic_format<Ch> &boost::io::detail::feed<boost::basic_format<Ch>::CharT,boost::basic_format<Ch>::Traits,const T&>(boost::basic_format<Ch> &,const int)' being compiled with [ Ch=char, T=int ] d:\Work\FakedReality\Source\Core\Platform\Win32\Version.cpp(38) : see reference to function template instantiation 'boost::basic_format<Ch> &boost::basic_format<Ch>::operator %<int>(const T &)' being compiled with [ Ch=char, T=int ] d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2275: 'Ch' : illegal use of this type as an expression d:\Sdk\boost\boost\format\feed_args.hpp(140) : see declaration of 'Ch' d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2039: 'put_head' : is not a member of 'operator``global namespace''' d:\Sdk\boost\boost\format\feed_args.hpp(79) : error C2780: 'void boost::io::detail::`anonymous-namespace'::put_head(std::basic_ostream<Ch,Tr> &,const T &)' : expects 2 arguments - 3 provided d:\Sdk\boost\boost\format\feed_args.hpp(78) : see declaration of 'boost::io::detail::`anonymous-namespace'::put_head' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2275: 'Ch' : illegal use of this type as an expression d:\Sdk\boost\boost\format\feed_args.hpp(155) : see declaration of 'Ch' d:\Sdk\boost\boost\format\feed_args.hpp(155) : see reference to function template instantiation 'void boost::io::detail::`anonymous-namespace'::put_last<char,std::char_traits<char>,T>(std::basic_ostream<_Elem,_Traits> &,const T &)' being compiled with [ T=int, _Elem=char, _Traits=std::char_traits<char> ] d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2039: 'put_last' : is not a member of 'operator``global namespace''' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C2780: 'void boost::io::detail::`anonymous-namespace'::put_last(std::basic_ostream<Ch,Tr> &,const T &)' : expects 2 arguments - 3 provided d:\Sdk\boost\boost\format\feed_args.hpp(82) : see declaration of 'boost::io::detail::`anonymous-namespace'::put_last' d:\Sdk\boost\boost\format\feed_args.hpp(83) : error C3861: 'disambiguater': identifier not found, even with argument-dependent lookup Generating Code...

4 4

uBlas Lapack binds - part 2
by geert_ceuppens 16 Apr '03

16 Apr '03

I realise that the binding will need quite some work to get right under VC6 (see previous posting). I would welcome suggestions to do the conversion directly (so without all the traits stuff) and not as general. Goal would be to quickly start and then (hopefully) one day be able to migrate to the traits version. Maybe the getrf function in Lapack would be a good starting point? Thanks ... By the way Toon, I think there is a typo in lapack.hpp: template < typename matrix_type, typename IpivIterator > int getrf(matrix_type& a, IpivIterator begin_ipiv, IpivIterator end_ipiv) { using namespace boost::numeric::bindings::traits ; typedef typename matrix_type::value_type value_type ; typedef typename boost::numeric::bindings::traits::value_traits< value_type >::value_type bind_type ; int m = boost::numeric::bindings::traits::matrix_size1( a ) ; int n = boost::numeric::bindings::traits::matrix_matrix_size2( a ) ; Shouldn't matrix_matrix_size2 be matrix_size2? (I can't compile so I did a visual compilation ;^) ... Rgds Geert

3 2

big int library
by Vaclav Vesely 16 Apr '03

16 Apr '03

Are there any plans about adding a big int library to the Boost? There are at least two implementations in the Sandbox. Thanks, w

2 1

command line parsing
by cl_corba_at 16 Apr '03

16 Apr '03

What is the "boost way" of parsing command line arguments? When I searched the archives, some user recommended spirit for this task, but this seems to heavy-weighted for me. I tried tokenizer, but a seperator for the command line is not provided. - you have to remove quotations marks (only escaped_list_separator can do this) - your have to ignore multiple tokens (only char_seperator can do this) If I write this seperator its still wouldn't provide the api I would like to see for a command line parser. I saw in the past there have been an attemp by Bill Kempf to provide a command line parser for boost but it didn't find its way into the library from reasons I don't know. So is there a recommended way to parse options I did overlook? Is a command line parser planned? Thanks for your answers, CL

5 8

GNU istream::scan
by Pat McCormack 15 Apr '03

15 Apr '03

Hi Is there anything like the GNU extension istream::scan() (to use a scanf style format string) in the Boost libraries? The boost format library seems designed to work solely on output streams (am I right on this?) Thanks Pat

1 0

RE: [Boost-Users] Boost weak_ptr<>
by bjorn.karlsson＠readsoft.com 15 Apr '03

15 Apr '03

> From: Edward Diener [mailto:eddielee@tropicsoft.com] > > I don't understand what the practical use of Boost weak_ptr<> > is ? Anyone > like to clue me in ? I understand that Boost weak_ptr<> > creates an internal > reference to a Boost shared_ptr<> but I don't understand what that is > supposed to accomplish. A weak_ptr is an observer of the pointee, i.e it doesn't take part of the lifetime management of the resource. One use for it is to break cyclic dependencies; another is to avoid dangling pointers. When a shared_ptr deallocates its pointee, it notifies the weak_ptrs. When the weak_ptr is subsequently used, it knows that the resource is gone (compare this with storing raw pointers, where there is no way to tell whether the pointer is still valid or not). Example: A Company class has a container that holds elements of type shared_ptr<Employee>. Now, the Supervisor class needs to know things about certain Employees, but it cannot/must not control their lifetime (in the context of the Company class, that is). Rather than asking the Company to search through each Employee (for example by name) each time Supervisor needs to access/update some Employee information, it caches information in weak_ptr<Employee> for efficient access. If an Employee quits (is removed from the container of shared_ptr<Employee>), the Supervisor class will be informed of this the next time it tries to access that weak_ptr<Employee>. At the time of access, a (temporary) shared_ptr is created from a weak_ptr, to ensure that the pointee isn't deleted in the midst of using it. Without weak_ptr, the above would be hard(er) to accomplish: * Asking Company to find an Employee each time it is needed may be too inefficient. * Storing shared_ptr<Employee> in the Supervisor class interferes with lifetime management and can cause inconsistent state. * Storing a raw pointer to an Employee is extremely dangerous (known as the dangling Employee problem); there's no way of telling if the pointer is valid. Bjorn Karlsson

2 1

shard_ptr
by Christian Gmeiner 15 Apr '03

15 Apr '03

Hi! I want to use boost's shared_ptr in my project, which is a dll. At the moment it looks like this: in the h: extern MY_API CCPUDetect * gCPU; in the cpp: CCPUDetect * gCPU = CCPUDetect::GetInstance(); I have problems to make the same with boost, so that gCPU is a smartpointer. Thanks, Christian Gmeiner [Non-text portions of this message have been removed]

2 1

Boost weak_ptr<>
by Edward Diener 15 Apr '03

15 Apr '03

I don't understand what the practical use of Boost weak_ptr<> is ? Anyone like to clue me in ? I understand that Boost weak_ptr<> creates an internal reference to a Boost shared_ptr<> but I don't understand what that is supposed to accomplish.

2 1

boost::date_time don't work under Kylix
by Konstantin M. Litvinenko 15 Apr '03

15 Apr '03

Hello, All! #include <boost/date_time/posix_time/posix_time.hpp> #include <iostream> int main() { using namespace boost::posix_time; using namespace boost::gregorian; date d(1990,1,1); ptime t(d); std::cout << to_simple_string(t); return 0; } I have tested this sample under 4 compilers - vc7, gcc 2.96 asplinux, builder 5.6.4 and kylix3 (compiler borland 5.6). Under Kylix it produce floating point error. I figure out that this happens because t, after this line ptime t(date(1990,1,1)); have {0} (as i see in debuger). Is this library or kylix bug? It seems to be a __int64 arithmetic error but i am not sure. With best regards, Konstantin M. Litvinenko aka Dark Angel. E-mail: darkangel(a)malva.ua

2 2

Question about SPirit parser terminology
by satya_vajja 14 Apr '03

14 Apr '03

Can anyone explain me what is the difference between a rule, grammar and parser in SPIRIT parser library. Thanks

2 1

Macintosh Carbon Thread Error with Boost 1.30.0
by Bobby Thomale 14 Apr '03

14 Apr '03

Hey guys, I just tried to get the new Boost 1.30.0 release to work on MacOS, using TARGET_CARBON (with threads turned on). I have a problem to report: there's a new "lightweight mutex" being used by the smart pointer library. There does not seem to be a lightweight mutex implementation for the "Carbon MP tasks" thread implementation. So - when I compile my code under Boost 1.30 I end up at: # error Unrecognized threading platform in lightweight_mutex.hpp when it tries to compile any of the smart pointer stuff. Help! :-) -- Bobby --------------------------------------------------------------------- Bobby Thomale Senior Software Developer Inoveon Corporation http://www.inoveon.com/ ---------------------------------------------------------------------

2 3

uBlas Lapack binds from Sandbox - pb with templ. specialisation
by geert_ceuppens 14 Apr '03

14 Apr '03

Toon & Kresimir, I've downloaded the lapack and blas binds from the boost sandbox, and I found out the hard way (again :^( ...) the 'features' of the VC6 compiler, i.e. the lack of support for template specialisation. The offending code is: (line nbr may differ due to versioning) *) boost\numeric\bindings\traits\algorithm.hpp(45): template < typename T > struct is_equal vs: template < typename T > struct is_equal< boost::numeric::ublas::matrix< T > > --> template class has already been defined as a non-template class *) libs\numeric\bindings\blas\test\blas.hpp(51): template < typename T > struct assign_multiplier vs: template < typename T > struct assign_multiplier< std::complex< T > > *) boost\numeric\bindings\traits\traits.hpp(46): template <typename V, typename T = typename V::value_type> struct default_vector_traits vs: template <typename V, typename T> struct default_vector_traits<V const, T> *) boost\numeric\bindings\traits\traits.hpp(112 & 114): This one gave more errors: template <typename MatrixT> inline typename matrix_traits<MatrixT>::pointer matrix_storage (MatrixT& m) { return matrix_traits<MatrixT>::storage (m); } error C2039: 'pointer' : is not a member of 'matrix_traits<M>' error C2146: syntax error : missing ';' before identifier 'matrix_storage' error C2433: 'pointer' : 'inline' not permitted on data declarations error C2501: 'pointer' : missing storage-class or type specifiers error C2059: syntax error : ';' error C2065: 'MatrixT' : undeclared identifier error C2065: 'm' : undeclared identifier 'matrix_storage' : missing storage-class or type specifiers syntax error : missing ';' before '<end Parse>' Any chance of providing non-specialised versions of these so I can use the VC6 compiler? Thanks - Geert

1 0

RE: [Boost-Users] command line parsing
by Jim.Hyslop 14 Apr '03

14 Apr '03

Joe Mariadassou [mailto:joemar@vcisnet.net] wrote: > The URL appears to be invalid. > joe > > -----Original Message----- > From: HartmutKaiser(a)t-online.de [mailto:HartmutKaiser@t-online.de] > Sent: Monday, April 14, 2003 3:04 AM > To: Boost-Users(a)yahoogroups.com > Subject: RE: [Boost-Users] command line parsing > > > As far as I remember Vladimir Prus has requested Boost review for his > program_options library. You can find it here: > > http://zigzag.cs.msu.su:7813/program_options > Works OK for me - maybe the problem is temporary. -- Jim

1 0

test
by iad929 13 Apr '03

13 Apr '03

plz ignore. Mohammed

1 0

RE: [Boost-Users] Unary predicate for shared_ptr == 'real' pointe r?
by dick.bridges＠tais.com 11 Apr '03

11 Apr '03

Thanks! In a sense, I can sympathize with your co-workers. I spent hours trying to get that predicate right - I kept screwing up the 'inner' bind. I keep hoping the more I work with the boost libraries the easier it will get. In my case, that's yet to be seen. ;) Thanks again. Tom Matelich <tmatelich@zetec To: "'Boost-Users(a)yahoogroups.com'" <Boost-Users(a)yahoogroups.com> .com> cc: Subject: RE: [Boost-Users] Unary predicate for shared_ptr == 'real' pointe r? 04/11/2003 01:55 PM Please respond to Boost-Users > From: Mark Storer [mailto:mstorer@cardiff.com] > Sent: Friday, April 11, 2003 1:51 PM > > The whole "new int" thing seems a little odd, but that's an > artifact of creating example code. precisely > What do your co-workers say? I feel like I'm missing something here. While the elegance of the double bind is appreciated, I've heard complaints about it being confusing. ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. ----------------------------------------------------------------------- Info: <http://www.boost.org> Wiki: <http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl> Unsubscribe: <mailto:boost-users-unsubscribe@yahoogroups.com> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

1 0

RE: [Boost-Users] Unary predicate for shared_ptr == 'real' pointe r?
by Tom Matelich 11 Apr '03

11 Apr '03

> From: Mark Storer [mailto:mstorer@cardiff.com] > Sent: Friday, April 11, 2003 1:51 PM > > The whole "new int" thing seems a little odd, but that's an > artifact of creating example code. precisely > What do your co-workers say? I feel like I'm missing something here. While the elegance of the double bind is appreciated, I've heard complaints about it being confusing. ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. -----------------------------------------------------------------------

1 0

RE: [Boost-Users] Unary predicate for shared_ptr == 'real' pointe r?
by Mark Storer 11 Apr '03

11 Apr '03

The whole "new int" thing seems a little odd, but that's an artifact of creating example code. What do your co-workers say? I feel like I'm missing something here. --Mark Storer Software Engineer Cardiff Software #include <disclaimer> typedef std::disclaimer<Cardiff> Discard;

1 0

RE: [Boost-Users] Unary predicate for shared_ptr == 'real' pointe r?
by Tom Matelich 11 Apr '03

11 Apr '03

> From: dick.bridges(a)tais.com [mailto:dick.bridges@tais.com] > Sent: Friday, April 11, 2003 11:54 AM > > This is probably more of a C++ than an boost question but, since it > involves shared_ptr, I thought I'd try it here first. I've > got a vector of > shared_ptrS (e.g., vector<shared_ptr<T> > v;) and I need to > find the entry > given the 'real' pointer to the underlying T. Looping thru > the vector like > this: > > <snippet> > for( vector<shared_ptr<T>::iterator itr = v.begin(); itr != > v.end(); itr++) > { > if( itr->get() == real_this){...} > } > </snippet> > > works but seems klunky. I should be able to use [find_if( > v.begin(), v.end > (), ???)] but I can't work out the predicate. Any help or > "pointers" ;) > will be greatly appreciated. > Here's how I like to do it, though my co-workers sometimes complain: #include "boost/shared_ptr.hpp" #include "boost/bind.hpp" #include <vector> #include <iostream> int main() { typedef std::vector<boost::shared_ptr<int> > V; V vec; int* foo = new int; vec.push_back(boost::shared_ptr<int>(foo)); V::iterator iter = std::find_if(vec.begin(), vec.end(), boost::bind(std::equal_to<int*>(), boost::bind(&boost::shared_ptr<int>::get, _1), foo)); if(iter != vec.end()) std::cout << "Life is good" << std::endl; //don't delete foo!!! return 0; } Tom ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. -----------------------------------------------------------------------

1 0

Unary predicate for shared_ptr == 'real' pointer?
by dick.bridges＠tais.com 11 Apr '03

11 Apr '03

This is probably more of a C++ than an boost question but, since it involves shared_ptr, I thought I'd try it here first. I've got a vector of shared_ptrS (e.g., vector<shared_ptr<T> > v;) and I need to find the entry given the 'real' pointer to the underlying T. Looping thru the vector like this: <snippet> for( vector<shared_ptr<T>::iterator itr = v.begin(); itr != v.end(); itr++) { if( itr->get() == real_this){...} } </snippet> works but seems klunky. I should be able to use [find_if( v.begin(), v.end (), ???)] but I can't work out the predicate. Any help or "pointers" ;) will be greatly appreciated. TIA

1 0

RE: [Boost-Users] Boost random's min < max assert
by Glen Knowles 11 Apr '03

11 Apr '03

From: Jens Maurer >> Uniform int contains an assert if min < max. If I ask for a random between 1 >> and 1 I'm expecting a 1, not an assert. Is there some reason this was deemed >> inappropriate? > >When I wrote that code, I never thought somebody would want to generate >a constant 1 as a random number. I agree that this looks non-orthogonal >on first sight. I'll look into it. In my specific case I have a number of objects determined at run time and am randomly picking a number between one and the number of objects in order to select one. The workaround is to simply test for one before the call, but the assert was unexpected. Thanks, Glen [Non-text portions of this message have been removed]

1 0

RE: [Boost-Users] Problem with boost::pool_alloc
by scleary＠jerviswebb.com 11 Apr '03

11 Apr '03

> -----Original Message----- > From: Ilya Chvetsov(work) [mailto:chvetsov@kdlab.com] > > Hi! > I try to use boost::pool_alloc with STLport-4.5-0119 snapshot. > STLPort call pool_alloc::allocate(const size_type n) with n = 0. > STLPort naitive allocator returns in such case NULL, but pool_alloc > tries return memory. Hmmm... You seem to have uncovered a corner case, not covered by the Standard requirements. I will change the pool_allocator (and fast_pool_allocator) to be able to handle this by returning NULL. > Method simple_segregated_storage<SizeType>::try_malloc_n( > void * & start, size_type n, const size_type partition_size) has > problem too if parameter n is 0. Problem is in while condition: > while (--n != 0)... This is correct; (n > 0) is a precondition for that function. -Steve

1 0