- Boost - lists.boost.org

Re: [boost] [regex]using boost::regex with STLport 5.0 on VS.net
by Wang Weiwei 26 Jan '06

26 Jan '06

sorry, the error msg is: BoostTest.obj : error LNK2019: missing symbol "void __cdecl boost::re_detail::raise_runtime_error(class _STL::runtime_error const &)" (?raise_runtime_error@re_detail@boost@@YAXABVruntime_error@_STL@@@Z) ，referenced in "void __cdecl boost::re_detail::raise_error<struct boost::regex_traits_wrapper<struct boost::regex_traits<char,class boost::w32_regex_traits<char> > > >(struct boost::regex_traits_wrapper<struct boost::regex_traits<char,class boost::w32_regex_traits<char> > > const &,enum boost::regex_constants::error_type)" (??$raise_error@U?$regex_traits_wrapper@U?$regex_traits@DV?$w32_regex_traits@D@boost@@@boost@@@boost@@@re_detail@boost@@YAXABU?$regex_traits_wrapper@U?$regex_traits@DV?$w32_regex_traits@D@boost@@@boost@@@1@W4error_type@regex_constants@1@@Z) BoostTest.obj : error LNK2019: missing symbol "class _STL::basic_string<char,class _STL::char_traits<char>,class _STL::allocator<char> > __cdecl boost::re_detail::w32_transform(unsigned long,char const *,char const *)" (?w32_transform@re_detail@boost@@YA?AV?$basic_string@DV?$char_traits@D@_STL@@V?$allocator@D@2@@_STL@@KPBD0@Z) ，referenced in "public: class _STL::basic_string<char,class _STL::char_traits<char>,class _STL::allocator<char> > __thiscall boost::w32_regex_traits<char>::transform(char const *,char const *)const " (?transform@?$w32_regex_traits@D@boost@@QBE?AV?$basic_string@DV?$char_traits@D@_STL@@V?$allocator@D@2@@_STL@@PBD0@Z) Thanks.

2 1

[octonion] Borland patch
by AlisdairM 26 Jan '06

26 Jan '06

The following patch has beeen tested with the new Borland compiler only, but looks like it should appy for earlier versions too: cvs diff -wb -- boost\math\octonion.hpp (in directory E:\sourceforge\devel\boost\) Index: boost/math/octonion.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/math/octonion.hpp,v retrieving revision 1.14 diff -w -b -r1.14 octonion.hpp 47c47 < return(octonion(static_cast<type>(0),b,c,d,e,f,g,h)); \ --- > return( octonion<type>(static_cast<type>(0),b,c,d,e,f,g,h)); \ Note that for once the patch is NOT supplied with a BOOST_WORKAROUND, as it is part of a large macro already, and we can't sneak #if statements into a macro ;¬) -- AlisdairM

2 1

[BOOST][asio] Compilation problem w/ Serialization example
by Marc Viala 26 Jan '06

26 Jan '06

Hi, We are just trying to use Asio (ver 0.3.6) w/ BOOST 1.33.1 and we have some troubles during the compilation of the serialization example, more precisely the file "connection.hpp" @ line 79 : "boost::bind can't deduced arguments..." I've tried to replace with boost::lambda::bind but without success... Is there any patch of this example with our C++ platform? Compilation platform: PC Windows XP Visual C++ 7.1.3088 BOOST 1.33.1 Thanks in advance, Marc Viala mailto:mviala@acticm.com <mailto:mviala@acticm.com>

3 2

[filesystem] new filesystsem library broken for Borland
by AlisdairM 26 Jan '06

26 Jan '06

The latest CVS version of the filesystem library is broken for the Borland compiler. It needs the same workaround already applied for MSVC and earlier: cvs diff -- boost\filesystem\path.hpp (in directory E:\sourceforge\devel\boost\) Index: boost/filesystem/path.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/filesystem/path.hpp,v retrieving revision 1.17 diff -r1.17 path.hpp 590c590,591 < # if !defined( BOOST_MSVC ) || BOOST_MSVC > 1300 // bypass VC++ 7.0 and earlier --- > // bypass VC++ 7.0 and earlier, and broken Borland compilers > # if !BOOST_WORKAROUND( BOOST_MSVC, <= 1300 ) && !BOOST_WORKAROUND( __BORLANDC__, BOOST_TESTED_AT(0x581) ) Note that the patch has been tested with the new compiler, as well as BCB6 in the online regression testing. I plan a more complete set of BCB2006 patches later this week, but will try to get BCB6 tests passing again first. There is a second problem in convenience.hpp, where system_message cannot be seen in the surrounding namespace. A quick and dirty fix would be: cvs diff -- boost\filesystem\convenience.hpp (in directory E:\sourceforge\devel\boost\) Index: boost/filesystem/convenience.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/filesystem/convenience.hpp,v retrieving revision 1.9 diff -r1.9 convenience.hpp 237a238,240 > # if BOOST_WORKAROUND( __BORLANDC__, BOOST_TESTED_AT(0x581)) > using boost::BOOST_FILESYSTEM_NAMESPACE::system_message; > #endif -- AlisdairM

2 1

just a test
by Andy Little 25 Jan '06

25 Jan '06

just a test

1 2

Re: [boost] [wave] pp-number bug?
by Hartmut Kaiser 25 Jan '06

25 Jan '06

> Alberto Ganesh Barbati wrote: > > > However, if run into Wave, the program above produces the error > > message: > > > > error: pasting the following two tokens does not give a valid > > preprocessing token: "e" and "-" > > > > Who's right? I'd agree with Paul Mensonides, because "1e" > is a valid > > pp token which is to be formed before applying the ## > operator. But my > > understanding is quite limited in this area so I commit > myself to the > > Gurus. > > Paul is right and Wave is wrong. That's a known issue with > Wave, it stems from the fact that Wave currently works on C++ > tokens, not pp-tokens as mandated by the Standard. This will > be fixed in the future, when I change Wave to operate on > pp-tokens, rescanning these into C++ tokens after > preprocessing. If you need to preprocess this kind of > constructs with Wave, you may want to use the --variadics > command line option, allowing concatenation of unrelated tokens. Wave now correctly recognizes pp-number tokens, which get converted to C++ tokens just before they get returned to the calling application. This fixes a couple of known Wave preprocessing problems (I even enabled a couple of additional tests failing so far). Regards Hartmut

2 1

[tuple] 'make_tuple()' missing from <tuple_basic_no_partial_spec.hpp>
by MB 25 Jan '06

25 Jan '06

Hi, 'make_tuple()' that makes an empty tuple seems missing from <detail/tuple_basic_no_partial_spec.hpp>: inline tuple<> make_tuple() { return tuple<>(); } Regards, MB p-stade.sourceforge.net

1 0

Re: [boost] Interest for policy based smart pointer (smart_ptr)
by Axter 25 Jan '06

25 Jan '06

"Larry Evans" <cppljevans(a)cox-internet.com> wrote in message news: <news:%3cdqpagd$mov$1@sea.gmane.org> <dqpagd$mov$1(a)sea.gmane.org>... > On 01/19/2006 05:13 PM, David Maisonave wrote: > [snip] > > I read in a previous thread that a policy base smart pointer was > > being considered for boost, but I could not find it in the sandbox. > http://cvs.sourceforge.net/viewcvs.py/boost-sandbox/boost-sandbox/boost/poli cy_ptr/ 1. Is there any usage documentation? 2. Has any performance test comparison been done? 3. Can it compile to VC++ 6.0? It looks like work started on this in 2004, but I don't see any logging pass 26 Jun 2005. 4. Is the work for this class at a stand-still?

3 9

Re: [boost] [algorithm] About the documentation of boost::algorithm::string.
by Matthew Henderson 25 Jan '06

25 Jan '06

Pavol, Here is an example of what we observed. This program compiles and runs without error. The point here is that, if the behaviour is as described in the documentation, then we would expect the last assert to fail, but is succeeds. #include <vector> #include <string> #include <iostream> #include <cassert> #include <boost/algorithm/string.hpp> int main() { std::string test_string("a,b,c,d,e"); std::vector<std::string> test_vector; test_vector.push_back("X"); assert( test_vector.size() == 1 ); boost::algorithm::split(test_vector,test_string,boost::algorithm::is_any_of(",")); assert( test_vector.size() == 5 ); } regards, Matthew Henderson On Tue, Jan 17, 2006 at 12:04:09PM +0000, Matthew Henderson wrote: > There are two minor problems with the documentation which I have noticed. > > 1. Undocumented headers. > boost/algorithm/string/config.hpp > boost/algorithm/string/iter_find.hpp > > 2. The statement "Each part is copied and added as a new element to the output > container." which appears both in the documentation of boost/algorithm/string > /split.hpp and in comments of that header file is not an accurate description > of what actually happens. Some testing by myself suggests that, in fact, the > contents of the original container are overwritten. > This is definitely not what the split is supposed to do. Could you please send some examples proving your claim. Algorithms in split.hpp does not contain any mutating code that I'm aware of. Regards, Pavol

2 3

[regex]using boost::regex with STLport 5.0 on VS.net 2003
by Wang Weiwei 25 Jan '06

25 Jan '06

hello, I'm using the boost::regex library with STLport 5.0 on VS.net 2003. But I failed with the following simple program: #include <rope> #include <boost/regex.hpp> int main(int argc, char* argv[]) { std::rope<int> r; std::string str; boost::regex reg("(\\d+),?"); std::string::const_iterator b = str.begin(); std::string::const_iterator e = str.end(); boost::sregex_iterator it(b, e, reg); boost::sregex_iterator end; return 0; } It's a Win 32 console program project. I have defined the _STLP_DEBUG macro for debug build and used the /MDd switch (multi-threaing, dynamic, debug, runtime lib). My procedures of building STLport 5.0 and boost are as follows(under the VS.net command prompt): STLport: In <STLport>\build\lib directory, type "nmake -fnmake-vc71.mak all install" boost: Run batch file with contents: SET PYTHON_ROOT="D:\Python24" SET PYTHON_VERSION=2.4.2 SET STLPORT_PATH="D:\" SET STLPORT_50_PATH="D:\STLport_50" SET STLPORT_VERSION=5.0 SET STLPORT_VERSIONS=5.0 SET VC71_ROOT="d:\Program Files\Microsoft Visual Studio .NET 2003\Vc7" Bjam.exe -sBOOST_ROOT=. -sTOOLS=vc-7_1 "-sBUILD=debug release <runtime-link>static/dynamic <threading>single/multi" stage Bjam.exe -sBOOST_ROOT=. -sTOOLS=vc-7_1-stlport "-sBUILD=debug release <runtime-link>static/dynamic <threading>single/multi" stage Is there anything wrong? Thanks. Max

3 2