April 2008 - Boost-users - lists.preview.boost.org

Retrieving edges list
by Chartier Nicolas 17 Apr '08

17 Apr '08

I'm rather new with the use of the BGL, but I'd like to know how I could retrieve a weighted edges list from a graph, without forgetting vertices and weight, in a list for example, i.e. for n edges: Element 1 : egde 1; node 1’; node 1’’; weight 1 Element 2 : egde 2; node 2’; node 2’’; weight 2 ... Element n : egde n; node n’; node n’’; weight n Also, I’d like to know if I have different lists from different graphs, is there a mean to merge them into only one big list and to remove the unnecessary elements? Any suggestion welcome ;) NC

1 0

[Regex]Trouble building library for WM5 using Visual C++ 8 - boost release 1.35
by Brady Reed 17 Apr '08

17 Apr '08

Let me say up front that my goal is to build the regex lib with Visual C++ 8.0 using nmake and the supplied makefiles with as little modification to the boost sources as possible. In fact, I would like to limit customization for Windows Mobile 5 entirely to adding #define's to boost/config/user.hpp. Is this possible? It would certainly minimize the amount of work when upgrading boost. In order to minimize the number of files to keep under source control, I've moved only the regex-relevant files from the full zip extraction to the source control directories. I tested this by copying files from the 1.35 boost distribution until I had a clean XP desktop VC8 build. However, the output below shows where the WM5 build fails right out of the gate. There seems to be an stdlib issue with a missing 'strxfrm'. In order to build for WM5, I've taken the following steps so far: (1) Added to boost\config\user.hpp: #define BOOST_NO_STD_LOCALE #define BOOST_USE_WINDOWS_H //Undefine the WIN32 that was defined in the makefile command line #undef WIN32 // Define the appropriate versions of WinCE #define _WIN32_WCE 500 #define UNDER_CE 500 // Define processor type #define _ARM_ // Define compiler version #define _VC80_UPGRADE 0x0710 (2) After navigating to <boost>/boost/libs/regex/build, set the INCLUDE environment variable to (broken into lines here to help readability): C:\Program Files\Microsoft Visual Studio 8\VC\ce\include; C:\Program Files\Windows CE Tools\wce500\Windows Mobile 5.0 Pocket PC SDK\include; C:\Program Files\Windows CE Tools\wce500\Windows Mobile 5.0 Pocket PC SDK\include\ARMV4I; then invoke nmake -fvc8.mak all (3) The resulting output on the command line is: cl /nologo /MT /W3 /EHsc /O2 /GF /Gy /I..\..\..\ /D_MT /DWIN32 /DNDEBU G /D_MBCS /D_LIB /FD /Zc:wchar_t /c -Y- -Fo./vc80/libboost_regex-vc80-mt-s-1_3 5/ -Fdvc80/libboost_regex-vc80-mt-s-1_35.pdb ../src/c_regex_traits.cpp c_regex_traits.cpp ../src/c_regex_traits.cpp(54) : error C2039: 'strxfrm' : is not a member of 'std' ../src/c_regex_traits.cpp(54) : error C3861: 'strxfrm': identifier not found ../src/c_regex_traits.cpp(54) : fatal error C1903: unable to recover from previous error(s); stopping compilation NMAKE : fatal error U1077: '"C:\Program Files\Microsoft Visual Studio 8\VC\bin\cl.EXE"' : return code '0x2' Stop. In http://article.gmane.org/gmane.comp.lib.boost.user/32566/ it was stated that the regex lib should build out of the box for Windows Mobile. What am I doing wrong?

2 1

[breadth_first_search.hpp] -Link error : Segmentation Fault
by bijas kn 17 Apr '08

17 Apr '08

While I am trying to make an application (johnson-test) in Metrowerks CodeWarrior (ARMV5 UDEB) , I am getting the message : *armcc.exe** has encountered a problem and needs to close. *With error* ***Segementation fault *. This problem i am getting for the highlighted code given bolow... template <class IncidenceGraph, class Buffer, class BFSVisitor, class ColorMap> void breadth_first_visit (const IncidenceGraph& g, typename graph_traits<IncidenceGraph>::vertex_descriptor s, Buffer& Q, BFSVisitor vis, ColorMap color) { function_requires< IncidenceGraphConcept<IncidenceGraph> >(); typedef graph_traits<IncidenceGraph> GTraits; typedef typename GTraits::vertex_descriptor Vertex; typedef typename GTraits::edge_descriptor Edge; function_requires< BFSVisitorConcept<BFSVisitor, IncidenceGraph> >(); function_requires< ReadWritePropertyMapConcept<ColorMap, Vertex> >(); typedef typename property_traits<ColorMap>::value_type ColorValue; typedef color_traits<ColorValue> Color; typename GTraits::out_edge_iterator ei, ei_end; put(color, s, Color::gray()); vis.discover_vertex(s, g); Q.push(s); while (! Q.empty()) { Vertex u = Q.top(); Q.pop(); vis.examine_vertex(u, g); for (tie(ei, ei_end) = out_edges(u, g); ei != ei_end; ++ei) { Vertex v = target(*ei, g); vis.examine_edge(*ei, g); ColorValue v_color = get(color, v); if (v_color == Color::white()) { vis.tree_edge(*ei, g); put(color, v, Color::gray()); vis.discover_vertex(v, g); Q.push(v); } else { vis.non_tree_edge(*ei, g); if (v_color == Color::gray()) vis.gray_target(*ei, g); else vis.black_target(*ei, g); } } // end for put(color, u, Color::black()); vis.finish_vertex(u, g); //} // end while } // breadth_first_visit *Q.empty(), Q.pop(), Q.top(), ei!=ei.end* etc are creating problem Thanks & Regards Bijas K N 09886892854

1 0

[statechart] How to keep transition actions separated from the interface?
by Andre Massing 16 Apr '08

16 Apr '08

Hi, I am implementing a fsm for a kind of a project class, which tracks certain changes like adding special files or starting a computation. To keep the interface as small/clean as possible all interactions with the fsm should be soley processed by the process_event method. But seemingl I have to declare transition actions public in an outer state to make them available for inner state reactions/transitions. Is there any nice way to avoid this and to keep transition actions separated from the public interface? (Declaring *all* inner states which need access to be friend seems too ugly to me...) A small code snippet to demonstrate my issue: In Project.h: struct Idle; struct Project : sc::state_machine< Project, Idle > { public: //.... void solve(const EvSolve &); //... }; //... struct Idle : sc::simple_state< Idle, Project, mpl::list<GeoState, SolverState> > { typedef mpl::list< sc::custom_reaction< EvCompile >, sc::custom_reaction< EvSolve > > reactions; sc::result react(const EvCompile &); sc::result react(const EvSolve &); }; In Project.h: sc::result Idle::react(const EvSolve & ev) { //... return transit<Solve>(&Project::solve, ev); } Regards, Andre

3 16

Detecting cycles
by Chartier Nicolas 16 Apr '08

16 Apr '08

I wonder if someone knows an easy way to detect cycles in a graph. Any suggestion welcome ;) NC

4 3

wstring in serialization?
by dennis＠conus.info 16 Apr '08

16 Apr '08

Hello, >> c:/boost/boost/serialization/access.hpp:109: error: 'struct >> std::basic_string<wchar_t, std::char_traits<wchar_t>, >> std::allocator<wchar_t> >' has no member named 'serialize' >> >> So, is there wstring support for serialization, or I should write my >> own? > > Hi Dennis, > sure there is, did you include <boost/serialization/wstring.hpp>? At least in latest boost distribution, this file (wstring.hpp) is not exist.

2 2

Building Boost as Universal Binary on Mac OS X
by Thomas Hauk 16 Apr '08

16 Apr '08

I would like to build 1.35.0 Boost libs (especially date_time, filesystem, regex, system, and thread) on my G5 as Universal Binaries with bjam 3.1.16, so I can use them in a project with Xcode on both my G5 and my Intel-based MacBook Pro. (I was surprised they did not build that way by default). The command I used was: $> bjam --toolset=darwin --build-type=complete --with-thread --with- date_time --with-filesystem --with-regex --with-system I know they're just built PPC because I used the `file` command, e.g.: $> file libboost_system-mt-d-1_35.dylib libboost_system-mt-d-1_35.dylib: Mach-O dynamically linked shared library ppc Can anyone help? TIA, T -- "It's a very sobering feeling to be up in space and realise that one's safety factor was determined by the lowest bidder on a government contract.'' -- Alan Shepherd

2 2

[MPI] compilation error with msvc 64bit due to size_t datatype
by Andreas Buhr 16 Apr '08

16 Apr '08

Hello, I cannot compile boost.mpi with msvc 8.0 64bit. The complete error message is attached. The reason is that is_mpi_datatype<std::size_t> derives from mpl::true_ with the 32 bit compiler while it derives from mpl::false_ with the 64bit compiler. Unfortunately, I do not understand the reason, so I cannot fix it. I would be grad if someone could help. best regards, Andreas Buhr -- Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

3 3

Unhandled exception using directory_iterator
by Silvio Reis Junior 16 Apr '08

16 Apr '08

Hello friends! I'm trying to use directory_iterator but when I try to list recursivelly an entire drive or a big directory I get an unhandled exception. I'm trying to figure out if I did something wrong but the code is very similar to the example on the boost web page. Here is the code: //------------------------------------------------------------------- VOID CFileSearch::InternalSearch( path RootDir ) { directory_iterator EndDirItr; for ( directory_iterator DirItr( RootDir ); DirItr != EndDirItr; ++DirItr ) { if ( is_regular( DirItr->status() ) ) { _tprintf( "FILE: %s\n", DirItr->path().string().c_str() ); } else if ( is_directory( DirItr->status() ) ) { _tprintf( "DIR : %s\n", DirItr->path().string().c_str() ); InternalSearch( DirItr->path() ); } } } //------------------------------------------------------------------- VOID CFileSearch::Search( string RootDir ) { path TempRootDir = system_complete( path( RootDir, native ) ); if ( exists( TempRootDir ) == true ) { InternalSearch( TempRootDir ); } } //------------------------------------------------------------------- CFileSearch FileSearch; FileSearch.Search( "c:\\" ); Is there something wrong in the above code? Thank you in advance. Silvio.

4 8

wstring in serialization?
by dennis＠conus.info 16 Apr '08

16 Apr '08

Hello, I have a simple piece of code: void save_tbl (std::map<std::wstring, std::map<WORD, std::map<WORD, std::map<WORD, int> > > > & tbl) { std::ofstream ofs ("prj12_tbl.dat", std::ios::trunc | std::ios::binary ); boost::archive::binary_oarchive oa(ofs); assert (ofs.good()); oa << tbl; ofs.close(); }; Got this errors: c:/boost/boost/serialization/access.hpp: In static member function `static voidboost::serialization::access::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::binary_oarchive, T = std::basic_string<wchar_t, std::char_tr aits<wchar_t>, std::allocator<wchar_t> >]': c:/boost/boost/serialization/serialization.hpp:81: instantiated from `void boost::serialization::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::binary_oarchive, T = std::basic_string<wchar_t, std::char_traits<wchar_ t>, std::allocator<wchar_t> >]' c:/boost/boost/serialization/serialization.hpp:140: instantiated from `void boost::serialization::serialize_adl(Archive&, T&, unsigned int) [with Archive = boost::archive::binary_oarchive, T = std::basic_string<wchar_t, std::char_traits<w char_t>, std::allocator<wchar_t> >]' c:/boost/boost/archive/detail/oserializer.hpp:151: instantiated from `void boost::archive::detail::oserializer<Archive, T>::save_object_data(boost::archive::detail::basic_oarchive&, const void*) const [with Archive = boost::archive::binar y_oarchive, T = std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> >]' collect.cpp:149: instantiated from here c:/boost/boost/serialization/access.hpp:109: error: 'struct std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> >' has no member named 'serialize' So, is there wstring support for serialization, or I should write my own?

2 1