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

[Unit Test] default FPC_STRONG limtations when checking close to 0.
by Chris Fairles 01 Jun '07

01 Jun '07

After wondering why my BOOST_CHECK_CLOSE(some_number_close_to_zero, 0.0) hadn't been working, I found that the default comparison type is FPC_STRONG which means the ration between the abs val of the difference and BOTH values must be < tolerance but this doesn't turn out to be true (running through what safe_fpt_division outputs atm ... difference between 0.0{0} and 0.0 std::numeric_limits<double>::epsilon(){-2.2204460492503131e-16} exceeds 0.01% It should work using FPC_WEAK but I can't find a way through the test_tools to easily change this. I have three solutions at this time. 1) Use BOOST_CHECK_SMALL. However, I have code that loops through matrix indicies like: for (int j = 0; j < matrix.col; j++) for (int i = 0; i < matrix.row; i++) BOOST_CHECK_CLOSE(matrix(i,j), data(i,j), tolerance); I'm going to have to use an if statement for the special case where I'm testing against 0 and it doesn't use the tolerance anyway so the test is completely different. Theres another number to determine "how small" the number should be. 2) Use check_is_close explicitly and change the default comparison type to FPC_WEAK. This means going outside the nice auto-unit-test and test-tools framework which although I'm fine with, other programmers writing test cases on my team might not be. 3) Define my own macro to simplify solution 2. Something along the lines of: CHECK_IS_CLOSE_WEAK(x,y,t) { check_is_close(x ,y,t, FPC_WEAK); } Are there any other recommended solutions?

2 3

[smart_ptr] shared_ptr and shared_array of const
by Tim Day 31 May '07

31 May '07

The ability to assign a shared_ptr<T> to shared_ptr<const T> is useful for controlling which parts of software have write access to the shared object, so I'm surprised that while e.g boost::shared_ptr<int> sp(new int); boost::shared_ptr<const int> csp=sp; compiles fine, and prohibits e.g *csp=0;, the equivalent for shared_array doesn't compile: boost::shared_array<int> sa(new int[10]); boost::shared_array<const int> csa=sa; // not allowed [The gcc 4.1.2 error is "conversion from ‘boost::shared_array<int>’ to non-scalar type ‘boost::shared_array<const int>’ requested"] I'm looking to prevent modification of elements of csa of course. The obvious alternative is to use a shared_ptr to a std::vector of course: boost::shared_ptr<std::vector<int> > sv(new std::vector<int>(10)); boost::shared_ptr<const std::vector<int> > csv=sv; // fine // (*csv)[0]=0; // prevented by const, as intended However, it's unclear to me exactly why the shared_array of const doesn't work as expected. Am I missing some necessary detail to make it work, is shared_array simply less featureful (the header is certainly shorter) or is there some reason it can't work ? Thanks for any explanation Tim [I'm using boost 1.33.1]

1 0

Error compiling Boost.Python library on Solaris
by Belousov, Arkadiy, GCM 31 May '07

31 May '07

Hello all! I have a compilation error with Boost 1.34.0, Sun Workshop 11 compiler. Compiling Python library, I get "libs/python/src/object/class.cpp", line 458: Error: Overloading ambiguity between "boost::python::api::getattr<boost::python::scope, char[11], boost::python::str>(const boost::python::scope&, const char(&)[11], const boost::python::str&)" and "boost::python::api::getattr(const boost::python::api::object&, const char*, const boost::python::api::object&)". The code in question is object module_prefix() { return object( PyObject_IsInstance(scope().ptr(), upcast<PyObject>(&PyModule_Type)) ? object(scope().attr("__name__")) : api::getattr(scope(), "__module__", str()) <<===== ); } Of course, casting the string constant to (const char*) solves the problem, even though I realize it's not a boost-like solution. The rest of 1.34 compiles perfectly for me (thanks are in order - excellent job!). The exact version of the compiler is $ CC -V CC: Sun C++ 5.8 Patch 121017-07 2006/11/02 Sincerely, Arkadiy Belousov ----------------------------------------- ******************************************************************* * This e-mail is intended only for the addressee named above. As this e-mail may contain confidential or privileged information, if you are not the named addressee, you are not authorized to retain, read, copy or disseminate this message or any part of it. ******************************************************************* *

2 1

unified documentation
by Rob Agar 31 May '07

31 May '07

hi all It seems to me (as a newcomer) that the one thing boost needs is documentation in a unified structure. I'm thinking of something along the lines of the MS .NET class library or the Java class references. The existing documentation for the libraries I have used so far are very hard to navigate - they seem to be more aimed at standards committees than end users. As an end user, the first thing I need is an easily navigated tree just listing all the namepaces, functions, classes, typedefs etc. As a start does anyone have a doxygen config file for generating a simple function list and class hierarchy? cheers Rob Agar

6 12

1_33_1 VC8 crash in program_options::cmdline
by Søren Holstebroe 31 May '07

31 May '07

This problem was found in boost 1.33.1 using VC++ 8.0. I haven't tried 1.34.0, but the cmdline code seems unchanged. from cmdline.cpp -------------------------------------- std::vector<option> cmdline::parse_long_option(std::vector<string>& args) { vector<option> result; const std::string& tok = args[0]; if (tok.size() >= 3 && tok[0] == '-' && tok[1] == '-') { string name, adjacent; string::size_type p = tok.find('='); if (p != tok.npos) { name = tok.substr(2, p-2); adjacent = tok.substr(p+1); if (adjacent.empty()) throw invalid_command_line_syntax(name, invalid_command_line_syntax::empty_adjacent_parameter); } else { name = tok.substr(2); } option opt; opt.string_key = name; if (!adjacent.empty()) opt.value.push_back(adjacent); opt.original_tokens.push_back(tok); result.push_back(opt); args.erase(args.begin()); } return result; } -------------------------------------- The line args.erase(args.begin()); crashes in MS VC++ 8.0 in release optimized configuration, though it works fine in debug mode. Could this be a compiler problem with loop unrolling and the const std::string& tok = args[0]; reference ? As seen in the stack trace, std::vector::erase tries to copy a "Bad Ptr" string and obviously fails miserably. Interestingly the program hangs in std::_Distance2 if I provide no command line arguments (it looks like a very long bad pointer loop, not an infinite loop). I might try recompiling boost with different compile options, but it is quite time consuming to experiment with. best regards Søren Holstebroe Code: namespace po = boost::program_options; po::options_description desc("Allowed options"); desc.add_options() ("help", "produce help message") ; po::variables_map vm; po::store(po::parse_command_line(argc, argv, desc), vm); po::notify(vm); Stack trace: > msvcr80.dll!memcpy(unsigned char * dst=0x00c20048, unsigned char * src=0xabababab, unsigned long count=458755) Line 188 Asm msvcr80.dll!memcpy_s(void * dst=0x00c20048, unsigned int sizeInBytes=458767, const void * src=0xabababab, unsigned int count=458755) Line 67 + 0xc bytes C msvcp80.dll!std::char_traits<char>::_Copy_s(char * _First1=0x00c20048, unsigned int _Size_in_bytes=458767, const char * _First2=0xabababab, unsigned int _Count=458755) Line 576 C++ msvcp80.dll!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::assign(const std::basic_string<char,std::char_traits<char>,std::allocator<char> > & _Right=<Bad Ptr>, unsigned int _Roff=0, unsigned int _Count=4294967295) Line 1059 + 0x2c bytes C++ msvcp80.dll!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::operator=(const std::basic_string<char,std::char_traits<char>,std::allocator<char> > & _Right=<Bad Ptr>) Line 1044 C++ PS.Vision.Console.exe!std::_Copy_opt<std::basic_string<char,std::char_traits<char>,std::allocator<char> > *,std::basic_string<char,std::char_traits<char>,std::allocator<char> > *>() + 0x4a bytes C++ PS.Vision.Console.exe!std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > >::erase() + 0x2d bytes C++ PS.Vision.Console.exe!boost::program_options::detail::cmdline::parse_long_option() + 0x28e bytes C++ PS.Vision.Console.exe!boost::_mfi::mf1<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::program_options::detail::cmdline,std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &>::operator()() + 0x40 bytes C++ PS.Vision.Console.exe!boost::_bi::list2<boost::_bi::value<boost::program_options::detail::cmdline *>,boost::arg<1> >::operator()<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::_mfi::mf1<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::program_options::detail::cmdline,std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &>,boost::_bi::list1<std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &> >() + 0x22 bytes C++ PS.Vision.Console.exe!boost::_bi::bind_t<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::_mfi::mf1<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::program_options::detail::cmdline,std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &>,boost::_bi::list2<boost::_bi::value<boost::program_options::detail::cmdline *>,boost::arg<1> > >::operator()<std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > >() + 0x2d bytes C++ PS.Vision.Console.exe!boost::detail::function::function_obj_invoker1<boost::_bi::bind_t<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::_mfi::mf1<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,boost::program_options::detail::cmdline,std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &>,boost::_bi::list2<boost::_bi::value<boost::program_options::detail::cmdline *>,boost::arg<1> > >,std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &>::invoke() + 0x1d bytes C++ PS.Vision.Console.exe!boost::function1<std::vector<boost::program_options::basic_option<char>,std::allocator<boost::program_options::basic_option<char> > >,std::vector<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::allocator<std::basic_string<char,std::char_traits<char>,std::allocator<char> > > > &,std::allocator<boost::function_base> >::operator()() + 0x5b bytes C++ PS.Vision.Console.exe!boost::program_options::detail::cmdline::run() + 0x371 bytes C++ PS.Vision.Console.exe!boost::program_options::basic_command_line_parser<wchar_t>::run() Line 102 + 0x12 bytes C++ PS.Vision.Console.exe!boost::program_options::parse_command_line<wchar_t>(int argc=5, wchar_t * * argv=0x003b5718, const boost::program_options::options_description & desc={...}, int style=0, boost::function1<std::pair<std::basic_string<char,std::char_traits<char>,std::allocator<char> >,std::basic_string<char,std::char_traits<char>,std::allocator<char> > >,std::basic_string<char,std::char_traits<char>,std::allocator<char> > const &,std::allocator<boost::function_base> > ext={...}) Line 119 + 0x9d bytes C++

1 0

Building Boost 1.33.1 with different #defines for release and debug
by Ben Pope 31 May '07

31 May '07

Hi all, I'm sure this must have been covered before, but I am completely unable to find what I'm after, so apologies if that is the case. I seem to have a feeling that BBv2 as shipped with 1.33.1 is not completely ready for compiling with msvc 8 on Windows with autolinking, but I'm not sure where I go that from. What I would like to do is build the following configurations with VC8: * debug with shared libraries * release shared and static libraries with _SECURE_SCL=0 defined. Can that be done in a single call of bjam? I don't mind if we get debug static at the same time. And for completeness and the sake of comparison (and for when we move to 1.34) what is the equivalent with BBv2? I've noticed that there is some conditional stuff: <variant>release:<define>_SECURE_SCL=0 But couldn't work out how to do it on the command line, do I need to create some jam config files? Thanks for your time, Ben

1 0

iostreams: stream vs stream_buffer
by Os 31 May '07

31 May '07

Hi, could someone elaborate on when I would use a stream and when I would use a stream_buffer (filtered or otherwise)? They look remarkably similar to me. cheers, os. -- os(a)collective.co.uk http://www.collective.co.uk/ http://www.myspace.com/darkroomtheband

1 0

[boost-1.33.1] broken mmap_file_iterator in spirit implementation details.
by Paweł Sikora 31 May '07

31 May '07

hi all, mmap based file_iterator implementation has a nasty bug. it treats empty files as non-existent. mmap() doesn't map empty files into memory, but boost's implementation omit this little detail of posix specification and return an error. in fact, empty file != doesn't exist. regards, pawel. -- to_be || !to_be == 1, to_be | ~to_be == -1

1 0

Re: [Boost-users] Cross compiling boost
by Peter Poulsen 31 May '07

31 May '07

Has anybody tried compiling boost 1.34 for ARM? /peter >> >> I'm trying to build boost for an arm processor, but all the >> hidden bjam stuff is simply to much, so I have to ask for help :) >> >> All my build tools are in ~/arm-tools/. I'm on a Linux x86 >> and the cross compiler I'm trying to use is a gcc version 3.4.4. >> >> I have tried various build commands. The latest being: >> >> >> ~/boost_1_34_0 $ bjam --build-dir=/tmp/build-boost >> --toolset-gcc-3.4.4 --with-regex -sGXX=arm-wrs-linux-gnueabi-g++ stage >> >> >> But I still get: >> >> >> ~/boost_1_34_0 $ file stage/lib/libboost_regex-gcc34.so >> stage/lib/libboost_regex-gcc34.so: ELF 32-bit LSB shared >> object, Intel 80386, version 1 (SYSV), stripped >> >> Can anybody explain to me what I must do to compile >> boost-regex for arm? >> >> btw I have found others trying to do the same but they seem to end up hacking the >> bjam files, and I have no idea of how to that. >> >> /peter >> >> >> -- >> A: Because it messes up the order in which people normally read text. >> Q: Why is it such a bad thing? >> A: Top-posting. >> Q: What is the most annoying thing on usenet and in e-mail? >> >> >> >> >> _______________________________________________ >> Boost-users mailing list >> Boost-users(a)lists.boost.org >> http://lists.boost.org/mailman/listinfo.cgi/boost-users >> >> > > This message (including any attachments) contains confidential > and/or proprietary information intended only for the addressee. Any unauthorized > disclosure, copying, distribution or reliance on the contents of this information is > strictly prohibited and may constitute a violation of law. If you are not the > intended recipient, please notify the sender immediately by responding to this > e-mail, and delete the message from your system. If you have any questions about > this e-mail please notify the sender immediately. > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > >

1 0

Wave Driver Program
by ivan＠0x4849.net 31 May '07

31 May '07

Hi, I am using debian linux and wanted to access the Wave Driver program. It seems they have a package for the wave library but that doesnt seem to include the driver binary. I had a look in the code but couldn't find where the driver program's code is either. Is it the intention to release the driver as a C++ pre-processor binary similar to how cpp is used, or is the driver program just meant for one-off testing? In any case could someone point me to the code for the driver within the boost source tree? Thanks, Ivan --------------------- Ivan Novick http://www.0x4849.net

2 1

iostream::mapped_file
by Richard_Morrey＠McAfee.com 31 May '07

31 May '07

A query about the above. I wish to use it to enable us to share memory amongst processes. Whilst it obviously should be possible to do this using CreateFileMapping on Windows and mmap on linux, which the class uses, the documentation is not clear as to whether it is designed be used in this way. Looking at the source code: (a) It does not provide a kernel name to the CreateFileMapping call on Windows. According to the MSDN docs this is the only way to share one mapping amongst multuiple processes, and therefore it is possible that a different mapping, albeit of the same file, is being created for each process. I don't want that as the point is to keep memory usage down. (b) On linux it only uses MEM_SHARED flag if the file is read only. If this flag isnt used, is there the possiblity that more than one copy of each page of the file will be loaded into physical memory too? Clarification on any of these points is most welcome Thanks Richard Morrey.

1 0

iostreams build error in Visual Studio 2005
by Os 31 May '07

31 May '07

In file_descriptor.cpp I get a deprecation warning from line 219: lseek is deprecated in favour of _lseek. I've just patched my local copy, but could this be changed in the official version? cheers, os. -- os(a)collective.co.uk http://www.collective.co.uk/ http://www.myspace.com/darkroomtheband

3 3

[multi_index] MSVC 8.0 visualizer for indexed_by<sequenced<>, ...> containers
by Filip Konvička 31 May '07

31 May '07

Hello all, for those interested, here is an alpha version of a MSVC-8.0 visualizer for multi_index_containers whose first index is sequenced<>. To try it out, insert autoexp.dat.txt contents into your Microsoft Visual Studio 8\Common7\Packages\Debugger\autoexp.dat For your container to be visualized, you need to call the VISUALIZE_MULTI_INDEX_CONTAINER macro at the global namespace level (see the attached example, test2.cpp, and the screenshot of a test2.cpp debugging session, sample.png). Comments / suggestions are welcome! Things that need attention: - of course, containers whose first index is not sequenced<> - maybe there is a better way of enforcing template instantiation...? - possible name conflict issues, you guys probably know better than I do :-) Cheers, Filip P.S.: My colleague says that this is black magic, so beware :-) #include "common.hpp" int wmain() { cont test_cont; test_cont.push_back(test(L"aaa")); test_cont.push_back(test(L"bbb")); test_cont.push_back(test(L"bbbx")); test_cont.push_back(test(L"bbbss")); test_cont.push_back(test(L"ccc", L"ddd")); return 0; // see screenshot } #ifndef MIC_VISUALIZER_HPP #define MIC_VISUALIZER_HPP template<typename T> struct multi_index_helper { }; template<typename T> struct multi_index_helper_2 { }; template<typename T> struct multi_index_helper_3 { }; #define __VISUALIZE_MIC_INDEX_DEF(Type) \ boost::multi_index::detail::multi_index_node_type< \ Type::value_type, \ Type::index_specifier_type_list, \ std::allocator<Type::value_type> >::type /// Enables MSVC visualizers for multi_index_container Type #define VISUALIZE_MULTI_INDEX_CONTAINER(Type) \ template<> \ struct multi_index_helper_2<Type> { \ typedef __VISUALIZE_MIC_INDEX_DEF(Type) index_node_type; \ typedef \ boost::base_from_member< \ boost::detail::allocator::rebind_to< \ std::allocator<Type::value_type>, \ index_node_type \ >::type> \ base_type; \ typedef \ boost::multi_index::detail::header_holder< \ index_node_type, Type> \ header_holder_type; \ typedef \ boost::multi_index::detail::multi_index_base_type< \ Type::value_type, \ Type::index_specifier_type_list, \ std::allocator<Type::value_type> >::type \ index_type; \ /* mimic multi_index_container layout */ \ base_type base; \ header_holder_type header; \ index_type index; \ unsigned int node_count; \ /* ensure that the debugger sees this template instance */ \ static void get() { \ multi_index_helper_2<Type> *ensure_inst=0; \ } \ }; \ template<> \ struct multi_index_helper_3<Type> { \ typedef __VISUALIZE_MIC_INDEX_DEF(Type) index_node_type; \ typedef \ boost::multi_index::detail::header_holder< \ index_node_type, Type> \ header_holder_type; \ typedef \ boost::multi_index::detail::multi_index_base_type< \ Type::value_type, \ Type::index_specifier_type_list, \ std::allocator<Type::value_type> >::type \ index_type; \ /* mimic multi_index_container layout from header field */ \ header_holder_type header; \ index_type index; \ unsigned int node_count; \ /* ensure that the debugger sees this template instance */ \ static void get() { \ multi_index_helper_3<Type> *ensure_inst=0; \ } \ }; \ template<> \ struct multi_index_helper<__VISUALIZE_MIC_INDEX_DEF(Type)> { \ typedef __VISUALIZE_MIC_INDEX_DEF(Type) index_node_type; \ /* mimic index_node_base<Type::value_type> layout */ \ Type::value_type value; \ /* ensure that the debugger sees this template instance */ \ static void get() { \ multi_index_helper<index_node_type> *ensure_inst=0; \ } \ }; #endif ;------------------------------- ; boost::multi_index_container ; boost::multi_index::detail::header_holder<boost::multi_index::detail::sequenced_index_node<*>,*>{ preview(#("multi_index_container data")) children( #( #list( head : *(((boost::multi_index::detail::sequenced_index_node_impl*)(boost::multi_index::detail::sequenced_index_node_trampoline<$T1>*)($c.member))->next_), size : ((multi_index_helper_3<$T2>*)&$c)->node_count, next : next_ ) : *(multi_index_helper<boost::multi_index::detail::sequenced_index_node<$T1> >*)(boost::multi_index::detail::sequenced_index_node<$T1>*)(boost::multi_index::detail::sequenced_index_node_trampoline<$T1>*)(&$e) ) ) } multi_index_helper<*> { preview(#($c.value)) children(#(value:$c.value)) } boost::multi_index::multi_index_container<*,*,*>{ preview(#($c.node_count, " nodes")) children( #( data:((multi_index_helper_2<boost::multi_index::multi_index_container<$T1,$T2,$T3> >*)&$c)->header, original members: [$c,!] ) ) }

1 1

[lambda] return statement wrapper
by Johan Nilsson 31 May '07

31 May '07

Hi, does a return statement wrapper exist in Boost.Lambda? I'd like to be able to do something conceptually similar to the following: -- #include <boost/lambda/bind.hpp> #include <boost/lambda/lambda.hpp> #include <cassert> namespace lambda = boost::lambda; void bar(int) { // whatever } void foo() { int const shouldBe3 = ( lambda::bind(&bar, 1), lambda::bind(&bar, 2), lambda::return(3) // Yes, this conflicts with the regular "return" statement, but you get the idea )(); assert(3 == shouldBe3); } --- I know I can roll my own implementation or work around it, but I was just wondering if something like that already exists. Couldn't locate anything in the docs - did I miss something? / Johan

2 2

unhandled exception (assertion) in Boost.Regexp regex_formatter with VS2005
by Andreas Sedlmeier 31 May '07

31 May '07

Hi. I am a relative newbie to two things: - Boost C++ libraries - Working with the standard C++ libraries and this is my first post on this list. Hello to everyone - I hope you'll forgive me if following question is simply stupid. I cant't even tell that. I currently need to wrap some of the functionality string algorithms and Boost.Regexp offers in Windows DLLs; the developemnt environment I use is VS2005. I started with the samples given in string algorithms which offer a different interface to some of the Boost.Regexp library. I slightly modified the sample and basically now have something like this: using namespace std; using namespace boost; string strInput ("abc__(456)__123__(123)__cde"); regex expression = "[0-9]+" string strReplaceWith("#$1#") string strResult = replace_all_regex_copy(str, expression, strReplaceWith ); This certainly does not make that much sense; its however user input I need to cope with and my problem is that I get an unhandled exception here which I cannot catch. Failure is exactly here (regex_format.hpp): template <class OutputIterator, class Results, class traits> void basic_regex_formatter<OutputIterator, Results, traits>::put(const sub_match_type& sub) { typedef typename sub_match_type::iterator iterator_type; iterator_type i = sub.first; while(i != sub.second) { put(*i); ++i; } } in the line "while(i != sub.second)" I saw that the MS standard C++ libraries compares the iterator types here. And, maybe because in my sample both - i and sub.second - are NULL here, it asserts with "string iterators incompatible". Does anybody know how to solve this problem ? Am I doing something completely wrong or is this maybe a incompatibility of Boost C++ with the MS standard C++ library and other compilers would behave different for my sample ? Any help or insight greatly appreciated. Andreas.

2 1

Specifying library locations for g++
by Scott Meyers 31 May '07

31 May '07

I've built Boost 1.34 for VC8 and g++ 4.1.1 on WinXP. I'm testing my installation with a simple program that uses Boost.FileSystem and Boost.Regex. Compilation is no problem, and for VC8, all I have to do to get things to link is add the location of the Boost binaries to the LIB environment variable. I was hoping that all I had to do for g++ was do the same for the LIBRARY_PATH environment variable (per http://gcc.gnu.org/onlinedocs/gcc/Environment-Variables.html) but I get unresolved symbol errors during linking even after doing that. I played around with the -l option (per http://gcc.gnu.org/onlinedocs/gcc/Link-Options.html#Link-Options) but nothing I tried worked. I have the following relevant binaries in my Boost binary directory: libboost_regex-mgw41-1_34.a libboost_regex-mgw41-d-1_34.a libboost_regex-mgw41-mt-1_34.a libboost_regex-mgw41-mt-d-1_34.a libboost_regex-mgw41-mt-s-1_34.a libboost_regex-mgw41-mt-sd-1_34.a libboost_regex-mgw41-s-1_34.a libboost_regex-mgw41-sd-1_34.a boost_filesystem-mgw41-1_34.a boost_filesystem-mgw41-d-1_34.a boost_filesystem-mgw41-mt-1_34.a boost_filesystem-mgw41-mt-d-1_34.a libboost_filesystem-mgw41-1_34.a libboost_filesystem-mgw41-d-1_34.a libboost_filesystem-mgw41-mt-1_34.a libboost_filesystem-mgw41-mt-d-1_34.a libboost_filesystem-mgw41-mt-s-1_34.a libboost_filesystem-mgw41-mt-sd-1_34.a libboost_filesystem-mgw41-s-1_34.a libboost_filesystem-mgw41-sd-1_34.a Can somebody explain or point me to an explanation of what I need to do to get a program using Boost binaries to link? I'm hoping there's something as simple as with VC8 where I can just specify a directory to look in and have the linker magically figure everything out from there. I'd like to avoid explicitly having to add a new library to link against each time I start using a new compiled Boost library in a project. Incidentally, in the above list of files, I notice that for Boost.FileSystem, there are some archives that start with lib and some that do not, while for regex, all start with lib. Is this meaningful? Thanks, Scott

9 23

Getting a "direct" iterator from Pointer Container
by Lost 30 May '07

30 May '07

Hi I'm using boost pointer container. Is it possible to get a "direct" iterator from Pointer Container? That's the iterator get form a normal stl iterator that points to the pointer? Thanks.

2 1

[UUID] Review results
by Hartmut Kaiser 30 May '07

30 May '07

Hi all, now, as Andy has finished his implementation (here: http://boost-consulting.com/vault/, file: uuid_v8.zip) we can move forward with the results of the review. This review has been kind of special in the sense that - it revealed some serious problems with the library to review (a license incompatibility problem and a security related problem around the random number generator scheme used), - it also spawned lots of dicussions on how to solve these problems. In the end we reached full agreement on what has to be done and Andy agreed to implement the changes. Normally, the mentioned problems would have prevented the acceptance of the library into Boost. OTOH, I think the following facts warrant not to reject the library in the first place: - The issues have already been discussed and a consensus has been reached on how to resolve the problems - Almost all reviewers found the proposed library to be incredibly useful and wanted to see the library accepted given the problems will be resolved - Andy fixed all of the significant problems in a short timeframe The final decision about the acceptance of the library will be made after a mini review of the proposed library (to be scheduled), which should concentrate on the new code presented by Andy and whether the mentioned problems have been appropriately fixed. Regards Hartmut Review Manager

1 0

compiler warnings in boost::serialization
by Os 30 May '07

30 May '07

Hi, I get a huge number of compiler warnings from the serialization and archive libraries. (This is boost 1.34.0 and Visual Studio 2005.) While I can turn them off, I'd rather not have to. The problem ones are: C4100 unreferenced formal parameter C4244 conversion from '...' to '...', possible loss of data C4265 class has virtual functions, but destructor is not virtual and new to 1.34.0 C4189 local variable is initialized but not referenced Can anything be done about this? cheers, os. -- os(a)collective.co.uk http://www.collective.co.uk/ http://www.myspace.com/darkroomtheband

1 0

Re: [Boost-users] [Boost.Test] Embedding unit testing in DLL projects
by Gennadiy Rozental 30 May '07

30 May '07

> "Tanner Davis" <tanner.davis(a)innovision.com> wrote in message > news:6704.32919073409$1180477224@news.gmane.org... > Hello- Hi, Tanner > What is the best practice for embedding test cases into library (versus > executable) projects? I have successfully used the Boost.Test framework > to build and > create EXE targeted, testing specific (MSVS) projects. However, much of > what I need to unit-test is compiled into a Win32 DLLs, and I would much > rather > embed the test cases directly within that DLL project code > (inside define blocks of course), rather than having to define a separate > project to produce an EXE > to execute the tests for each DLL. When I say "I would rather", it > really means my manager demands it. It appears that the boost.test MACROs > nicely > obfuscate the normal main()-ish entry point for the testing. I have > searched for information on overriding or extending the kick-off of > whatever calls the > init_unit_test_suite() so I could write my own custom entry points for > testing the DLL as a post-build step, but have not found what I am looking > for. Boost 1.34.0 introduced new Boost.Test variant usage specifically for this case. It's called "external test runner usage variant". The idea is to build your test module as DLL (define BOOST_TEST_DYN_LINK during compilation). Now you can run the test modult using any external test runner tool. Boost.Test supply simple command line test runner: boost_test_runner utility (it's located in libs/test). What it does essenciallty is it loads DLL dynamically, init the framework and test tree and run all the registerred test cases. > So, my question is this. Is there a provided framework for testing code > within a DLL project? And if not, is there a way to kick start the > testing from a non- > main method (say a custom DLL entry point), and would this strategy work > within the confines of Boost.Test? I believe that other well known unit > testing > framework (which shall remain nameless) supports such a thing, and I would > suspect Boost.Test does also. > > I apologize if this question is often asked or obviously documented > elsewhere. > Thank you for the assistance. The upcoming docs will cover this usage vasriants in more details. HTH, Gennadiy

1 0

[regex] ECMAScript conformance
by Armel Asselin 30 May '07

30 May '07

[boost 1.34.0, with last update to regex sources on 28th february 2007] Hello, first of all I must tell that I am very happy with boost.regex, it's fantastic. though, there are a few inconsistencies though with the ECMAScript spec. 1. it seems that re-matching an atom does not first clear its sub-matches, as "step 4" in page 135 of ECMA 262 3rd edition tells to. that is, it does what is in the 'and not' part here: /(z)((a+)?(b+)?(c))*/.exec("zaacbbbcac") which returns the array ["zaacbbbcac", "z", "ac", "a", undefined, "c"] and not ["zaacbbbcac", "z", "ac", "a", "bbb", "c"] 2. the {n,m} notation should not accept spaces in ECMAScript mode (though it's practical, it's not conformant). I easily admit that its not really important. Best regards Armel

3 2

problem with building program_options (VC8)
by Sergey Skorniakov 30 May '07

30 May '07

I have build boost libraries with command line bjam --toolset=msvc --user-config=config32.jam --stagedir=. debug-symbols=on stage File config32.jam is: using msvc : 8.0 : : <cxxflags>"-wd4996 -wd4103 -Zp4" <define>_CRT_SECURE_NO_DEPRECATE <define>_CRT_NONSTDC_NO_DEPRECATE <define>_SECURE_SCL=0 <define>BOOST_SPIRIT_THREADSAFE <define>BOOST_THREAD_USE_DLL <define>BOOST_THREAD_BUILD_DLL ; The resulting libboost_program_options-vc80-mt-1_34.lib has size 7,420,264 bytes and my program, builded with /D "_CRT_SECURE_NO_DEPRECATE" /D "_CRT_NONSTDC_NO_DEPRECATE" /D "_SECURE_SCL=0" /D "BOOST_THREAD_USE_DLL", crashes. When I changed config32.jam to using msvc : 8.0 : : <cxxflags>"-wd4996 -wd4103 -Zp4 -D_CRT_SECURE_NO_DEPRECATE -D_CRT_NONSTDC_NO_DEPRECATE -D_SECURE_SCL=0 -DBOOST_SPIRIT_THREADSAFE -DBOOST_THREAD_USE_DLL -DBOOST_THREAD_BUILD_DLL " ; The resulting libboost_program_options-vc80-mt-1_34.lib size is 7,233,568 bytes and my program works well. This looks like <define> in config file sometimes has no effect. Furthermore, program options library is alignment-sensitive.

2 2

callbacks to function template (solved)
by Ares Lagae 30 May '07

30 May '07

With the help of the boost lists I finally solved the problem of callbacks to a function template. I will describe what exactly the problem was, and how I have solved it, in the hope that it will be usefull for somebody. I am also open for suggestions and comments. Problem description ------------------- I am building a set of parsers. The parser classes pass data to the application using callbacks. Because there are potentially a lot of callbacks, and some of the callbacks are optional, I wrap each callback into a std::tr1::function object, and then pass it to the parser. The problem I came across can be simplified and abstracted to parsing a format like this one. bool 1 char 2 short 3 int 4 long 5 float 6.0 double 7.0 Each line contains a type and is followed by a value of that type. This data should be parsed and passed to the application. The types are fixed (bool, char, short, int, long, float, double). (In the actual real problem, there are more types, and "declarations" involving several types.) Solution 1 ---------- The first solution (see listing_1.cpp) registers a callback for each data type. Because each data type is handled in a similar way, there is a lot of code repetition. Solution 2 ---------- The second solution (see listing_2.cpp and variable_list.hpp) replaces the callback function by a function template. Instead of having a member variable for the callback for each data type, the member variables are grouped into a "variable_list". This is a class template that, for each type in a type list, keeps a variable. The code repetition in the parser class is eliminated, but the user still has to call "set_callback" to register a callback for each of the data types. Solution 3 ---------- The third solution (see listing_3.cpp) eliminates the similar code at the user side. Instead of calling "set_callback" to register a callback for each of the data types, a functor is constructed, that for each type constructs a std::tr1::function object that wraps the callback. That function object is passed to the parser, and the parser sets all the callbacks. This again results in similar code in the parser "set_callbacks". Solution 4 ---------- The fourth solution (see listing_4.cpp) eliminates the similar code in the parser "set_callbacks" by executing a functor for each type in the type list. This code is approximately equivalent to callbacks to a function template. Best regards, -- Ares Lagae Computer Graphics Research Group, Katholieke Universiteit Leuven http://www.cs.kuleuven.be/~ares/

2 2

filesystem/convenience.hpp build warning
by Os 30 May '07

30 May '07

Having just moved to v1.34.0, I now get a build warning from this file. boost/filesystem/convenience.hpp:46: warning C4245: 'argument' : conversion from 'int' to 'boost::filesystem::system_error_type', signed/unsigned mismatch Is this a bug? The line in question is boost::throw_exception( basic_filesystem_error<Path>( "boost::filesystem::create_directories", ph, -1 ) ); Changing it to boost::throw_exception( basic_filesystem_error<Path>( "boost::filesystem::create_directories", ph, (boost::filesystem::system_error_type)-1 ) ); seems to fix it. cheers, os. -- os(a)collective.co.uk http://www.collective.co.uk/ http://www.myspace.com/darkroomtheband

1 0

Why even use Boost preprocessor library over my own code generator ?
by Keriask 30 May '07

30 May '07

Hi. I find the 2 pros given in the docs to use the preprocessor library not valid! I was about to use the preprocessor library for the 1st time but after seing and studying some of the preprocessed output from various compilers it killed the only apparently valid argument the preprocessor docs present themselves to use it: integration and read to run. Obviously integration can be made easily with the #line and #error directives, wich will even allow my custom code generator to have syntax error checking. The other argument is not really true since a simple text processor can be made portable very easily, hey Linux does it with much more complicated stuff and it works most times. BTW, I'm not trying to bash it by any means(the work and effort put there are amazing), I just want to find a good reason to use it myself which unfortunately I haven't so far. Best regs Jose ____________________________________________________________________________________Be a better Globetrotter. Get better travel answers from someone who knows. Yahoo! Answers - Check it out. http://answers.yahoo.com/dir/?link=list&sid=396545469

8 12

Re: [Boost-users] Cross compiling boost
by Peter Poulsen 30 May '07

30 May '07

On Ons, Maj 30, 2007 15:23, Hughes, James wrote: > Check the archive of boost messages - I posted something on this last > year sometime....Google should find it. > > James Yes, and it probably works fine with 1.33, but with the new 1.34 something has changed (for instance you don't seem to be required to give the "-sTOOL=gcc" argument anymore) I have seen a couple of examples on how to cross compile 1.33, but none of them seems to work with 1.34. /peter > > > > ====== > Q. What is the most annoying thing on usenet and in e-mail? > A: Bottom-posting > Q: Why is it such a bad thing? > A: Because it takes ages to actually find what you are looking for in > the pages of unnecessary cak when all you wanted was the answer above. > > > >> -----Original Message----- >> From: boost-users-bounces(a)lists.boost.org >> [mailto:boost-users-bounces@lists.boost.org] On Behalf Of >> Peter Poulsen >> Sent: 30 May 2007 08:48 >> To: boost-users(a)lists.boost.org >> Subject: [Boost-users] Cross compiling boost >> >> >> I'm trying to build boost for an arm processor, but all the >> hidden bjam stuff is simply to much, so I have to ask for help :) >> >> All my build tools are in ~/arm-tools/. I'm on a Linux x86 >> and the cross compiler I'm trying to use is a gcc version 3.4.4. >> >> I have tried various build commands. The latest being: >> >> >> ~/boost_1_34_0 $ bjam --build-dir=/tmp/build-boost >> --toolset-gcc-3.4.4 --with-regex -sGXX=arm-wrs-linux-gnueabi-g++ stage >> >> >> But I still get: >> >> >> ~/boost_1_34_0 $ file stage/lib/libboost_regex-gcc34.so >> stage/lib/libboost_regex-gcc34.so: ELF 32-bit LSB shared >> object, Intel 80386, version 1 (SYSV), stripped >> >> Can anybody explain to me what I must do to compile >> boost-regex for arm? >> >> btw I have found others trying to do the same but they seem to end up hacking the >> bjam files, and I have no idea of how to that. >> >> /peter >> >> >> -- >> A: Because it messes up the order in which people normally read text. >> Q: Why is it such a bad thing? >> A: Top-posting. >> Q: What is the most annoying thing on usenet and in e-mail? >> >> >> >> >> _______________________________________________ >> Boost-users mailing list >> Boost-users(a)lists.boost.org >> http://lists.boost.org/mailman/listinfo.cgi/boost-users >> >> > > This message (including any attachments) contains confidential > and/or proprietary information intended only for the addressee. Any unauthorized > disclosure, copying, distribution or reliance on the contents of this information is > strictly prohibited and may constitute a violation of law. If you are not the > intended recipient, please notify the sender immediately by responding to this > e-mail, and delete the message from your system. If you have any questions about > this e-mail please notify the sender immediately. > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > > -- A: Because it messes up the order in which people normally read text. Q: Why is it such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?

1 0

Cross compiling boost
by Peter Poulsen 30 May '07

30 May '07

I'm trying to build boost for an arm processor, but all the hidden bjam stuff is simply to much, so I have to ask for help :) All my build tools are in ~/arm-tools/. I'm on a Linux x86 and the cross compiler I'm trying to use is a gcc version 3.4.4. I have tried various build commands. The latest being: ~/boost_1_34_0 $ bjam --build-dir=/tmp/build-boost --toolset-gcc-3.4.4 --with-regex -sGXX=arm-wrs-linux-gnueabi-g++ stage But I still get: ~/boost_1_34_0 $ file stage/lib/libboost_regex-gcc34.so stage/lib/libboost_regex-gcc34.so: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), stripped Can anybody explain to me what I must do to compile boost-regex for arm? btw I have found others trying to do the same but they seem to end up hacking the bjam files, and I have no idea of how to that. /peter -- A: Because it messes up the order in which people normally read text. Q: Why is it such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?

2 1

[mpl] call functor for each type in a type list
by Ares Lagae 30 May '07

30 May '07

I want to call a functor for each type in a type list. Consider the following program. #include <iostream> #include <boost/mpl/apply.hpp> #include <boost/mpl/empty_base.hpp> #include <boost/mpl/fold.hpp> #include <boost/mpl/identity.hpp> #include <boost/mpl/inherit.hpp> #include <boost/mpl/joint_view.hpp> #include <boost/mpl/lambda.hpp> #include <boost/mpl/pair.hpp> #include <boost/mpl/transform.hpp> #include <boost/mpl/vector.hpp> template <typename T, typename Functor> class functor_list_member { public: functor_list_member() { Functor f; f.template operator()<T>(); } }; template<typename Sequence, typename Functor> class functor_list: private boost::mpl::fold< Sequence, boost::mpl::empty_base, boost::mpl::inherit< boost::mpl::_1, functor_list_member<boost::mpl::_2, Functor> > >::type {}; typedef boost::mpl::vector<bool, char, short, int, long, float, double> types; struct functor { template <typename T> void operator()() { std::cout << __PRETTY_FUNCTION__ << std::endl; } }; typedef functor_list<types, functor> my_functor_list; int main(int argc, char* argv) { my_functor_list(); } The program outputs void functor::operator()() [with T = bool] void functor::operator()() [with T = char] void functor::operator()() [with T = short int] void functor::operator()() [with T = int] void functor::operator()() [with T = long int] void functor::operator()() [with T = float] void functor::operator()() [with T = double] However, I want to create a functor object first, and pass it to each functor_list_member, like this: template <typename T, typename Functor> class functor_list_member { public: functor_list_member(Functor f) { f.template operator()<T>(); } }; template<typename Sequence, typename Functor> class functor_list: private boost::mpl::fold< Sequence, boost::mpl::empty_base, boost::mpl::inherit< boost::mpl::_1, functor_list_member<boost::mpl::_2, Functor> > >::type { public: functor_list(Functor f) /* how do I call base(f) ? */ { } }; typedef boost::mpl::vector<bool, char, short, int, long, float, double> types; struct functor { template <typename T> void operator()() { std::cout << __PRETTY_FUNCTION__ << std::endl; } }; typedef functor_list<types, functor> my_functor_list; int main(int argc, char* argv) { functor f; my_functor_list(f); } but I do not know how to call the constructor of the base class in functor_list::functor_list(Functor f). (see comment /* how do I call base(f) ? */). Can someone help me? Best regards, -- Ares Lagae Computer Graphics Research Group, Katholieke Universiteit Leuven http://www.cs.kuleuven.be/~ares/

2 2

help with fusion filter_view
by Christian Holmquist 30 May '07

30 May '07

Hello, I'm trying to create a fusion::filter_view with a predicate that must access, for each element, the element at the same position in an 'external' typelist. Let's say I have two sequences like this typedef mpl::vector<int&> mpl_vec; typedef fusion::vector<int> f_vec; Now, I would like a view into f_vec of, for instances, all references in mpl_vec. typedef fusion::filter_view<f_vec, is_reference<mpl::_> > f_view; This is obviously wrong; is_reference should evaluate with mpl_vec as its source. If I wrap is_reference like this template<class Sequence> struct is_reference_in { template<class T> struct apply : boost::is_reference<T> // Wrong, T should be extracted from Sequence { }; }; and create the view like typedef fusion::filter_view<f_vec, is_reference_in<mpl_vec> > f_view2; I gain access to the mpl_vec in my predicate, but since the predicate is immutable I cannot iterate over the Sequence template parameter to extract the type to send to is_reference. The way i'm solving this now is to store everything in the same vector, like typedef fusion::vector<fusion::pair<int, int&> > f_vec and create predicates that work with fusion::pair instead. This seems clumsy. Could zip_view help me out here? Any other suggestions? Thanks, Christian

1 0

function template as template parameter
by Ares Lagae 30 May '07

30 May '07

consider the following program: template <typename T> void function_template(T value) { std::cout << __PRETTY_FUNCTION__ << std::endl; } struct function_template_wrapper { template <typename T> static void call(T value) { function_template<T>(value); } }; int main(int argc, char* argv[]) { int i; function_template_wrapper::call<int>(i); } Is it possible to somehow parameterize the struct function_template_wrapper such that I can pass function_template as a parameter? E.g. function_template_wrapper<function_template>::call<int>(i); would call function_template<int>(i), and function_template_wrapper<other_function_template>::call<int>(i); would call other_function_template<int>(i) Underlying problem: I want to somehow pass a name of a function template to a class, that will instantiate the function template for a number of types (known at compile time), and store these instantiations in function objects, for later use. Best regards, -- Ares Lagae Computer Graphics Research Group, Katholieke Universiteit Leuven http://www.cs.kuleuven.be/~ares/

3 6

Serialization Performance
by Andrew VanUitert 30 May '07

30 May '07

I am using Boost Serialization as a basic RPC mechanism on an embedded Linux system, however it seems that the serialization library uses a lot of CPU cycles to process some rather simple messages. Each of the messages I am using passes a pointer to an object that is a serializable object representing the needed parameters for each type of message. As each message/command is passed from one system to another, the pointer is filled in with the de-marshaled parameters. So far this has worked splendidly except for the CPU utilization. The embedded processor is an Arm running at 300 MHz, and when I send it repeated messages the CPU utilization pegs. Profiling has revealed that the embedded processor is spending all its time in serialization library. For the most part, the message being sent back and forth are arrays of chars under 255 in size. Are there some general guidelines when considering performance with this library ? For instance, I read that the performance of 1.34 increased substantially with the use of a streambuf and a binary archive for arrays of basic types. However, the constructor that takes a streambuf is only available in the binary archive, and I am using a text archive to avoid ending and alignment issues. Is the use of a pointer for the parameters a place for optimization ? Andy

2 1

[iostreams] Are streams really models of Device?
by Chad Walters 30 May '07

30 May '07

Standard stream and stream buffers are supposed to be models of the Device concept (http://www.boost.org/libs/iostreams/doc/concepts/device.html) However, I am finding that their treatment by various parts of the library doesn't seem to be consistent. For example, streams don't actually seem to be usable everywhere a Device can be used. For example, if I define SinkOrStream as follows: using namespace boost::iostreams; typedef file_sink SinkOrStream; Then the following declarations compile properly: SinkOrStream mySink(mFileName); stream<composite<gzip_compressor, SinkOrStream> > out1(gzip_compressor(), mySink); stream<SinkOrStream> out2(mFileName); However, neither out1 nor out2 compile properly if I define SinkOrStream as: typedef stream<file_sink> SinkOrStream; If I understand the documentation correctly, these should both be legal, no? Does this work for others? I also find it a little strange that streams are supposed to be models of Device and they have a close() method, yet they do not support the closable_tag. Is this by design? Thanks for any help on clarifying these issues... Chad Walters

1 0

Any existing base64 encoding class/function in boost ?
by Raja Cherukuri 29 May '07

29 May '07

Is there a base64 encoding class/function in boost 1.33.x ? Thanks in Advance. - Raja

1 0

[Boost.Test] Embedding unit testing in DLL projects
by Tanner Davis 29 May '07

29 May '07

Hello- What is the best practice for embedding test cases into library (versus executable) projects? I have successfully used the Boost.Test framework to build and create EXE targeted, testing specific (MSVS) projects. However, much of what I need to unit-test is compiled into a Win32 DLLs, and I would much rather embed the test cases directly within that DLL project code (inside define blocks of course), rather than having to define a separate project to produce an EXE to execute the tests for each DLL. When I say "I would rather", it really means my manager demands it. It appears that the boost.test MACROs nicely obfuscate the normal main()-ish entry point for the testing. I have searched for information on overriding or extending the kick-off of whatever calls the init_unit_test_suite() so I could write my own custom entry points for testing the DLL as a post-build step, but have not found what I am looking for. So, my question is this. Is there a provided framework for testing code within a DLL project? And if not, is there a way to kick start the testing from a non-main method (say a custom DLL entry point), and would this strategy work within the confines of Boost.Test? I believe that other well known unit testing framework (which shall remain nameless) supports such a thing, and I would suspect Boost.Test does also. I apologize if this question is often asked or obviously documented elsewhere. Thank you for the assistance.

1 0

[Building 1.34] Linking against 1.33.1 and/or 1.34
by Christoph Duelli 29 May '07

29 May '07

Is it possible to configure Boost so that the library files built (linux, gcc 4.1) are *not* named like My problem: Boost 1.33.1 libs are a.t.m. named like libboost_regex-gcc.so Boost 1.34.0 libs are now named like libboost_regex-gcc41.so Consequently, my makefile that contain lines like -lboost_regex-gcc do not work with 1.34. If I 'fix' them for 1.34 my other machines (and colleagues) will be in trouble til they switch to 1.34 as well. Question: Is it possible to build boost (or write my makefiles) in a way so that my projects build with both Boost 1.33.1, 1.34 and so on (whatever is present)? Is there a standard or best practice for that? Thank you and best regards Christoph

4 3

[Boost.graph] compilation error for dijkstra_shortest_paths
by Nicolas Ray 29 May '07

29 May '07

Hello all, I have developed an application using the dijkstra_shortest_paths function of the Graph library of version 1.31 of Boost. When upgrading to Boost version 1.34, I have the following compilation error under Borland C++ Builder 6: [C++ Error] relaxed_heap.hpp(52): E2247'relaxed_heap<unsigned int, indirect_cmp<float*,less<float>>,vec_adj_list_vertex_id_map<VertexProperty, unsigned int>>::largest_key' is not accessible If I comment the call to the "dijkstra_shortest_paths" function in my code, there is no compilation error anymore. Is it related to how the default vertex_index is defined? If I add and use a custom vertex_index in the vertex property map, I get the same compilation error. I have tried to compile with version 1.33 and add the same error. I defined my graph as follow: struct vertex_altitude_t {typedef vertex_property_tag kind;}; typedef property<vertex_altitude_t, float> VertexProperty; typedef property<edge_weight_t, float > EdgeProperty; typedef adjacency_list<vecS, vecS, undirectedS, VertexProperty, EdgeProperty> Graph; Any help would be greatly appreciated, thank you very much, nicolas

2 2

Working with 64 bit dynamic bitsets....(uncertain of my method)
by Graham Reitz 29 May '07

29 May '07

Can anyone suggest a more efficient method to break apart a 64 bit dynamic bitset into 32 bit unsigned int chunks? The conversion is needed because eventually a C style system IO call needs to be performed that can't deal with a boost::dynamic_bitset or std::bitset; Thanks for any help/advice, Graham This works, but it feels inefficient: #include <iostream> #include "boost/dynamic_bitset.hpp" using namespace std; using namespace boost; // Create a 64 bit dynamic bitset from two 32 bit words, not 64 bit portable, illustration only dynamic_bitset<> create_64_bitset(unsigned int ms_word, unsigned int ls_word); int main(int argc, char* argv[]) { dynamic_bitset<> bitset_64( create_64_bitset(static_cast<unsigned>(0xaaaaaaaa), static_cast<unsigned>(0xbbbbbbbb))); cout << bitset_64 << endl; // This whole portion, below, seems inefficient. Is it? Is there a better method? // Break 64 bit dynamic bitset into 32 bit words dynamic_bitset<> temp_bitset(bitset_64); // Get least sig. 32 bits unsigned int ls_word = ((bitset_64 <<= 32) >>= 32).to_ulong(); // Get most sig 32 bits unsigned int ms_word = (temp_bitset >>= 32).to_ulong(); cout << ls_word << endl << ms_word << endl; cin >> argc; return 0; } dynamic_bitset<> create_64_bitset(unsigned int ms_word, unsigned int ls_word) { // Create a 64 bitset dynamic_bitset<> most_sig_word(64, ms_word); most_sig_word <<= 32; dynamic_bitset<> least_sig_word(64, ls_word); // Combine the two into one 64 bit bitset most_sig_word |= least_sig_word; return most_sig_word; }

1 0

[filesystem] dirent64* compilation error on PGI
by Server Levent Yilmaz 29 May '07

29 May '07

I am trying to build Boost.Filesystem (release 1.34) on a GNU/Linux x86_64 system with Portland Group compilers (there is no official support for this compiler). Everything else seems to work nice and dandy, except this single error with compiling operations.cpp: "operations.cpp", line 1247: error: a value of type "dirent64 *" cannot be assigned to an entity of type "dirent *" if ( (p = ::readdir( dirp )) == 0 ) ^ If I comment out the following defines in operations.cpp:22 and operations.cpp:25 #define _FILE_OFFSET_BITS 64 #define __USE_FILE_OFFSET64 the error goes away. But I doubt this is the correct way to fix this. Any ideas? thanks Levent -- Server Levent Yilmaz Mechanical Engineering University of Pittsburgh

1 0

[date time] Something leaks in operator<<
by Seweryn Habdank-Wojewódzki 29 May '07

29 May '07

Hi I do not know if date time leaks or not, but I want to clarify that before I will send information to libstdc++ bugzilla. Regards. The code: #include <iostream> #include <boost/date_time/gregorian/gregorian.hpp> int main() { using namespace boost::gregorian; using namespace std; date b1 (2007,2,3); date b2 (2007,2,7); cout << b1 << '\n'; // line 12 cout << b2 << '\n'; // line 13 date_period p1 (b1, b2); cout << p1 << '\n'; // line 15 } And the result of: valgrind --leak-check=full --show-reachable=yes ./main ==2382== Memcheck, a memory error detector. ==2382== Copyright (C) 2002-2007, and GNU GPL'd, by Julian Seward et al. ==2382== Using LibVEX rev 1732, a library for dynamic binary translation. ==2382== Copyright (C) 2004-2007, and GNU GPL'd, by OpenWorks LLP. ==2382== Using valgrind-3.2.3, a dynamic binary instrumentation framework. ==2382== Copyright (C) 2000-2007, and GNU GPL'd, by Julian Seward et al. ==2382== For more details, rerun with: -v ==2382== 2007-Feb-03 2007-Feb-07 [2007-Feb-03/2007-Feb-06] ==2382== ==2382== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 7 from 1) ==2382== malloc/free: in use at exit: 856 bytes in 26 blocks. ==2382== malloc/free: 58 allocs, 32 frees, 2,772 bytes allocated. ==2382== For counts of detected errors, rerun with: -v ==2382== searching for pointers to 26 not-freed blocks. ==2382== checked 107,996 bytes. ==2382== ==2382== 20 bytes in 1 blocks are still reachable in loss record 1 of 7 ==2382== at 0x4021FB4: operator new(unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x8049CDD: std::locale::locale<boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > > >(std::locale const&, boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >*) (locale_classes.h:568) ==2382== by 0x804F4E2: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:57) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== ==2382== 48 bytes in 1 blocks are still reachable in loss record 2 of 7 ==2382== at 0x4021C34: operator new[](unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x40AE6B5: std::locale::_Impl::_Impl(std::locale::_Impl const&, unsigned) (in /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/libstdc++.so.6.0.8) ==2382== by 0x8049CF9: std::locale::locale<boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > > >(std::locale const&, boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >*) (locale_classes.h:568) ==2382== by 0x804F4E2: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:57) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== ==2382== 80 bytes in 2 blocks are still reachable in loss record 3 of 7 ==2382== at 0x4021FB4: operator new(unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x804C497: std::vector<std::string, std::allocator<std::string> >::_M_insert_aux(__gnu_cxx::__normal_iterator<std::string*, std::vector<std::string, std::allocator<std::string> > >, std::string const&) (new_allocator.h:90) ==2382== by 0x804CB25: boost::date_time::date_generator_formatter<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >::date_generator_formatter() (stl_vector.h:610) ==2382== by 0x804ED2E: boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >::date_facet(unsigned) (date_facet.hpp:86) ==2382== by 0x804F4B5: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:56) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== ==2382== 112 bytes in 1 blocks are still reachable in loss record 4 of 7 ==2382== at 0x4021FB4: operator new(unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x804F4A5: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:56) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== ==2382== 128 bytes in 1 blocks are still reachable in loss record 5 of 7 ==2382== at 0x4021C34: operator new[](unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x40AE262: std::locale::_Impl::_M_install_facet(std::locale::id const*, std::locale::facet const*) (in /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/libstdc++.so.6.0.8) ==2382== by 0x8049D13: std::locale::locale<boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > > >(std::locale const&, boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >*) (locale_classes.h:571) ==2382== by 0x804F4E2: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:57) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== ==2382== 128 bytes in 1 blocks are still reachable in loss record 6 of 7 ==2382== at 0x4021C34: operator new[](unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x40AE1EF: std::locale::_Impl::_M_install_facet(std::locale::id const*, std::locale::facet const*) (in /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/libstdc++.so.6.0.8) ==2382== by 0x8049D13: std::locale::locale<boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > > >(std::locale const&, boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >*) (locale_classes.h:571) ==2382== by 0x804F4E2: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:57) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== ==2382== 340 bytes in 19 blocks are possibly lost in loss record 7 of 7 ==2382== at 0x4021FB4: operator new(unsigned) (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==2382== by 0x40F08F3: std::string::_Rep::_S_create(unsigned, unsigned, std::allocator<char> const&) (in /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/libstdc++.so.6.0.8) ==2382== by 0x40F1A94: (within /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/libstdc++.so.6.0.8) ==2382== by 0x40F1C48: std::string::string(char const*, std::allocator<char> const&) (in /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/libstdc++.so.6.0.8) ==2382== by 0x804EC4B: boost::date_time::date_facet<boost::gregorian::date, char, std::ostreambuf_iterator<char, std::char_traits<char> > >::date_facet(unsigned) (date_facet.hpp:86) ==2382== by 0x804F4B5: std::basic_ostream<char, std::char_traits<char> >& boost::gregorian::operator<< <char, std::char_traits<char> >(std::basic_ostream<char, std::char_traits<char> >&, boost::gregorian::date const&) (gregorian_io.hpp:56) ==2382== by 0x8049B01: main (main.cpp:12) ==2382== ==2382== LEAK SUMMARY: ==2382== definitely lost: 0 bytes in 0 blocks. ==2382== possibly lost: 340 bytes in 19 blocks. ==2382== still reachable: 516 bytes in 7 blocks. ==2382== suppressed: 0 bytes in 0 blocks. -- |\/\/| Seweryn Habdank-Wojewódzki \/\/

3 4

BOOST: interprocess library
by dhruva 29 May '07

29 May '07

Hi, I would like to know if the "interprocess" library can work with STL packaged with Visual Studio. My current understanding is that "interprocess" uses a modified STL to use shared memory and this is packaged with "interprocess". Since a lot of code we have uses STL from VS (MS Visual Studio), mixing might not be an option. Is there a clean workaround path or is it a known limitation? with best regards, dhruva -- Dhruva Krishnamurthy Contents reflect my personal views only!

2 1

Bellman-Ford Example
by Cigdem Gueler 29 May '07

29 May '07

Hallo,I have sent one more email previously. I 've got my little example for which Bellman-Ford fails to detect an existing negative cycle. This occurs after I modify the graph!I would be glad if you could help me. Bellman-Fors shortest paths is a very important routine for my code, which I have to call several times. And it has to work reliably so that the rest of the code also does...Thank you.cheers,Cigdem _________________________________________________________________ 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…

2 1

Bellman-Ford shortest paths
by Cigdem Gueler 29 May '07

29 May '07

Hi,I have a little problem about Bellman Ford shortest paths algorithm. When I call this algorithm for the graphs with parallel edges, sometimes it fails to recognize the negative cycles. Doesn't it work for graphs with parallel edges? Or am I doing a mistake in my code? Well, the code is huge to send it to the mailing list. I am trying to find a little example for which this phenomenon occurs!But I still would like to be confirmed that Bellman Ford should have actually worked for graphs with parallel edges. It is not really mentioned in the documentation if it works or not!!Thank you,Cigdem _________________________________________________________________ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx

2 1

Re: [Boost-users] [date time] Something leaks in operator<<
by Seweryn Habdank-Wojewódzki 29 May '07

29 May '07

Hi > If you want to try to confirm it as a leak - change your cout references > to a [non-static] stream you declare directly in your program scope (not > cout, cerr, clog). You should then either see the "possible" move to > "leaked" or go away entirely. OK. thanks for suggestion to day I will check it, and put here results :-). Regards. -- |\/\/| Seweryn Habdank-Wojewódzki \/\/

1 0

Suppress generation of unversioned .lib
by Braden McDaniel 29 May '07

29 May '07

Is it possible to suppress generation of the unversioned .lib file when building on Windows? Braden

6 12

[iostreams] access to specialized stream attributes?
by Christopher Lux 28 May '07

28 May '07

Hi, i am trying to get access to specialized stream attributes in my own filters. I created a specialized stringstream class which holds some kind of log level, which i want to use in my input filters. Therefore i wrote a input device which handles my specialized class. With debug outputs i see that this device is used to read my stream. Now i want to create a filter which is able to access the log level attribute from the specialized stream. I tried to specialize the template function for reading from the source for my stream device and for the specialized stream but none of these functions are getting used. Is there a way to be able to access either the device or the source stream in the filters? The example code below gives a short example of the mentioned problem. #include <iostream> #include <sstream> #include <string> #include <boost/iostreams/concepts.hpp> #include <boost/iostreams/operations.hpp> #include <boost/iostreams/filtering_stream.hpp> #include <boost/iostreams/char_traits.hpp> #include <boost/iostreams/stream.hpp> // specialized stream class class console_out_stream : public std::stringstream { public: console_out_stream() : std::stringstream() {} virtual ~console_out_stream() {} int get_log_level() const { return (_log_level); } void set_log_level(int l) { _log_level = l; } private: int _log_level; }; // ability to set log level using the << operator on the new stream class class log_level { public: log_level(int l) : _log_level(l) {} int _log_level; }; console_out_stream& operator << (console_out_stream& con_str, const log_level& log) { con_str.set_log_level(log._log_level); return (con_str); } // source device class my_console_source : public boost::iostreams::source { public: my_console_source(console_out_stream& con) : _console(con) {} std::streamsize read(char* s, std::streamsize n) { std::cout << "ran through the source" << std::endl; return (boost::iostreams::read(_console, s, n)); } const console_out_stream& get_console() const {return _console; } protected: console_out_stream& _console; }; // input filter // just filters out the character s class test_in_filter_m : public boost::iostreams::multichar_input_filter { public: explicit test_in_filter_m() {} template<typename Source> std::streamsize read(Source& src, char* s, std::streamsize n) { for (std::streamsize z = 0; z < n; ++z) { int c; while (true) { if ((c = boost::iostreams::get(src)) == EOF) return z != 0 ? z : -1; else if (c == boost::iostreams::WOULD_BLOCK) return z; if (c != 's') break; } s[z] = c; } return n; } template<> std::streamsize read<my_console_source>(my_console_source& src, char* s, std::streamsize n) { std::cout << "ran through the source filter: my_console_source" << std::endl; return (n); } template<> std::streamsize read<console_out_stream>(console_out_stream& src, char* s, std::streamsize n) { std::cout << "ran through the source filter: console_out_stream" << std::endl; return (n); } }; int main(int argc, char **argv) { console_out_stream console; console << log_level(20) << "sick, sad world!"; console << log_level(5) << "from mtv daria" << std::endl; boost::iostreams::stream<my_console_source> in_stream_buf(console); boost::iostreams::filtering_istream out_stream; out_stream.push(test_in_filter_m()); out_stream.push(in_stream_buf); std::string line; while (std::getline(out_stream, line)) { std::cout << line << std::endl; } return (0); } -chris -- Christopher Lux | | Bauhaus University Weimar | Faculty of Media - Virtual Reality Systems Group

1 0

[Boost.Signals] bug
by Yashin Evgeny 28 May '07

28 May '07

Here is, minimal example. struct Conn : public boost::signals::trackable { void operator()(){} }; void f() { boost::signal<void ()> sig; Conn c; sig.connect( c ); c = Conn(); // crash inside.. } The reason is simple. During clear() method of std::list, there are signals::connection destructor calls are happening, which, in turn, calling erase method of the same std::list, and as a result we're getting _DEBUG_ERROR("list erase iterator outside range"); I guess, the bug is with "dying" variable: in destructor, it was set to true, and then signals::connection destructors are not trying to remove themself from the std::list. But in boost::signals::trackable::operator=() there is no dying flag marking. Evgeny.

1 0

[ptr_vector] vector<ptr_vector < T > >
by Seweryn Habdank-Wojewódzki 28 May '07

28 May '07

Hi! Can anyone tell me why std::vector < boost::ptr_vector < T > > can not be reserved and resized I can not push_back anything inside vector, due to compile errors? I am using ptr_vector for storing copyable, but nonassignable objects of type T - in fact objects contains mainly const elements. Regards. #include <iostream> #include <vector> #include <boost/ptr_container/ptr_vector.hpp> class Foo { public: Foo ( int const i, int const & j ) :i_(i), j_(j) {} Foo ( Foo const & f ) : i_(f.i_), j_(f.j_) {} private: int const i_; int const & j_; void operator= ( Foo const & f ); }; int main() { using namespace std; using namespace boost; int k = 10; ptr_vector<Foo> g; g.push_back ( new Foo(1,k) ); vector<ptr_vector<Foo> > d; //d.push_back ( g ); <-- Errors //d.resize(5) //d.reserve(5) } -- |\/\/| Seweryn Habdank-Wojewódzki \/\/

6 9

Boost 1.34.0 Builds *Almost* Perfectly - 4 Failures with BJam [Regex/ICU-related]
by Lawrence Spector 28 May '07

28 May '07

Any ideas? I think they're related to ICU. I downloaded ICU 3.6 (the latest). However, I am using this with VS2005, which was "rarely tested" according to the website. Still, after converting the ICU project to VS2005 and building, I had no build errors, but I do get these link errors with boost, which all seem to be ICU-related. See below: -------------- D:\3rdparty\boost_1_34_0>call "C:\Program Files\Microsoft Visual Studio 8\VC\vcvarsall.bat" x86 Setting environment for using Microsoft Visual Studio 2005 x86 tools. Building Boost.Regex with Unicode/ICU support enabled Using ICU in D:\3rdparty\icu4c-3_6/include ...patience... ...patience... ...found 5406 targets... ...updating 10 targets... msvc.link.dll bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.dll bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.lib Creating library bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.lib and object bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.exp icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: virtual __thiscall icu_3_6::Locale::~Locale(void)" (__imp_??1Locale@icu_3_6@@UAE@XZ) referenced in function __unwindfunclet$??0icu_regex_traits_implementation@re_detail@boost@@QAE@ ABVLocale@icu_3_6@@@Z$0 icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: static class icu_3_6::Collator * __cdecl icu_3_6::Collator::createInstance(class icu_3_6::Locale const &,enum UErrorCode &)" (__imp_?createInstance@Collator@icu_3_6@@SAPAV12@ABVLocale@2@AAW4UErrorC ode@@@Z) referenced in function "public: __thiscall boost::re_detail::icu_regex_traits_implementation::icu_regex_traits_impl ementation(class icu_3_6::Locale const &)" (??0icu_regex_traits_implementation@re_detail@boost@@QAE@ABVLocale@icu_3 _6@@@Z) icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: __thiscall icu_3_6::Locale::Locale(class icu_3_6::Locale const &)" (__imp_??0Locale@icu_3_6@@QAE@ABV01@@Z) referenced in function "public: __thiscall boost::re_detail::icu_regex_traits_implementation::icu_regex_traits_impl ementation(class icu_3_6::Locale const &)" (??0icu_regex_traits_implementation@re_detail@boost@@QAE@ABVLocale@icu_3 _6@@@Z) icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: __thiscall icu_3_6::Locale::Locale(void)" (__imp_??0Locale@icu_3_6@@QAE@XZ) referenced in function "public: __thiscall boost::icu_regex_traits::icu_regex_traits(void)" (??0icu_regex_traits@boost@@QAE@XZ) icu.obj : error LNK2019: unresolved external symbol _u_tolower_3_6 referenced in function "public: unsigned __int64 __thiscall boost::icu_regex_traits::lookup_classname(int const *,int const *)const " (?lookup_classname@icu_regex_traits@boost@@QBE_KPBH0@Z) icu.obj : error LNK2019: unresolved external symbol _u_toupper_3_6 referenced in function "public: int __thiscall boost::icu_regex_traits::toupper(int)const " (?toupper@icu_regex_traits@boost@@QBEHH@Z) icu.obj : error LNK2019: unresolved external symbol _u_digit_3_6 referenced in function "public: bool __thiscall boost::icu_regex_traits::isctype(int,unsigned __int64)const " (?isctype@icu_regex_traits@boost@@QBE_NH_K@Z) icu.obj : error LNK2019: unresolved external symbol _u_isspace_3_6 referenced in function "public: unsigned __int64 __thiscall boost::icu_regex_traits::lookup_classname(int const *,int const *)const " (?lookup_classname@icu_regex_traits@boost@@QBE_KPBH0@Z) icu.obj : error LNK2019: unresolved external symbol _u_charFromName_3_6 referenced in function "public: class std::vector<int,class std::allocator<int> > __thiscall boost::icu_regex_traits::lookup_collatename(int const *,int const *)const " (?lookup_collatename@icu_regex_traits@boost@@QBE?AV?$vector@HV?$allocato r@H@std@@@std@@PBH0@Z) icu.obj : error LNK2019: unresolved external symbol _u_isblank_3_6 referenced in function "public: bool __thiscall boost::icu_regex_traits::isctype(int,unsigned __int64)const " (?isctype@icu_regex_traits@boost@@QBE_NH_K@Z) icu.obj : error LNK2019: unresolved external symbol _u_charType_3_6 referenced in function "public: bool __thiscall boost::icu_regex_traits::isctype(int,unsigned __int64)const " (?isctype@icu_regex_traits@boost@@QBE_NH_K@Z) bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.dll : fatal error LNK1120: 11 unresolved externals call "C:\Program Files\Microsoft Visual Studio 8\VC\vcvarsall.bat" x86 >nul link /NOLOGO /INCREMENTAL:NO /DLL /DEBUG /subsystem:console /out:"bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex -vc80-mt-gd-1_34.dll" /IMPLIB:"bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_re gex-vc80-mt-gd-1_34.lib" @"bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc8 0-mt-gd-1_34.dll.rsp" if %ERRORLEVEL% NEQ 0 EXIT %ERRORLEVEL% if exist "bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80 -mt-gd-1_34.dll.manifest" ( mt -nologo -manifest "bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80 -mt-gd-1_34.dll.manifest" "-outputresource:bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\ boost_regex-vc80-mt-gd-1_34.dll;2" ) ...failed msvc.link.dll bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.dll bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.lib... ...removing bin.v2\libs\regex\build\msvc-8.0\debug\threading-multi\boost_regex-vc80- mt-gd-1_34.lib ...skipped <pstage\lib>boost_regex-vc80-mt-gd-1_34.dll for lack of <pbin.v2\libs\regex\build\msvc-8.0\debug\threading-multi>boost_regex-vc8 0-mt-gd-1_34.dll... ...skipped <pstage\lib>boost_regex-vc80-mt-gd-1_34.lib for lack of <pbin.v2\libs\regex\build\msvc-8.0\debug\threading-multi>boost_regex-vc8 0-mt-gd-1_34.lib... ...skipped <pstage\lib>boost_regex-vc80-mt-gd.lib for lack of <pstage\lib>boost_regex-vc80-mt-gd-1_34.lib... msvc.link.dll bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.dll bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.lib Creating library bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.lib and object bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.exp icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: __thiscall icu_3_6::Locale::Locale(class icu_3_6::Locale const &)" (__imp_??0Locale@icu_3_6@@QAE@ABV01@@Z) referenced in function "public: class icu_3_6::Locale __thiscall boost::re_detail::icu_regex_traits_implementation::getloc(void)const " (?getloc@icu_regex_traits_implementation@re_detail@boost@@QBE?AVLocale@i cu_3_6@@XZ) icu.obj : error LNK2019: unresolved external symbol _u_tolower_3_6 referenced in function "public: int __thiscall boost::icu_regex_traits::translate_nocase(int)const " (?translate_nocase@icu_regex_traits@boost@@QBEHH@Z) icu.obj : error LNK2019: unresolved external symbol _u_toupper_3_6 referenced in function "public: int __thiscall boost::icu_regex_traits::toupper(int)const " (?toupper@icu_regex_traits@boost@@QBEHH@Z) icu.obj : error LNK2019: unresolved external symbol _u_digit_3_6 referenced in function "public: int __thiscall boost::icu_regex_traits::value(int,int)const " (?value@icu_regex_traits@boost@@QBEHHH@Z) icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: __thiscall icu_3_6::Locale::Locale(void)" (__imp_??0Locale@icu_3_6@@QAE@XZ) referenced in function "public: class icu_3_6::Locale __thiscall boost::icu_regex_traits::getloc(void)const " (?getloc@icu_regex_traits@boost@@QBE?AVLocale@icu_3_6@@XZ) icu.obj : error LNK2019: unresolved external symbol _u_isspace_3_6 referenced in function "public: bool __thiscall boost::icu_regex_traits::isctype(int,unsigned __int64)const " (?isctype@icu_regex_traits@boost@@QBE_NH_K@Z) icu.obj : error LNK2019: unresolved external symbol _u_isblank_3_6 referenced in function "public: bool __thiscall boost::icu_regex_traits::isctype(int,unsigned __int64)const " (?isctype@icu_regex_traits@boost@@QBE_NH_K@Z) icu.obj : error LNK2019: unresolved external symbol _u_charType_3_6 referenced in function "public: bool __thiscall boost::icu_regex_traits::isctype(int,unsigned __int64)const " (?isctype@icu_regex_traits@boost@@QBE_NH_K@Z) icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: virtual __thiscall icu_3_6::Locale::~Locale(void)" (__imp_??1Locale@icu_3_6@@UAE@XZ) referenced in function __unwindfunclet$??0icu_regex_traits_implementation@re_detail@boost@@QAE@ ABVLocale@icu_3_6@@@Z$0 icu.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: static class icu_3_6::Collator * __cdecl icu_3_6::Collator::createInstance(class icu_3_6::Locale const &,enum UErrorCode &)" (__imp_?createInstance@Collator@icu_3_6@@SAPAV12@ABVLocale@2@AAW4UErrorC ode@@@Z) referenced in function "public: __thiscall boost::re_detail::icu_regex_traits_implementation::icu_regex_traits_impl ementation(class icu_3_6::Locale const &)" (??0icu_regex_traits_implementation@re_detail@boost@@QAE@ABVLocale@icu_3 _6@@@Z) icu.obj : error LNK2019: unresolved external symbol _u_charFromName_3_6 referenced in function "public: class std::vector<int,class std::allocator<int> > __thiscall boost::icu_regex_traits::lookup_collatename(int const *,int const *)const " (?lookup_collatename@icu_regex_traits@boost@@QBE?AV?$vector@HV?$allocato r@H@std@@@std@@PBH0@Z) bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.dll : fatal error LNK1120: 11 unresolved externals call "C:\Program Files\Microsoft Visual Studio 8\VC\vcvarsall.bat" x86 >nul link /NOLOGO /INCREMENTAL:NO /DLL /subsystem:console /out:"bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_reg ex-vc80-mt-1_34.dll" /IMPLIB:"bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_ regex-vc80-mt-1_34.lib" @"bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-v c80-mt-1_34.dll.rsp" if %ERRORLEVEL% NEQ 0 EXIT %ERRORLEVEL% if exist "bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc 80-mt-1_34.dll.manifest" ( mt -nologo -manifest "bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc 80-mt-1_34.dll.manifest" "-outputresource:bin.v2\libs\regex\build\msvc-8.0\release\threading-mult i\boost_regex-vc80-mt-1_34.dll;2" ) ...failed msvc.link.dll bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.dll bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.lib... ...removing bin.v2\libs\regex\build\msvc-8.0\release\threading-multi\boost_regex-vc8 0-mt-1_34.lib ...skipped <pstage\lib>boost_regex-vc80-mt-1_34.dll for lack of <pbin.v2\libs\regex\build\msvc-8.0\release\threading-multi>boost_regex-v c80-mt-1_34.dll... ...skipped <pstage\lib>boost_regex-vc80-mt-1_34.lib for lack of <pbin.v2\libs\regex\build\msvc-8.0\release\threading-multi>boost_regex-v c80-mt-1_34.lib... ...skipped <pstage\lib>boost_regex-vc80-mt.lib for lack of <pstage\lib>boost_regex-vc80-mt-1_34.lib... ...failed updating 4 targets... ...skipped 6 targets... -------------------- Thanks in advance for any help that can be provided, Lawrence

3 7

[1.34.0] building on cygwin fails out of the box
by Andy Stevenson 28 May '07

28 May '07

All, Trying to build 1.34.0 out of the box on Cygwin yields the following errors when trying to install it. Ideas? ./tools/jam/src/bin.cygwinx86/bjam --user-config=user-config.jam --prefix=/usr/local --exec-prefix=/usr/local --libdir=/usr/local/lib --includedir=/usr/local/include install Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (don't panic: this is a strictly optional feature). /boost/boost_1_34_0/tools/build/v2/build/virtual-target.jam:996: in virtual-target.register-actual-name from module virtual-target error: Duplicate name of actual target: <p/usr/local/lib>exact.dll.a error: previous virtual target { common%common.hard-link-exact.dll.a.IMPORT_LIB { common%common.copy-boost_signals-gcc41-mt-d-1_34.dll.a.IMPORT_LIB { gcc%gcc.link.dll-boost_signals-gcc41-mt-d-1_34.dll.a.IMPORT_LIB { gcc%gcc.compile.c++-trackable.o.OBJ { trackable.cpp.CPP } } { gcc%gcc.compile.c++-connection.o.OBJ { connection.cpp.CPP } } { gcc%gcc.compile.c++-named_slot_map.o.OBJ { named_slot_map.cpp.CPP } } { gcc%gcc.compile.c++-signal_base.o.OBJ { signal_base.cpp.CPP } } { gcc%gcc.compile.c++-slot.o.OBJ { slot.cpp.CPP } } } } } error: created from ./install-unversioned error: another virtual target { common%common.hard-link-exact.dll.a.IMPORT_LIB { common%common.copy-boost_prg_exec_monitor-gcc41-mt-d-1_34.dll.a.IMPORT_LIB { gcc%gcc.link.dll-boost_prg_exec_monitor-gcc41-mt-d-1_34.dll.a.IMPORT_LIB { gcc%gcc.compile.c++-execution_monitor.o.OBJ { execution_monitor.cpp.CPP } } { gcc%gcc.compile.c++-cpp_main.o.OBJ { cpp_main.cpp.CPP } } } } } error: created from ./install-unversioned error: added properties: <define>BOOST_TEST_DYN_LINK=1 error: removed properties: <define>BOOST_SIGNALS_DYN_LINK=1 <define>BOOST_SIGNALS_NO_LIB=1 /boost/boost_1_34_0/tools/build/v2/build/virtual-target.jam:459: in actualize-no-scanner from module object(file-target)@16510 /boost/boost_1_34_0/tools/build/v2/build/virtual-target.jam:111: in object(file-target)(a)16510.actualize from module object(file-target)@16510 /boost/boost_1_34_0/tools/build/v2/build-system.jam:476: in load from module build-system /boost/boost_1_34_0/tools/build/v2/kernel/modules.jam:261: in import from module modules /boost/boost_1_34_0/tools/build/v2/kernel/bootstrap.jam:132: in boost-build from module /boost/boost_1_34_0/boost-build.jam:9: in module scope from module Not all Boost libraries built properly. Andy Stevenson

7 12

[Boost.Program_Options] Assertion failed on a Windows Form Application
by JD 28 May '07

28 May '07

Hi, When I create a "Windows Forms Application" under Visual C++ Express and add these lines to the main .cpp file: #include <boost/program_options.hpp> static boost::program_options::variables_map vm; and then set the the project property 'Common Language Runtime support' to /clr (/clr:pure make compilation fails), I got an assertion failure on dbgheap.c line 1414, expression: _CrtIsValidHeapPointer(pUserData) on the line: BOOST_PROGRAM_OPTIONS_DECL std::string arg("arg"); in value_semantic.cpp. Compiler command line: /Od /I "G:\Program Files\boost-cvs\boost" /D "WIN32" /D "_DEBUG" /D "_UNICODE" /D "UNICODE" /FD /EHa /MDd /Yu"stdafx.h" /Fp"Debug\po_assert.pch" /Fo"Debug\\" /Fd"Debug\vc80.pdb" /W3 /nologo /c /Zi /clr /TP /errorReport:prompt /FU "c:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\System.dll" /FU "c:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\System.Data.dll" /FU "c:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\System.Drawing.dll" /FU "c:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\System.Windows.Forms.dll" /FU "c:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\System.XML.dll" Linker command line: /OUT:"G:\Projects\po_assert\Debug\po_assert.exe" /INCREMENTAL /NOLOGO /LIBPATH:"G:\Program Files\boost-cvs\boost\stage\lib" /MANIFEST /MANIFESTFILE:"Debug\po_assert.exe.intermediate.manifest" /DEBUG /ASSEMBLYDEBUG /PDB:"g:\projects\po_assert\debug\po_assert.pdb" /SUBSYSTEM:WINDOWS /ENTRY:"main" /MACHINE:X86 /FIXED:No /ERRORREPORT:PROMPT Using a pointer and dynamically allocation a variables_map ends the same way... What am I doing wrong ? Or is it a known problem ? Thank you. JD

2 1

Re: [Boost-users] [iostreams] Issue with closing of filters [PATCH]
by Chad Walters 28 May '07

28 May '07

gchen <chengang31(a)gmail.com> wrote: > It seems like a issue with close_impl<closable_tag>::close(...). > ... > (link: http://www.boost.org/libs/iostreams/doc/functions/close.html) Yes, that looks like the right answer. The code is clearly not performing what the document describes. When I fix this up, my code works as expected. Proposed patch attached. >> I think the wfilter versions of these typedefs should use >> multichar_wfilter<...>. Agreed? > > A copy-n-paste bug !? Yep, that's my take -- proposed patch attached. > My $0.02 Worth much more than that... many thanks! Chad Walters

2 1

auto-link problem
by Rob Agar 27 May '07

27 May '07

hi all, Apologies if this question has been asked a million times before, but having searched the archives and documentation I'm none the wiser. Basically the problem is this: the auto-linker is trying to link to the wrong boost::filesystem lib and I can't figure out out to fix it. The error from MS Visual Studio 2005 is "LINK : fatal error LNK1104: cannot open file 'libboost_filesystem-vc80-mt-gd-1_34.lib'". Fair enough because this file does not exist. I want it to use the dll import lib boost_filesystem-vc80-mt-gd-1_34.lib. This file does exist, and is in a directory on the MSVC lib search path, along with the matching dll. So the question is, how do I tell the auto-linker to link to the file I want, or if this is not possible, how do I switch off auto-linking? I would rather not recompile the boost libs if I can help it. Thanks Rob

3 4

[date time] are months and years derived from date_duration?
by Seweryn Habdank-Wojewódzki 27 May '07

27 May '07

Hi I have simple problem if months and years derived from date_duration? The goal is that user can dynamically choose a time of events that appear periodically in shedule e.g. every 20th of any month is a meeting or every 3 months something other happens or birthsday is one per year. Regards. -- |\/\/| Seweryn Habdank-Wojewódzki \/\/

2 5

[iostreams] Issue with closing of filters (esp. tee_filter)
by Chad Walters 27 May '07

27 May '07

I am running into an issue with iostreams related to the closing of filters (tee_filter in particular) and I think I really need a little help. Thanks in advance to anyone patient enough to read through the following. I created a multichar_output_filter called ChecksumFilter. It outputs a crc for every 4K of data streamed into it. I am using in conjunction with a tee_filter to stream out a file of checksums in parallel with another file stream. Unfortunately, ChecksumFilter::close() is not invoked at times when I believe it should be, resulting in the checksum for the final chunk of data not getting pushed out. The following code shows the problem: typedef boost::iostreams::composite<ChecksumFilter, boost::iostreams::file_sink> ChecksumComposite; char buffer1[BufferSize]; memset(buffer1, DefaultValue, sizeof(buffer1)); std::string fileName1(tmpnam(0)); std::string fileName2(tmpnam(0)); boost::iostreams::file_sink sink1(fileName1.c_str()); boost::iostreams::file_sink sink2(fileName2.c_str()); ChecksumFilter chksum; ChecksumComposite comp(chksum, sink2); boost::iostreams::tee_filter<ChecksumComposite> tee(comp); { boost::iostreams::filtering_stream<boost::iostreams::output> ostream; ostream.push(tee); ostream.push(sink1); result = boost::iostreams::write(ostream, buffer1, sizeof(buffer1)); } // ChecksumFilter::close() is not invoked here!!! ChecksumFilter itself is pretty basic -- it is simply a multichar_output_filter: class ChecksumFilter : public boost::iostreams::multichar_output_filter { public: ChecksumFilter(void) : mImpl(new details::ChecksumFilterImpl()) { // Nothing to do here } template<typename Sink> std::streamsize write(Sink& sink, const char* s, std::streamsize n) { return mImpl->write<Sink>(sink, s, n); } template<typename Sink> void close(Sink& sink) { mImpl->close<Sink>(sink); } private: boost::shared_ptr<details::ChecksumFilterImpl> mImpl; }; I don't think the details of ChecksumFilterImpl are relevant. I spent some time tracing through the closing code path. I'm sure this comes as no surprise to anyone but it's quite a twisty maze. I will be the first to admit that my understanding of the library is incomplete, but here is something that I think is a bit odd -- somewhere along the way between the code in closer.hpp and tee.hpp, the openmode gets converted from BOOST_IOS::out to BOOST_IOS::in | BOOST_IOS::out. >From closer.hpp: template<> struct close_impl<closable_tag> { template<typename T> static void close(T& t, BOOST_IOS::openmode which) { printf("close closable_tag 1\n"); typedef typename category_of<T>::type category; const bool in = is_convertible<category, input>::value && !is_convertible<category, output>::value; if (in == ((which & BOOST_IOS::in) != 0)) { t.close(); } } template<typename T, typename Sink> static void close(T& t, Sink& snk, BOOST_IOS::openmode which) { typedef typename category_of<T>::type category; const bool in = is_convertible<category, input>::value && !is_convertible<category, output>::value; if (in == ((which & BOOST_IOS::in) != 0)) { non_blocking_adapter<Sink> nb(snk); t.close(nb); } } }; >From tee.hpp: template<typename Device> class tee_filter : public detail::basic_adapter<Device> { public: ... template<typename Next> void close( Next&, BOOST_IOS::openmode which = BOOST_IOS::in | BOOST_IOS::out ) { iostreams::close(this->component(), which); } ... }; The original openmode at the start of the close chain is BOOST_IOS::out, but the close_impl's two close methods discard that openmode when they call t.close() or t.close(nb). This means that tee_filter's close() sets openmode to its default value of BOOST_IOS::in | BOOST_IOS::out. This mode is passed along to iostreams::close(), which ends up back in close_impl close() again as it is attempting to close the ChecksumFilter. However, now the check "if (in == ((which & BOOST_IOS::in) != 0))" fails, and so ChecksumFilter.close() is never invoked. If I change the default value for openmode in tee_filter::close() to BOOST_IOS::out, everything works OK for me, but I am not really sure that this is the right thing to do. It appears that some kind of custom closing in closer.hpp could be done for tee_filters that would pass through the openmode, much like the handling for classes that support the two_sequence tag (sorry if I am jumbling terminology a bit here). But I am not sure my understanding of the overall code is strong enough to do the right thing here either. Can anyone help me out or put me in touch with someone who can? Much thanks! Chad Walters PS: And if you read this far, you probably understand the iostreams library more than well enough to confirm whether the possible (likely?) bug I reported earlier this week is really a bug (nobody has replied as of yet): At the bottom of concepts.hpp, the following typedefs appear: typedef multichar_filter<input> multichar_input_filter; typedef multichar_filter<input> multichar_input_wfilter; typedef multichar_filter<output> multichar_output_filter; typedef multichar_filter<output> multichar_output_wfilter; typedef multichar_filter<dual_use> multichar_dual_use_filter; typedef multichar_filter<dual_use> multichar_dual_use_wfilter; I think the wfilter versions of these typedefs should use multichar_wfilter<...>. Agreed?

2 1

Re: [Boost-users] Boost-users Digest, Vol 1279, Issue 2
by Yaoltzin Gomez 26 May '07

26 May '07

Thank you for you time to answer. Allthough your answer has helped me, I still have the bad_weak_ptr exception. Every time I try to access the shared_from_this pointer from inside Base member functions, I get that exception. What I want to know is how can I access safely an without errors, the shared_from_this pointer inside my Base class, when I create a smart pointer instance of my Derived class. In derived I don't have that problem (I can access shared_from_this) because this is the class the instance is created. Thanks in advance Peter Dimov wrote: > Yaoltzin Gomez wrote: > > > Basically, what I have is this: > > > > class Base : public boost::enable_shared_from_this<Base> > > { > > boost::shared_ptr<Base> GetBasePointer() > > { > > return shared_from_this(); > > } > > > > void Foo() > > { > > shared_ptr<Child> child( new Child ); > > AddChild( child, GetBasePointer() ); > > } > > }; > > > > class Derived: public Base, public > > boost::enable_shared_from_this<Derived> { > > boost::shared_ptr<Derived> GetDerivedPointer() > > { > > return shared_from_this(); > > } > > [...] > > One way to avoid putting more than one enable_shared_from_this<> base into > Derived is to use > > class Derived: public Base > { > boost::shared_ptr<Derived> GetDerivedPointer() > { > return dynamic_pointer_cast<Derived>( shared_from_this() ); > } > > // ... > }; > > static_pointer_cast will also work in this case (it will fail if Derived > uses virtual inheritance). >

2 1

Yet another boost::enable_shared_from_this question
by Yaoltzin Gomez 26 May '07

26 May '07

Hello everyone Here's my problem. I have a small set of classes (I'm working on an API). Some of the classes use inheritance hierarchy. Sometimes I need to use (well lots of times) the "this" pointer and work with raw pointers inside each class. However, I would like to work with smart pointers. So I need to change the signature of my member functions to accept smart pointers, but now I need an "this" smart pointer equivalent. I derived the classes from boost::enable_shared_from_this. Basically, what I have is this: class Base : public boost::enable_shared_from_this<Base> { boost::shared_ptr<Base> GetBasePointer() { return shared_from_this(); } void Foo() { shared_ptr<Child> child( new Child ); AddChild( child, GetBasePointer() ); } }; class Derived: public Base, public boost::enable_shared_from_this<Derived> { boost::shared_ptr<Derived> GetDerivedPointer() { return shared_from_this(); } void Bar() { AnotherFunction( GetDerivedPointer() ); ... } }; Now, in the main function I have: int main(...) { boost::shared_ptr<Derived> derived( new Derived ); derived->Foo(); derived->Bar(); } Obviously, this causes a bad_weak_ptr exception, because I'm creating a Derived instance, and the shared_from_this called inside Foo throws the exception, but I need to call this functions defined in the Base class. In other words, I need access to the shared_from_this, in Base and Derived, and right now I can't do this. I would like to know if it's possible to use shared_from_this in this way, or not, and if not, what other options do I have (besides using raw pointers) This is a simplified problem, in my code, generally I have more than 2 leves of class inheritance, so things get more complicated. I hope I explained myself correctly. Thanks in advance.

2 1

[BJam/Boost.Python] Issues Building with Boost.Build v2 using Boost 1.34.0
by Lawrence Spector 26 May '07

26 May '07

I successfully was building Boost.Python extensions using bjam version 1 in Boost 1.33.1. Since the upgrade, I converted my Jamfile to Jamfile.v2 and am getting the following errors when I try to build: bjam -sTOOLS=vc-8_0 -sBOOST_ROOT=C:\3rdparty\boost_1_34_0 -sBUILD=debug error: Could not find parent for project at '.' error: Did not find Jamfile or project-root.jam in any parent directory. Here's my Jamfile: # This is the top of our own project tree project-root ; import python ; extension MyLib # Declare a Python extension : MyLib.cpp # source # requirements and dependencies for Boost.Python extensions <template>@boost/libs/python/build/extension ## Requirements ## : <include>$(MyDir)/Include <define>_ATL_DLL <define>_AFXDLL <define>_UNICODE <define>UNICODE <runtime-link>dynamic ; -------------------------------- Here's my Jamrules: path-global BOOST_ROOT : C:/3rdparty/boost_1_34_0 ; project boost : $(BOOST_ROOT) ; -------------------------------- Any idea how to correct this? Thanks in advance, Lawrence

2 3

bjam define=_SECURE_SCL shows no effect
by Christopher Lux 26 May '07

26 May '07

hi, after reading through this thread [1] i tried the following to build boost::program options: bjam define=_SECURE_SCL=0 define=_TESTING_SOMTHING=1 --toolset=msvc-8.0 --with-program_options variant=release to test the effect i added the folowing to the store function in program_options: #if _SECURE_SCL == 1 std::cout << "_SECURE_SCL == 1" << std::endl; #else std::cout << "_SECURE_SCL == 0" << std::endl; #endif #if _TESTING_SOMTHING == 1 std::cout << "_TESTING_SOMTHING == 1" << std::endl; #else std::cout << "_TESTING_SOMTHING == 0" << std::endl; #endif the strange thing now is, that _TESTING_SOMTHING is correctly set to 1 but _SECURE_SCL still is set to 1 and not to 0. did i miss something or is there another way to get this set properly to disable all runtime checks for release builds? [1] http://thread.gmane.org/gmane.comp.lib.boost.devel/143980/focus=144063 thx -chris -- Christopher Lux | | Bauhaus University Weimar | Faculty of Media - Virtual Reality Systems Group

8 19

Re: [Boost-users] [date_time] UNICODE and wcout problems
by fang.wenbiao＠mail.scut.edu.cn 26 May '07

26 May '07

atoi(beg->c_str()); ------------------------ From: Kirit S忙lensminde <kirit.saelensminde(a)gmail.com> To: boost-users(a)lists.boost.org Sent: 2007-05-26 11:35 Subject: Re: [Boost-users] [date_time] UNICODE and wcout problems >Filip Konvi膷ka wrote: >> Hi, >> >> in MSVC 8.0 _UNICODE build, after I send a ptime to wcout, I can no >> longer print "international" characters. When I use a temporary >> wostringstream for printing the ptime, everything is OK. Minimal repro >> see below (the "2" at the end is never printed). With some characters >> like the "拧" in the example, the output is totally cut off; with others, >> like "谩", the codepage is changed, so the characters are displayed >> incorrectly. >> >> Any suggestions? > >This is a problem in the MSVC libraries. If you print a character above >code 255 then the stream crashes and is good for nothing afterwards. >Only std::wstringstream doesn't have this problem. > >I think if you buy the Dinkumware libraries this works. The _cputws >function does work as you would expect, but it can't be piped. The >behaviour also changes if you run the program from a command shell with >Unicode turned on ( cmd.exe /u). > >I did talk to PJ Plauger about it on clc. This is what he explained: > >"When you write to a wofstream, the wchar_t sequence you write gets >converted to a byte sequence written to the file. How that conversion >occurs depends on the codecvt facet you choose. Choose none any you get >some default. In the case of VC++ the default is pretty stupid -- the >first 256 codes get written as single bytes and all other wide-character >codes fail to write. " > >http://groups.google.com/group/comp.lang.c++/browse_thread/thread/3c2032537… > > >Kirit > > > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

Using the block iterator version of the dynamic_bitset constructor?
by Graham Reitz 26 May '07

26 May '07

How do you use the BlockInputIterator version of the dynamic_bitset? template <typename BlockInputIterator> dynamic_bitset(BlockInputIterator first, BlockInputIterator last, const Allocator& alloc = Allocator()) What kind of iterator should I use? Thanks in advance for any help, Graham

1 0

Moving from Loki::Functor to Boost
by Scott Ribe 25 May '07

25 May '07

I have an application that depends pretty heavily on Loki::Functor, including some extensions/fixes I've made. That works great. But I want to use smart pointers more extensively (in particular I have a need for weak references), and so before I go further with advanced template use, I'd like to switch to Boost, since it's being actively developed. Can anybody point me to a reference comparing Loki's Functor's to Boost's function wrappers? I'm ready to dive into "Beyond the C++ Standard Library", but first I'd like to know if there's anything like a guide for moving my current code. (I only use other parts of Loki very lightly, BTW.) In addition to figuring out the straight substitution of one library for the other, I have to cope with the modifications I've made to Functor: - they can be "null" in the sense of not being bound to a function call and throwing an exception if called (so that they can be embedded in a class and assigned a value later) - there are initialization forms, and corresponding macros, that record the file & line where a functor is created, and log invocations - they can wrap method calls to Objective-C objects Heh, the last one reveals a bit about the platform I'm developing for ;-) -- Scott Ribe scott_ribe(a)killerbytes.com http://www.killerbytes.com/ (303) 722-0567 voice

4 5

Question about Boost string tokenizer
by Meryl Silverburgh 25 May '07

25 May '07

Hi, I have the following loop to tokenize a string. But how can I pass the sub-string to ato()? for(boost::tokenizer<>::iterator beg=tok.begin(); beg!=tok.end();++beg){ int value = atoi(*beg); cout << value << "\n"; i++; } I get this error, in my code: : error: cannot convert 'const std::basic_string<char, std::char_traits<char>, std::allocator<char> >' to 'const char*' for argument '1' to 'int atoi(const char*)'

2 2

mkdir -p and cp -r
by Patrick Lacasse 25 May '07

25 May '07

What is the boost::filesystem way to do a system( "mkdir -p somedir" ); or a system( "cp -r somedir_orig somedir_dest" ); ? Should I write a recursive function for that ?

1 0

[Serialization] Thread-unsafe singleton
by Sohail Somani 25 May '07

25 May '07

Robert, I've seen this thread here: http://thread.gmane.org/gmane.comp.lib.boost.user/21885 And I believe I am coming across the same issue (static *serializer instances) manifesting itself as an assertion failure: Assertion failed: lookup(eti) == m_self->m_map.end(), file vendor/boost/libs/serialization/src/extended_type_info.cpp, line 77 That assertion means that a type_info key is being registered twice. I've verified this via cout (\o/ for debugging). Now I am also of the opinion that there is no way to do this initialization without bringing a threading library into Boost.Serialization. From reading the comments in the code, it appears that the only reason the library is using these singletons is because of VC6. Is it possible to have some thread-safe way (perhaps via ifdefs?) to do these singletons? Or do you think there is another reason I am having this issue? Thanks, Sohail PS: I'm using 1.33.1. If the issue is fixed in 1.34, I'd be very glad!

3 13

boost::filesystem::path throws exception: invalid name
by Vivek Nallur 24 May '07

24 May '07

My code looks something like this: try{ fs::path full_path (fs::initial_path()); full_path = fs::system_complete(fs::path(filePath, fs::native)); fs::create_directories(full_path); INFO << "Directory successfully created: " << full_path.native_directory_string() << std::endl; retval = true; }catch(std::exception& ex){ WARN << "Could not create directory: " << filePath << " " << ex.what() << std::endl; } return retval; ========== When filePath is "c:\a\b\", boost throws an exception saying invalid path name c:\a\b\ in path "c:\a\b\" when filePath is "c:\\a\\b\\", boost throws exactly the same exception when filePath is "c:/a/b/", the exception string changes saying invalid path name "c:" in path "c:/a/b/" What am I doing wrong? -Vivek

4 8

[MPL] Runtime traversal of a type sequence until a condition is true
by Scott Meyers 24 May '07

24 May '07

I'm interested in taking Andrei Alexandrescu's Ad Hoc Visitor technique (http://tinyurl.com/2k8h2h) and implementing it using the MPL. I need to implement a runtime function that iterates over an MPL type sequence until a dynamic_cast succeeds. Here's some pseudocode, where I use "typelist" to mean an MPL type sequence: class UnknownType { ... }; // for exceptions template <typename TL> // TL = "typelist" void visit(ASTNode *pn) { for (each type T in TL) { // try dynamic_ if (T *pT = dynamic_cast<T*>(pn)) { // casting to all v.process(pT); // the types in TL return; } throw UnknownType(typeid(*pn)); // throw an ex. if } // no type matches I can make this work if I replace the for loop with a use of mpl::for_each, but for_each iterates over all elements of the type sequence, and I want to stop when a dynamic_cast succeeds. I can hack the function object I pass to for_each to have it be a no-op once a successful dynamic_cast has been encountered, but the operative word there is "hack." What I really want is something more like do_until<Sequence>(unaryFunctionObject) that will loop over all the entries in Sequence until the unaryFunctionObject returns true. Is there something like this in the MPL, must I roll my own, or am I approaching this the wrong way entirely? Thanks, Scott

5 7

removing a subset of bits from a boost::dynamic_bitset...
by Graham Reitz 24 May '07

24 May '07

What is an efficient method using dynamic_bitset to do the following? bitset1<16> = 0xabcd mask<16> = 0x000f result<4> = (bitset1 & mask) What is the syntax to remove a subset of bits from a dynamic bitset? Thanks, Graham

1 0

BOOST_FOREACH and iterating by reference
by Daniel Gehriger 24 May '07

24 May '07

I tried BOOST_FOREACH on a slightly modified version of the sample in the documentation. Instead of iterating by value, I iterate by reference: extern std::vector<float> get_vector_float(); BOOST_FOREACH( float& f, get_vector_float() ) { } When compiling with VisualC++ 8.0, I receive the following error message: error C2440: 'initializing' : cannot convert from 'const float' to 'float &' Conversion loses qualifiers Is there a work-around ? - Daniel -- View this message in context: http://www.nabble.com/BOOST_FOREACH-and-iterating-by-reference-tf3806531.ht… Sent from the Boost - Users mailing list archive at Nabble.com.

3 7

[program_options] unknown options in config files
by Ram V 24 May '07

24 May '07

Hi, This is a feature request for the program_options library. A clean and extensible design for my code stipulates that it will be best to pass the user input (command line / config file) around to different modules which then extract the relevant options using the program_options library. This will mean neglecting all options that are irrelevant to the current module. However, it appears that this is possible only from the command line and not from the config file. Ideally, it will be great if the api for parsing config files mirrors what exists for the command line. That would mean a basic_config_file_parser class with chainable allow_unregistered() and collect_unrecognized() methods for the config file parsing. Some searching shows that there have been a couple of similar requests on this list: http://thread.gmane.org/gmane.comp.lib.boost.user/25035 http://thread.gmane.org/gmane.comp.lib.boost.user/24398 <http://thread.gmane.org/gmane.comp.lib.boost.user/24398> and on the developers' list: http://thread.gmane.org/gmane.comp.lib.boost.devel/150415 http://thread.gmane.org/gmane.comp.lib.boost.devel/86380 <http://thread.gmane.org/gmane.comp.lib.boost.devel/150415> It also appears that a issue ticket was opened on sourceforge: http://thread.gmane.org/gmane.comp.lib.boost.devel/145941 I am wondering if there are any plans to add support for this soon. If not, any thoughts from the library developers or other users on how they think its best to do-it-myself will be appreciated. Thanks RamV

1 0

Distributing Boost.Program_options
by Chris Thachuk 24 May '07

24 May '07

Hello all, I'm a first time user of boost libraries and I'm finding them extremely useful. I will be incorporating their use into a number of open source research projects. I've searched the archives on how to distribute the boost source with my code. It seems the bcp tool is the way to go for this, however, I'm confused on how to use it for my purpose. I would like to keep the code as portable as possible without users needing to install additional libraries on their own. Since most of the boost libraries are all contained in header files, it seems easy to distribute. However, I always getting linking errors unless I link against a pre-built library for Boost.Program_options. Is it necessary for this library to be built first, then linked against? How could I use the 'bcp' tool to grab the needed Boost dependencies for my project? Can I use a standard make file as the users wouldn't have bjam installed? My apologies if this is covered in an obvious location. I've found the documentation to be helpful, but I haven't found what I've needed yet to accomplish this. Any help would be greatly appreciated. Best, Chris Thachuk Graduate Student Simon Fraser University Vancouver, BC, Canada

3 4

Bug :: bzip2_decompressor
by scan man 24 May '07

24 May '07

I modified the test program shipped in the boost::iostreams to decompress a 1.5 GB file which expand to 6GB. The program exits after writing 900KB of data without throwing exceptions. It is reproduced below, I was told in the irc channel that this was a bug in alloc and will be probably fixed in 1.34.1. -r #include <fstream> #include <iostream> #include <boost/iostreams/filtering_stream.hpp> #include <boost/iostreams/copy.hpp> #include <boost/iostreams/filter/bzip2.hpp> using namespace std; namespace bio = boost::iostreams; int main ( int argc, char ** argv ) { if (argc < 2) { fprintf(stderr, "Usage: %s data_filename\n", argv[0]); exit(0); } ifstream file(argv[1], ios_base::in|ios_base::binary ); if ( !file.good() ) { fprintf(stderr, "Unable to open file%s\n", argv[1]); exit(0); } else { try { bio::filtering_stream < bio::input > in; in.push( bio::bzip2_decompressor () ); in.push( file ); bio::copy ( in, cout ); } catch ( const std::exception& e ) { std::cerr<<"\nCaught std::exception " << ": "<< e.what() << std::endl; } } }

1 0

boost condition deadlock
by jc 24 May '07

24 May '07

hello all i'm new to boost thread. i'm converting my win32 based thread application to boost so that in future i can be portable. this is my code void CVectorCAN::xlVectorThreadByCh(void){ //do something here while(m_bVectorThreadAlive){ WaitForSingleObject(m_hxlEventHandle, 50); xlStatus = XL_SUCCESS; while(!xlStatus){ if(m_bReceiveStatus){ { boost::mutex::scoped_lock lock1(m_AckMutex1); m_bVectorThreadDead = true; m_cond1.notify_one(); } return; } //process data here }//while xl_success }//while the thread is alive m_bVectorThreadDead = true; } int CVectorCAN::KillVectorBoostThread(void){ boost::mutex::scoped_lock lock2(m_AckMutex1); m_bVectorThreadAlive = false; m_bReceiveStatus = true; while(!m_bVectorThreadDead){ m_cond1.wait(lock2); } m_pVectorThread->join(); delete m_pVectorThread; m_pVectorThread = NULL; return 0; } my function deadlocks in the m_cond1.wait(lock2) i have no idea what i'm doing wrong. any pointer in the right direction will be great thanks -- jc (jayachandran kamaraj)

1 0

Re: [Boost-users] [filesystem] is_directory on root throws
by james.jones＠firstinvestors.com 24 May '07

24 May '07

From: "David Neubelt" <david(a)readyatdawn.com> > I checked and verified the folder does exist, I'm guessing the problem > is that the directory is in the format c:folder instead of c:\folder. That most likely is the problem, since those are (probably) different directories. c:folder refers to a directory or file named "folder" in the **current working directory** of C:, whereas c:\folder refers to a directory or file named "folder" in the root of C:. - James Jones Administrative Data Mgmt. Webmaster 375 Raritan Center Pkwy, Suite A Data Architect Edison, NJ 08837

2 2

[Serialisation] Explicit instantiation of serialize when used in library
by Hughes, James 24 May '07

24 May '07

Hello all (but probably mostly Robert) I have a class in a library which has a serialise function. It's defined in the header but implemented in the cpp file. However, I am getting an unresolved error, because when the library is build there is no need to instantiate the serialise function, as its only called by something using the library. So, I have tried to explicitly instantiate the missing function in the cpp file in the library, after the serialise implementation template class LCPanel::serialize<boost::archive::xml_iarchive>(boost::archive::xml_iar chive&, unsigned int) However, this line gives an error Expected ')' before '&' token I'm not sure where I am going wrong here as syntax appears correct (but I am not sure of), any ideas? James This message (including any attachments) contains confidential and/or proprietary information intended only for the addressee. Any unauthorized disclosure, copying, distribution or reliance on the contents of this information is strictly prohibited and may constitute a violation of law. If you are not the intended recipient, please notify the sender immediately by responding to this e-mail, and delete the message from your system. If you have any questions about this e-mail please notify the sender immediately.

1 1

Cannot build boost with Intel Compiler 9.1 on Windows
by Iliyan Georgiev 24 May '07

24 May '07

Hi, Today was a painful and frustrating day. I've been trying to build Boost 1.34 with Intel C++ Compiler 9.1 for hours L I have tried on 4 machines - all of them running either XP or Vista, with installed Visual Studio 2005 and Intel Compiler 9.1. I get the same errors. Here's what I do: 0. Run iclvars.bat to set the environment variables 1. Download the latest boost distribution sources 2. Extract them 3. Download the latest version of bjam (ot build it myself with VC8) 4. Put bjam.exe to the root of boost 5. Run bjam from the boost root (tried with several command lines, i.e. bjam -build-dir=build -toolset=intel-win32 stage) The process starts (warns me about not having Python, which is ok), but the first time compile-c-c++ is run, the windows message "I don't know how to open this file" appears for the file extension .rsp. I dismiss the window for every file (quite annoying) but then the build script cannot find some files and it all messes up. What happens here? What's wrong? In the list archives I found a guy reporting similar troubles with an older version, and the reply suggested to set some variables and to follow a link with more info, which was dead. Anyway, this was for the old version 1. How to deal with this now? I really feel helpless L

2 2

installing boost on mingW (Dev-Cpp) compiler in nice fashion.
by chun ping wang 24 May '07

24 May '07

i know in boost 1.33.1 it will install boost for you into a specific directory with a the lib located in one nice place.. however in 1.34.0 i get.. bin.v2/libs/${boost_name}/${compiler_name}/release/static_link/${lib_file} and i want to be able to get that instant "install" Thanks for any help.

2 1

[Win32] Do I need both versioned and unversioned boost libs?
by Iliyan Georgiev 24 May '07

24 May '07

I noticed that the build process generates duplicate .lib files - versioned and unversioned. Is there a redundancy and do I need both in order auto-linking to work properly? If so, which type is redundant? Thanks in advance!

3 2

[filesystem] is_directory on root throws
by David Neubelt 24 May '07

24 May '07

I'm iterating through a directory on windows (c:\) and it is throwing an exception saying the file isn't found when I call is_directory on it. Here is my code filesystem::path boostdir("c:\\"); // -- make sure the path exists if(!filesystem::exists(boostdir)) { // .. error handle } // -- make sure its a directory if(!filesystem::is_directory(boostdir)) { // .. error handle } filesystem::directory_iterator enditr; for(filesystem::directory_iterator itr(boostdir); itr != enditr; ++itr ) { if(filesystem::is_directory(*itr)) // .. code } filesystem::is_directory(*itr) throws with m_what = "boost::filesystem::is_directory: "c:8f876d52712ede0273c42e26d4af6847": The system cannot find the file specified." I checked and verified the folder does exist, I'm guessing the problem is that the directory is in the format c:folder instead of c:\folder. Thanks, David

1 0

Re: [Boost-users] Cannot build boost with Intel Compiler 9.1 on Windows
by Iliyan Georgiev 23 May '07

23 May '07

Oh, wow, there's a nasty bug. Bjam was looking for icc.exe all the time, and the new name of the Intel C++ Compiler 9.1 executable is icl.exe. Renamed icl.exe to icc.exe and it all worked out! Please, fix this! Hi, Today was a painful and frustrating day. I've been trying to build Boost 1.34 with Intel C++ Compiler 9.1 for hours L I have tried on 4 machines - all of them running either XP or Vista, with installed Visual Studio 2005 and Intel Compiler 9.1. I get the same errors. Here's what I do: 0. Run iclvars.bat to set the environment variables 1. Download the latest boost distribution sources 2. Extract them 3. Download the latest version of bjam (ot build it myself with VC8) 4. Put bjam.exe to the root of boost 5. Run bjam from the boost root (tried with several command lines, i.e. bjam -build-dir=build -toolset=intel-win32 stage) The process starts (warns me about not having Python, which is ok), but the first time compile-c-c++ is run, the windows message "I don't know how to open this file" appears for the file extension .rsp. I dismiss the window for every file (quite annoying) but then the build script cannot find some files and it all messes up. What happens here? What's wrong? In the list archives I found a guy reporting similar troubles with an older version, and the reply suggested to set some variables and to follow a link with more info, which was dead. Anyway, this was for the old version 1. How to deal with this now? I really feel helpless L

3 2

[Boost-Users] [Boost 1.34.0] no support for mipspro compiler?
by Sébastien Fortier 23 May '07

23 May '07

It seems like there is no more support the SGI mipspro compiler? When I tried to build the boost libraries I got the following message mipspro.jam: No such file or directory .../boost_1_34_0/tools/build/v2/build/toolset.jam:27: in toolset.using rule mipspro.init unknown in module toolset. .../boost_1_34_0/tools/build/v2/build/project.jam:825: in using user-config.jam:5: in modules.load .../boost_1_34_0/tools/build/v2/build-system.jam:138: in load .../boost_1_34_0/tools/build/v2/kernel/modules.jam:261: in import .../boost_1_34_0/tools/build/v2/kernel/bootstrap.jam:132: in boost-build .../boost_1_34_0/boost-build.jam:9: in module scope Not all Boost libraries built properly. -- Sébastien Fortier

2 1

BGL - retrieving a subgraph
by Noah Roberts 23 May '07

23 May '07

To describe what I want to do I will use the following graph: digraph G { 0[label="test1"]; 1[label="test2"]; 2[label="test3"]; 3[label="test4"]; 2->0 ; 2->1 ; 3->0 ; 3->2 ; } Now, what I want to do is two things. Both are for vertices such as 2. 2 depends on 0 and 1, it has 3 as its only client. Is there a graph algorithm that will extract those two subgraphs for me? Take any vertex in a directer graph and give me a subgraph of all its descendants or a subgraph for all its ancestors (I imagine those are two different algorithms). I looked at the connected components algorithms and none seemed to apply. Still looking at others but there are a lot of functions. Any ideas to narrow down the search?

2 5

Creating a test file with mutilple test suites
by Stephen Torri 23 May '07

23 May '07

I would link to have one file that has multiple test suites. Right now I have one test suite for all the various tests. I tried to find an example of what I want to do on the boost.org site. There was nothing close to what I wanted. If such a page exists its not easy to find. The close I came to was to create a test_suite for each group and add the test. That is as far as I have gotten. I did find an exampe that use BOOST_AUTO_TEST_CASE but there was only one test suite in it. Here is what I have so far. Is there plans to have a more comprehensive documentation on how to use the boost test library? Can users provide more examples? Stephen BOOST_AUTO_TEST_CASE(test_result_reports) { // test_suite* // init_unit_test_suite ( int, char** ) // { test_suite* test1 = BOOST_TEST_SUITE ("Memory Map constructors"); // Allow Memory_Map to allocate our memory at any location test1->add ( BOOST_TEST_CASE ( &test_default_constructor ) ); test_suite* test2 = BOOST_TEST_SUITE ("Memory Map seek (pass)"); test2->add ( BOOST_TEST_CASE ( &test_seek_PASS_LOW_SEEK ) ); test2->add ( BOOST_TEST_CASE ( &test_seek_PASS_HIGH_SEEK ) ); test_suite* test3 = BOOST_TEST_SUITE ("Memory Map seek (fail)"); test3->add ( BOOST_TEST_CASE ( &test_seek_FAIL_HIGH_SEEK ) ); test3->add ( BOOST_TEST_CASE ( &test_read_FAIL_EMPTY_READ ) ); test_suite* test4 = BOOST_TEST_SUITE ("Memory Map copy"); test4->add ( BOOST_TEST_CASE ( &test_copy ) ); test_suite* test5 = BOOST_TEST_SUITE ("Memory Map read tests"); test5->add ( BOOST_TEST_CASE ( &test_uint8_read ) ); test5->add ( BOOST_TEST_CASE ( &test_uint16_read ) ); test5->add ( BOOST_TEST_CASE ( &test_uint32_read ) ); test5->add ( BOOST_TEST_CASE ( &test_uint64_read ) ); test_suite* test6 = BOOST_TEST_SUITE ("Memory Map address tests"); test6->add ( BOOST_TEST_CASE ( &test_seek_ADDRESS_PASS_LOW ) ); test6->add ( BOOST_TEST_CASE ( &test_seek_ADDRESS_PASS_HIGH ) ); test6->add ( BOOST_TEST_CASE ( &test_seek_ADDRESS_FAIL_LOW ) ); test6->add ( BOOST_TEST_CASE ( &test_seek_ADDRESS_FAIL_HIGH ) ); test_suite* test7 = BOOST_TEST_SUITE ("Memory Map empty name"); test7->add ( BOOST_TEST_CASE ( &test_empty_name ) ); test_suite* test8 = BOOST_TEST_SUITE ("Memory Map subset"); test8->add ( BOOST_TEST_CASE ( &test_subset ) ); //return test; }

3 4

[Boost.Thread] How to Stop/Exit/Interrupt
by JD 23 May '07

23 May '07

Hi, I was looking for a way for a thread to kill another one. William Kempf wrote some time ago answering a question about this: WK > Today you should use a flag and a condition/mutex combination. WK > Direct support for cancellation is being provided in a future release. I don't think it has been implemented AFAIK. Is it still planned for future release? JD

2 1

Anyone know or heard when the Boost Installer will be available?
by Allen 23 May '07

23 May '07

Hi everyone, was wondering if anyone knew when the boost installer will be available. Does anyone know? -- Sincerely, Allen Gene Allen Saucier, Jr Software Engineer EDO-CAS, Inc 100 Quality Circle Huntsville, AL 35806 (256) 922-6453 (w) "...As for I and my house, we shall follow the Lord" Joshua 24:15

4 3

[Python] Exposing foo(std::vector<double>)
by Sohail Somani 23 May '07

23 May '07

Hi, Given: foo(const std::vector<double> &) Is there anything special I need to do to expose foo above? I have: ..MODULE(test) { def("foo",foo); } I've also tried: ...MODULE(test) { class_<std::vector<double> >("vector_d") .def(vector_indexing_suite<...>) ; def("foo",foo); } In both cases, I get Boost.Python throwing an error saying that the signature doesn't match when being called from Python. If there is some documentation I have overlooked, I would appreciate a pointer. Sohail

1 1

Thank you for the response on when the Boost Installer will be available
by Allen 23 May '07

23 May '07

Just wanted to say thank you Dave Abrahams for answering my question on when the installer for Boost will be available. I really look forward to it. However, I was also able to get the current version of boost up & functional for my system. Great Job Developers!!!! & Documentation Up-Keepers. I really appreciate you efforts & can't live without Boost!!!!!! -- Sincerely, Allen Gene Allen Saucier, Jr Software Engineer EDO-CAS, Inc 100 Quality Circle Huntsville, AL 35806 (256) 922-6453 (w) "...As for I and my house, we shall follow the Lord" Joshua 24:15

1 0

message based TCP protocol?
by bringiton bringiton 23 May '07

23 May '07

Hi guys, I'm wondering whether you know a solution to the following problem: I wish to have a tcp connection that sends binary messages of dynamic length. I plan on using asio, but I am concerned with the following: say i send 5 messages with lengths (10, 5, 30, 2, 3). As far as I understand TCP, when I read the stream on the other end, I will not receive the data 1 message at a time. Instead I will receive a stream that may hold N messages and may even hold portions of a message. eg the stream may hold 13 bytes, which is the first message and 3 bites of the 2nd. Are there any solutions to this? ie building messages as they come in through the TCP stream. I know how to implement this myself however I assume that this is probably a common problem. Thanks.

6 7

How do you catch a ABORT signal?
by Stephen Torri 23 May '07

23 May '07

I have code that has error checking to see if pointers are null at run-time. If a pointer is null I am calling abort(). How can I check it with the boost test library? Stephen

2 1

bind + thread
by Nicolas Emiliani 23 May '07

23 May '07

Hi, I was programming a stupid server using boost::thread and boost::bind... and sadly I encountered a problem... probably due to my inexpertice with the library. The thing is that i'm creting a boost thread object like this : struct bla{ void member(){} } bla b; boost::thread t( boost::bind( &bla::member , &b) ); this works fine but, i need to bind a member pointer that returns an int, so as i understand i need to do this : boost::bind<int>( &bla::member_that_returns_int , &b) but boost::thread won't accept it in the constructor... So... what I'm asking is.... is there a trick to do this that I'm not aware of?... basically what occucrred to me was to encapsulate de method that returns an int using the () operator... but it seems pretty nasty to me. is there a way to do it staright forward with boost::bind? thanks! Nico

2 1

mis-filed emails!
by Gottlob Frege 22 May '07

22 May '07

I just (finally) realized that for the past month or so, gmail has been marking some boost mail as spam. In fact, all of Dave Abrahams emails have been going straight to the spam folder (sorry Dave!), as well as emails from a few others (not sure of the overall pattern). Anyhow, just in case anyone was expecting a reply from me and didn't get one, that would be why. I now have an extra 100 or so emails to catch up on, but I doubt I'll reply to any, so as to not re-open any old threads. Tony

3 3

setting program_options programmatically
by Joshua ChaitinPollak 22 May '07

22 May '07

Hello, I'm gradually porting my application to Boost, and I'm starting to use the program_options library. I'm also using the boost::unit_test library, and for some higher level unit tests its often convenient to programmatically setup some properties specific to the test before it runs. I can't figure out how to programmatically set program_options in the variable_map however. I currently wrap the boost::program_options interface with a simple Config class, so what I'd like to have is something like this: Config::setProperty(std::string &option, std::string &value); Config::setProperty(std::string &option, int value); Config::setProperty(std::string &option, bool value); But I can't even get a generic std::string implementation working. It looks like I have go through creating basic_option and basic_parsed_options objects and calling po::store(parsed_options, vm); except that instead of "std::string& option" above, I need to pass in the option_description& object, which starts to get messy. Is there an elegant way of doing this? Thanks, Josh -- Joshua ChaitinPollak Software Engineer Kiva Systems

1 0

executable statecharts
by NFarnsworth＠web.de 22 May '07

22 May '07

Hello, I have a question to the boost statechart library. I work on a Project which has the ambition to execute statecharts which a user has designed. Is it possible to implement this with the help of the boost statechart library or are there any restrictions concerning the boost statechart library to do this? I not mean to design statecharts but to simulate them. I think I can implement executable statecharts by representing the statechart- model as a graph, wich gives me the information how to build the statechart objects in the correct order and howto attach the transactions. Thank you for your help in advance and best regards, Niklas _______________________________________________________________ SMS schreiben mit WEB.DE FreeMail - einfach, schnell und kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192

2 1

Possible bug in iostreams: multichar_wfilter vs multichar_filter
by Chad Walters 22 May '07

22 May '07

I was looking through the iostreams sources and came across this in boost/iostreams/concepts.hpp: //------Definitions of helper templates for multi-character filter cncepts----// template<typename Mode, typename Ch = char> struct multichar_filter : filter<Mode, Ch> { struct category : filter<Mode, Ch>::category, multichar_tag { }; }; template<typename Mode, typename Ch = wchar_t> struct multichar_wfilter : multichar_filter<Mode, Ch> { }; typedef multichar_filter<input> multichar_input_filter; typedef multichar_filter<input> multichar_input_wfilter; typedef multichar_filter<output> multichar_output_filter; typedef multichar_filter<output> multichar_output_wfilter; typedef multichar_filter<dual_use> multichar_dual_use_filter; typedef multichar_filter<dual_use> multichar_dual_use_wfilter; //-------------------------------------------------------------------------- --// It seems to me that the wide versions of these typedefs should be defined using multichar_wfilter, making the typedefs appear as follows: typedef multichar_filter<input> multichar_input_filter; typedef multichar_wfilter<input> multichar_input_wfilter; typedef multichar_filter<output> multichar_output_filter; typedef multichar_wfilter<output> multichar_output_wfilter; typedef multichar_filter<dual_use> multichar_dual_use_filter; typedef multichar_wfilter<dual_use> multichar_dual_use_wfilter; Also, the default arguments in multichar_filter and multichar_wfilter termplates aren't wrapped in BOOST_IOSTREAMS_DEFAULT_ARG(), unlike the other ones in this file. Seems like they should be, for consistency if nothing else. Chad

1 0

Xpressive with non char data
by gast128 22 May '07

22 May '07

Dear all, The new Boost version ships with an extra regular expression library. Now I need a dynamic regualr expression on a std::vector<std::string>. The character type would be a string (since in that case multiple characters are allowed on the same position). Is this possible with xpressive? Boost.Regex didn't seem to support it according to a previous post. wkr, me

2 2

building specific versions of boost
by Filip Konvička 22 May '07

22 May '07

Hi, I wanted to build just the <threading>multi <link>static <runtime-link>dynamic version of Boost 1.34 with MSVC 8.0. However, I did not find any info on how to do that, do I ended up deleting the extra targets from Jamfile.v2's "default-build". Sure there is a better way of doing this (some command-line flags)? Thanks, Filip

2 3

[1.34.0] x86_64 rpms available with Boost.Asio 0.3.8rc2 and Boost.GIL 2.0 included.
by Chris Fairles 22 May '07

22 May '07

I've seen the odd request for asio to be included with boost. I hacked together a boost rpm package that includes asio and just for fun I stuck Boost.GIL 2.0 in there as well. You can find the rpms here: http://blog.tripthelight.net/?p=4 They were built on a CentOS 5 platform (rhel5 compatible) with gcc 4.1.1 The asio docs are missing. I'm not a BBv2 expert and couldnt really figure out how the docs are built and how to add asio's (attempted several things by mimicing how other docs are built to no avail). Gil didn't include docs, just a html file pointing to the homepage. Of course, I don't take responsibility if these mess up your system. They work for me "as is" but your milage may vary. Do let me know if theres serious issues however and I'll see what I can do. Chris

2 2

BOOST 1.33 msvc/gcc
by Nadine Faizant 22 May '07

22 May '07

Hello, I'm testing same code compiling under MSVC 8.0 or gcc 3.4.2 on the same machine. I face two different problems around boost::dynamic_bitset : - with msvc version : dynamic_bitset allocation is very, very slow (compare to gcc execution time), - with gcc version, it seems (i'm quiete sure) that dynamic_bitset allocation take more and more time after several hours of intensive use. The two problem sound like a problem around dynamic_bitset allocation. Somebody did notice that ? Thanks for your help, Nadine

1 0

problem building asio on Windows
by Yunior Peralta González 22 May '07

22 May '07

I tried to build boost in Windows using bjam and I made it without any error (I just compiled the thread library). the problem came when I included asio.hpp I think that's because not all thread libraries were installed (just 6 files were installed) when I did the same in Debian I got just 4 files and also had errors with asio, but when I installed this way: ./configure --with-libraries=thread sudo make install instead of bjam, I got 10 files and asio worked perfectly. I'm not sure but if I need the 10 files as I guess I think that I'm not be able of using asio on Windows, but this is a contradiction because asio like boost is a multi-platform set of libraries. I'm very confused and I'll be very grateful if you try to help me.

2 1

[Boost.asio] daytime server example doesn't work.
by Schrotti 22 May '07

22 May '07

Hello, I have a problem with compiling the daytime server example of the Boost.asio tutorial (http://asio.sourceforge.net/asio-0.3.7/doc/tutorial/tutdaytime2src.html) I always get an exception which says "A File which already exists can't be created". Does anyone know the reason of the problem? btw: I'm using Boost 1.33.1 under Windows XP Sp2 with the MS Visual Studio IDE (express) and its compiler. Thanks for your efforts, Schrotti.

2 1

Using asio SSL
by Daniel Adolfsson 22 May '07

22 May '07

Hello. I'm trying to write a small FTP client using asio, and I'm having trouble getting SSL/TLS working. What I want to do is to connect to the FTP server normally, and send "AUTH TLS" in cleartext to the server, and if successfull switch over to SSL. I have tried to use a ssl::stream to connect to the server, and while not in SSL mode, using async_write() / async_read_until() with stream_.next_layer(). This works fine, and everything is sent, but when the client enter SSL mode, the server only receives junk from the client, while the client receives good responses from the server. I've checked and double checked- I am using stream_ while in SSL mode, not stream_.next_layer(). So, async_read_until() seems to work with stream_ in SSL mode, but async_write() doesn't. Any ideas? Is there a better way to do this? Daniel

2 1