- Boost-users - lists.boost.org

asio socket options
by Eugene M. Kim 23 Jan '08

23 Jan '08

Greetings, On UNIX-like platforms, boost::asio::detail::socket_option::{boolean,integer} templates are very useful when defining socket option classes for non-standard options that are supported by the platform, e.g.: namespace so = boost::asio::detail::socket_option; typedef so::boolean<SOL_SOCKET, SO_REUSEADDR> reuse_addr; typedef so::boolean<SOL_SOCKET, SO_REUSEPORT> reuse_port; typedef so::boolean<SOL_SOCKET, SO_ERROR> pending_error; typedef so::integer<SOL_SOCKET, SO_SNDLOWAT> send_low_watermark; typedef so::integer<SOL_SOCKET, SO_RCVLOWAT> recv_low_watermark; typedef so::boolean<IPPROTO_IP, IP_HDRINCL> include_header; typedef so::boolean<IPPROTO_IP, IP_DONTFRAG> do_not_fragment; typedef so::integer<IPPROTO_IP, IP_TOS> type_of_service; typedef so::integer<IPPROTO_IP, IP_TTL> time_to_live; ... and so on. It is not suitable to include these non-standard options in the asio library itself, but it seems justifiable to move the boolean/integer templates out of boost::asio::detail and into boost::asio. Also, many socket options take a C struct as an argument, so a template for such options might be useful as well: template <int Level, int Name, class Value> class typed : public Value { typedef Value value; template <typename Protocol> int level(const Protocol &) const { return Level; } template <typename Protocol> int name(const Protocol &) const { return Name; } template <typename Protocol> value *data(const Protocol &) { return (value *)this; } template <typename Protocol> std::size_t size(const Protocol &) const { return sizeof(Value); } template <typename Protocol> void resize(const Protocol &, std::size_t s) const { if (s != sizeof(Value)) throw std::length_error("resizing fixed-length socket option"); } }; Then it becomes trivial to define and use such socket options: typedef typed<SOL_SOCKET, SO_SNDTIMEO, struct ::timeval> send_timeout; typedef typed<SOL_SOCKET, SO_RCVTIMEO, struct ::timeval> recv_timeout; send_timeout sndtimeo; sndtimeo.tv_sec = 1; sndtimeo.tv_usec = 0; Would this template be appropriate for inclusion in a future version of asio? Just my $.02. Cheers, Eugene

1 0

meta-programming library
by noman javed 23 Jan '08

23 Jan '08

Hi I need to work with boost-mpl library. Should I install the whole boost library or is there any possibility to just install boost-mpl ? Thanks in anticipation Noman Javed --------------------------------- Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now.

2 1

The Scanner Business
by jono 23 Jan '08

23 Jan '08

hello possums! i have the following rules declared in a spirit grammar definition (using boost 1.8.4 )..... //----------------------- typedef typename lexeme_scanner<ScannerT>::type lexeme_scanner_t; rule<lexeme_scanner_t> legal_initial_p ,legal_p ; //----------------------- and initialized in definition::definition() .... //----------------------- legal_initial_p = alpha_p | ch_p('_') |'<'|'>'|'='|'?'|'/'|'*'|'&'|'-'|'+'|'%' |'!'|'$'|':'|'@' ; legal_p = legal_initial_p | digit_p | ch_p('.') ; //----------------------- the above code compiles and functions as expected under MSVC 7.1 but causes g++ 4.1.1 to hang. if i use type rule<ScannerT> for legal_p and legal_initial_p, the code compiles on both, but doesn't work properly. i'm using |dynamic_distinct_parser ||so i can handle the '.' character as a syntactic token in certain contexts. |||//---------------------- |some declarations ... || struct keywords : symbols<Object> { ... } keywords_p; || ||dynamic_distinct_parser<ScannerT> keyword_p; ||||| //----------------------|some definitions ...| | ||chset<> distinct(chset<>(legal_ident) | chset<>("0-9")); distinct_directive<> keyword_d(distinct); distinct_parser<> keyword_p(distinct); |||| || ||rule<ScannerT, object_closure::context_t> ||Symbol_p = lexeme_d [ ( legal_initial_p >> *legal_p ) - keywords_p ] [ Symbol_p.value_ = intern_(arg1, arg2) ] ; || ||| dot_p = keyword_p("."); //---------------------- |// .... ||it's quite possible there's a better way to do the parsing, but it was fiddly, and i don't have any more time to mess with it. (the rest of the grammar is a recursive scheme reader, hence the closures) i would be very grateful for any suggestions as to what might be causing g++ to go walkabout! cheers, jono jp(a)day-one.com | | |

1 0

Regex newbie problem
by Barberi 23 Jan '08

23 Jan '08

Newbie to boost. I used the MSVC 8.0 installer, and got the first couple of example programs to work, but failed on the regex example. I'm using MSVC 2008 (aka MSVC-9.0) (Express Edition) IDE, in Debug configuration. The linker can't resolve several symbols with libboost_regex_vc80-mt-gd-1_34_1.lib. Stripped of the qualifiers these are: _Orphan_all(void) const, ~_Container_base(void), _Swap_all(...), and insert(...). I switched to release configuration, now the linker looks in libboost_regex_vc80-mt-1_34_1.lib and can't find insert(...). Any suggestions? Should I wait for a MSVC-9.0 installer, and go back to MSVC 2005 in the meantime? Stephen B.

3 8

Boost, HP-UX and gcc
by Martin_Thomas＠McAfee.com 22 Jan '08

22 Jan '08

Hello, I am trying to use use Boost::Regex on HP 11i with gcc 4.2.2 and I notice that in config/platform/hpux.hpp there is a define to turn off wide-char support unless using aCC and have a1 source eg: #if !(defined(__HP_aCC) || !defined(_INCLUDE__STDC_A1_SOURCE)) # define BOOST_NO_SWPRINTF # define BOOST_NO_CWCTYPE #endif This will never allow gcc to use wregex even if wchar support is available. Is this because there are so few people using Zgcc oin HP-ux or am I missing something? On the assumption that it is related to usage, I have added the following to allow gcc 4 to use wregex: #if defined(__GNUC__) # if (__GNUC__ >3) # undef BOOST_NO_CWCTYPE # endif #endif Is there anything obviously wrong with this? Thanks // Martin

4 8

[program_options] crash on application exit
by ope 22 Jan '08

22 Jan '08

Hi, to get the main() more clean, I moved the command line options parser to an own class (some stuff missing): ---8--- class CommandLineOptions { public: CommandLineOptions( int argc, char** argv ); private: po::options_description generic_options; po::options_description qt4_common; #if defined( Q_OS_UNIX ) || defined( Q_OS_MAC ) po::options_description qt4_x11; #endif #if defined( Q_OS_WIN ) po::options_description qt4_win; #endif #if defined DEBUG po::options_description qt4_dbg; #endif po::options_description hidden_options; po::options_description usage_options; po::options_description cmdline_options; po::positional_options_description pos_options; po::variables_map vm; }; --->8--- with ---8<--- CommandLineOptions::CommandLineOptions( int argc, char** argv ) : generic_options("Generic options"), qt4_common("Qt options"), #if defined( Q_OS_UNIX ) || defined( Q_OS_MAC ) qt4_x11( "Qt X11 specific options" ), #endif #if defined( Q_OS_WIN ) qt4_win( "Qt Windows specific options" ), #endif #if defined( DEBUG ) qt4_dbg( "Qt debug options" ), #endif hidden_options("Hidden options") { generic_options.add_options() ( "version,v", "print version string") ( "help,h", "produce help message") ; hidden_options.add_options()(...); qt4_common.add_options()(... #if defined( Q_OS_UNIX ) || defined( Q_OS_MAC ) qt4_x11.add_options()(... #if defined( Q_OS_WIN ) qt4_win.add_options()(... #if defined( DEBUG )(... usage_options .add( generic_options ) .add( qt4_common ) #if defined( Q_OS_UNIX ) || defined( Q_OS_MAC ) .add( qt4_x11 ) #elif defined( Q_OS_WIN ) .add( qt4_win ) #endif #if defined( DEBUG ) .add( qt4_dbg ) #endif ; cmdline_options.add( usage_options ).add( hidden_options ); pos_options.add( "input-file", -1 ); po::store(po::command_line_parser(argc, argv). options(cmdline_options).positional(pos_options).run(), vm); po::notify(vm); if (vm.count( "version" )) { std::cout << Version() << "\n"; } ... } --->8--- and the main(): ---8<--- int main( int argc, char **argv ) { try { logging::init(); CommandLineOptions cmdline( argc, argv ); QT_REQUIRE_VERSION( argc, argv, QT_VERSION_STR ); Q_INIT_RESOURCE( workbench ); QApplication app( argc, argv ); QMainWindow wb; wb.show(); // using boost.logging v2 L_ << "execute application."; return app.exec(); ... --->8--- It works fine until the app exits. My gdb session wasn't very useful to me: (gdb) target exec ./edas (gdb) run Starting program: /.../edas2/edas Using host libthread_db library "/lib/i686/cmov/libthread_db.so.1". (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) [Thread debugging using libthread_db enabled] [New Thread 0xb70506c0 (LWP 13445)] [New Thread 0xb6aa5b90 (LWP 13450)] 28:31 [1] execute application. Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0xb70506c0 (LWP 13445)] 0x08051980 in ?? () (gdb) backtrace #0 0x08051980 in ?? () If I comment out my CommandLineOptions class instance it works. Thanks, Olaf

1 0

boost::filesystem::rename - behavior/code vs. docs
by Andriy Gapon 22 Jan '08

22 Jan '08

This is what documentation on boost::filesystem::rename states: http://www.boost.org/libs/filesystem/doc/tr2_proposal.html template <class Path1, class Path2> void rename(const Path1& from_p, const Path2& to_p); Requires: Path1::external_string_type and Path2::external_string_type are the same type. Effects: Renames from_p to to_p, as if by POSIX rename(). Postconditions: !exists(from_p) && exists(to_p), and the contents and attributes of the file originally named from_p are otherwise unchanged. [Note: If from_p and to_p resolve to the same file, no action is taken. Otherwise, if to_p resolves to an existing file, it is removed. A symbolic link is itself renamed, rather than the file it resolves to being renamed. -- end note] Actual effect is that an exception is thrown when renaming a file to an existing file on the same fs. rename(2) performs an 'atomic' rename in this case without problems, but in the boost code there is a specila check against "too permissive" posix rename(): http://boost.cvs.sourceforge.net/boost/boost/libs/filesystem/src/operations… 1126 rename_api( const std::string & from, const std::string & to) 1127 { 1128 // POSIX is too permissive so must check 1129 fs::system_error_type dummy; 1130 if ( fs::exists( status_api( to, dummy ) ) ) 1131 return EEXIST; 1132 return std::rename( from.c_str(), to.c_str() ) != 0 1133 ? errno : 0; 1134 } And rename testing code also checks for the above behavior. I think that this is a huge difference between the docs and the code, and this is very counter-intuitive too. -- Andriy Gapon

1 0

[Boost.Iostreams] Question about gzip codec and dual_use filter
by Marc Viala Perso 22 Jan '08

22 Jan '08

Hello, I'm just wondering why zlib and bzip2 codec are "dual_use" filter but not gzip? Can anybody enlighten me on this restriction? In fact, I have the following use case: I need to manage an "iostream" instance for interfacing with an external library but I would like to be able to support the GZIP codec during writing/reading operations, that is, I must deal with the following "external" serializer engine: SerializerEngine ar(std::iostream& ios, OpeningMode mode) The SerializerEngine interface provides some read/write raw operations. For doing that, I've just tried to implement the following class: class gzfstream: public boost::iostreams::filtering_stream<boost::iostreams::dual_use> { public: template <class Path> gzfstream(const Path& file_path, std::ios_base::openmode mode): _file(file_path, mode) { if( _file.fail() ) return ; namespace fs = boost::filesystem ; namespace io = boost::iostreams ; const typename Path::string_type& ext = fs::extension(file_path) ; if( gz_extension_exists(ext) ) { const bool in_op = (mode & std::ios::in) != 0 ; const bool out_op = (mode & std::ios::out) != 0 ; if( in_op && out_op ) throw std::logic_error("can't support input/output operation with codec") ; if( out_op ) push(io::gzip_compressor()) ; else push(io::gzip_decompressor()) ; } push(_file) ; } ~gzfstream() { reset() ; } private: boost::filesystem::fstream _file ; bool gz_extension_exists(const std::string& ext) { return ext.size() == 3 && ext == ".gz" ; } bool gz_extension_exists(const std::wstring& ext) { return ext.size() == 3 && ext == L".gz" ; } } ; But this class doesn't compile only for GZIP codec because, as explained in the documentation, GZIP doesn't support dual_use concept - contrary to the ZLib and BZip2 codecs. Is there any issue to deal with my use case? Any help will be appreciate? Best regards, Marc VIALA

4 7

[boost-users] [boost-thread linker error on msvc]
by chun ping wang 22 Jan '08

22 Jan '08

Hi i get the following linker error when compling a program that links to boost-thread. Here is the error... "C:/Program Files/Microsoft Visual Studio 8/VC/bin/lib.exe" -LTCG -out:"./lib/libAlgorithm.lib" ./lib/BitOperation.obj ./lib/ContainerOperation.obj ./lib/DiffieHellman.obj ./lib/dsa.obj ./lib/Elucid.obj ./lib/IsPrime.obj ./lib/mathfunc.obj ./lib/median.obj ./lib/NpComplete.obj ./lib/polynomial.obj ./lib/rsa.obj ./lib/sortingAlgo.obj ./lib/Fourier.obj ./lib/AES.obj Microsoft (R) Library Manager Version 8.00.50727.762 Copyright (C) Microsoft Corporation. All rights reserved. rsa.obj : warning LNK4006: "class boost::random::shuffle_output<class boost::random::linear_congruential<unsigned long,1366,150889,714025,0>,97,139726> Random::generator" (?generator@Random@@3V?$shuffle_output@V ?$linear_congruential@K$0FFG@$0CENGJ@$0KOFCJ@$0A@@random@boost @@$0GB@$0CCBMO@@random@boost@@A) already defined in sortingAlgo.obj; second definition ignored rsa.obj : warning LNK4006: "void __cdecl Random::initialize(void)" (?initialize@Random@@YAXXZ) already defined in sortingAlgo.obj; second definition ignored NpComplete.obj : warning LNK4006: "class boost::random::shuffle_output<class boost::random::linear_congruential<unsigned long,1366,150889,714025,0>,97,139726> Random::generator" (?generator@Random@@3V?$shuffle_output@V ?$linear_congruential@K$0FFG@$0CENGJ@$0KOFCJ@$0A@@random@boost @@$0GB@$0CCBMO@@random@boost@@A) already defined in sortingAlgo.obj; second definition ignored NpComplete.obj : warning LNK4006: "void __cdecl Random::initialize(void)" (?initialize@Random@@YAXXZ) already defined in sortingAlgo.obj; second definition ignored IsPrime.obj : warning LNK4006: "class boost::random::shuffle_output<class boost::random::linear_congruential<unsigned long,1366,150889,714025,0>,97,139726> Random::generator" (?generator@Random@@3V?$shuffle_output@V ?$linear_congruential@K$0FFG@$0CENGJ@$0KOFCJ@$0A@@random@boost @@$0GB@$0CCBMO@@random@boost@@A) already defined in sortingAlgo.obj; second definition ignored IsPrime.obj : warning LNK4006: "void __cdecl Random::initialize(void)" (?initialize@Random@@YAXXZ) already defined in sortingAlgo.obj; second definition ignored dsa.obj : warning LNK4006: "class boost::random::shuffle_output<class boost::random::linear_congruential<unsigned long,1366,150889,714025,0>,97,139726> Random::generator" (?generator@Random@@3V?$shuffle_output@V ?$linear_congruential@K$0FFG@$0CENGJ@$0KOFCJ@$0A@@random@boost @@$0GB@$0CCBMO@@random@boost@@A) already defined in sortingAlgo.obj; second definition ignored dsa.obj : warning LNK4006: "void __cdecl Random::initialize(void)" (?initialize@Random@@YAXXZ) already defined in sortingAlgo.obj; second definition ignored DiffieHellman.obj : warning LNK4006: "class boost::random::shuffle_output<class boost::random::linear_congruential<unsigned long,1366,150889,714025,0>,97,139726> Random::generator" (?generator@Random@@3V?$shuffle_output@V ?$linear_congruential@K$0FFG@$0CENGJ@$0KOFCJ@$0A@@random@boost @@$0GB@$0CCBMO@@random@boost@@A) already defined in sortingAlgo.obj; second definition ignored DiffieHellman.obj : warning LNK4006: "void __cdecl Random::initialize(void)" (?initialize@Random@@YAXXZ) already defined in sortingAlgo.obj; second definition ignored rm -f ./lib/BitOperation.obj ./lib/ContainerOperation.obj ./lib/DiffieHellman.obj ./lib/dsa.obj ./lib/Elucid.obj ./lib/IsPrime.obj ./lib/mathfunc.obj ./lib/median.obj ./lib/NpComplete.obj ./lib/polynomial.obj ./lib/rsa.obj ./lib/sortingAlgo.obj ./lib/Fourier.obj ./lib/AES.obj "C:/Program Files/Microsoft Visual Studio 8/VC/bin/link.exe" -LTCG -out: PolyTestProject.exe ./project/PolyTest.obj -LIBPATH:"C:/Program Files/Microsoft Visual Studio 8/VC/lib" -LIBPATH:"C:/boost_msvc/lib" -LIBPATH:"./lib" libAlgorithm.lib libboost_thread-mt.lib -MACHINE:x86 -nologo -NODEFAULTLIB:msvcrt msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: void __thiscall std::_Mutex::_Lock(void)" (?_Lock@_Mutex@std@@QAEXXZ) already defined in libcpmt.lib(xmutex.obj) msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: void __thiscall std::_Mutex::_Unlock(void)" (?_Unlock@_Mutex@std@@QAEXXZ) already defined in libcpmt.lib(xmutex.obj) msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: class std::basic_ostream<char,struct std::char_traits<char> > & __thiscall std::basic_ostream<char,struct std::char_traits<char> >::flush(void)" (?flush@?$basic_ostream@DU?$char_traits@D@std@@@std@@QAEAAV12@XZ) already defined in PolyTest.obj msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: void __thiscall std::basic_ostream<char,struct std::char_traits<char> >::_Osfx(void)" (?_Osfx@?$basic_ostream@DU?$char_traits@D@std@@@std@@QAEXXZ) already defined in PolyTest.obj msvcprt.lib(MSVCP80.dll) : error LNK2005: "bool __cdecl std::uncaught_exception(void)" (?uncaught_exception@std@@YA_NXZ) already defined in libcpmt.lib(uncaught.obj) msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: void __thiscall std::basic_ios<char,struct std::char_traits<char> >::setstate(int,bool)" (?setstate@?$basic_ios@DU?$char_traits@D@std@@@std@@QAEXH_N@Z) already defined in PolyTest.obj msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: int __thiscall std::basic_streambuf<char,struct std::char_traits<char> >::sputn(char const *,int)" (?sputn@?$basic_streambuf@DU?$char_traits@D@std@@@std@@QAEHPBDH@Z) already defined in PolyTest.obj msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: int __thiscall std::basic_streambuf<char,struct std::char_traits<char> >::sputc(char)" (?sputc@?$basic_streambuf@DU?$char_traits@D@std@@@std@@QAEHD@Z) already defined in PolyTest.obj msvcprt.lib(MSVCP80.dll) : error LNK2005: "public: __thiscall std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> >::~basic_string<char,struct std::char_traits<char>,class std::allocator<char> >(void)" (??1?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@QAE@XZ) already defined in PolyTest.obj PolyTest.obj : error LNK2001: unresolved external symbol "class ?? ::InputIterator __cdecl algocpw::horner( ?? ) throw( ?? )" (?horner@algocpw@@YA?AVInputIterator) PolyTest.obj : error LNK2001: unresolved external symbol "class ?? ::InputIterator __cdecl algocpw::evaluate( ?? ) throw( ?? )" (?evaluate@algocpw@@YA?AVInputIterator) PolyTestProject.exe : fatal error LNK1120: 2 unresolved externals I am suing msvc 2005.

2 3

How to format gregorian::date-output dynamicly
by thom_schu＠gmx.de 22 Jan '08

22 Jan '08

Hi, to be compatible with a given interface I need to encapsulate a boost::gregorian::date. I have a problem with the formatted output of the date-object. I tried a given example: date d(2005,Jun,25); date_facet* facet(new date_facet("%A %B %d, %Y")); std::stringstream ss; ss.imbue(std::locale(std::cout.getloc(), facet)); ss << d; std::cout << ss.str() << std::endl; facet->format("%Y"); ss.str(""); ss << d; std::cout << ss.str() << std::endl; this works fine. But the same inside an object causes an segmentation-fault: class MyDate { date m_date; std::stringstream ss; date_facet* facet; Date() : m_date(day_clock::local_day()), facet(new date_facet("%A %B %d, %Y")) { ss.imbue(std::locale(std::cout.getloc(), facet)); } std::string Date::format(std::string pattern) { facet->format(pattern.c_str()); ss.str(""); ss << m_date; return ss.str(); } } Can anybody tell me what Im doing wrong and how to do it better ? Thanks a lot gizmo -- Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten Browser-Versionen downloaden: http://www.gmx.net/de/go/browser -- Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer

1 0