- Boost-users - lists.boost.org

[boost.function] member function with default values for parameters
by Yan Drugalya 17 Apr '08

17 Apr '08

Folks, What is wrong with the code below: class SomeClass { public: int doSomething(int i=0) { return ++i; } }; boost::function2<int, SomeClass*, int> func = &SomeClass::doSomething; SomeClass sc; std::cout << func(&sc, 1); Currently I'm having following error: An object or reference of type "boost::function2<int,SomeClass *,int,std::allocat or<boost::function_base> >" cannot be initialized with an rvalue of type "int (SomeClass::*)(int)". If I remove default value for parameter i - everything compiles. Complier: VisualAge 7.0 Thanks

2 1

Xcode 3 compile issue
by Dave DeLong 17 Apr '08

17 Apr '08

Hi everyone, I'm attempting to build a project that links against the Boost library, but I keep running into this error: /Developer/SDKs/MacOSX10.5.sdk/usr/include/c++/4.0.0/i686-apple- darwin9/bits/c++locale.h:72: error: 'snprintf' is not a member of 'std' This has come up before on this list ( http://lists.boost.org/boost-users/2008/02/34157.php ), but it doesn't seem like there was an answer. How can I get rid of this? My project requires boost.... Specs: Mac OS X 10.5, Xcode 3.0, CoreDuo chipset Thanks! Dave DeLong

2 1

Mac Xcode 3.0 Boost compile issue
by Daniel Lord 17 Apr '08

17 Apr '08

This is a specific error with Xcode 3.0 on Leopard not a generic Boost question: Something is screwy with the paths between Boost and Xcode: Xcode 3.0 needs to find the Boost includes, so I added my directory containing the top of the Boost tree (the boost_1_34_1 directory) to the User header Search paths in the build configuration for the target. No problem, everything appears to be found. However, adding that search path screws up something else: An error pops up in c++locale.h highlighting the line: const int __ret = std::snprintf(__out, __size, __fmt, __prec, __v); and claiming: "error: 'snprintf' is not a member of 'std'." What the freak? It compiles fine from the command line with g++ and using -I<same path to boost as above>. So it is some wierd interaction with Boost and Xcode paths but it make no sense to me. It's almost as if Boost is redefining something. But what? This does not occur if I don't add the Boost search paths but of course then the Boost definitions are not found. The program is a simple one from O'Reilly's C+++ Cookbook show below: #include <boost/filesystem/operations.hpp> #include <boost/filesystem/fstream.hpp> #include <iostream> using namespace boost::filesystem; int main(int argc, char** argv) { if (argc < 2) { std::cerr << "Usage: " << argv[0] << " [dir name]\n"; return(EXIT_FAILURE); } path fullPath = // Create the full, absolute path name system_complete(path(argv[1], native)); if (!exists(fullPath)) { std::cerr << "Error: the directory " << fullPath.string( ) << " does not exist.\n"; return(EXIT_FAILURE); } if (!is_directory(fullPath)) { std::cout << fullPath.string( ) << " is not a directory!\n"; return(EXIT_SUCCESS); } directory_iterator end; for (directory_iterator it(fullPath); it != end; ++it) { // Iterate through each // element in the dir, std::cout << it->leaf( ); // almost as you would if (is_directory(*it)) // an STL container std::cout << " (dir)"; std::cout << '\n'; } return(EXIT_SUCCESS); }

2 1

Strange use of boost::scoped_ptr in Boost.Process
by Boris 17 Apr '08

17 Apr '08

I found this in the source code of Boost.Process (which is not an official Boost library): // XXX I'm not sure this usage of scoped_ptr is correct... boost::scoped_ptr< STARTUPINFO > si((STARTUPINFO*)new char[setup.m_startupinfo->cb]); I'm not sure either if this is correct. I don't know either why a char array is allocated and not STARTUPINFO directly. Does anyone know though if the code above is correct? STARTUPINFO is a POD (see http://msdn2.microsoft.com/en-us/library/ms686331(VS.85).aspx) Boris

1 0

[ublas] traits.hpp(97) vc8 error in DEBUG mode
by Markus Werle 17 Apr '08

17 Apr '08

Hi! This error occurs in DEBUG mode only. No testcase yet, but you probably have an idea anyway? regards, Markus [..snipped ...]\boost_1_35_0\boost\numeric\ublas\traits.hpp(97) : error C2784: 'boost::numeric::interval<T,Policies> boost::numeric::sqrt(const boost::numeric::interval<T,Policies> &)' : could not deduce template argument for 'const boost::numeric::interval<T,Policies> &' from 'const double' [..snipped...]\boost_1_35_0\boost\numeric\interval\arith2.hpp(198): see declaration of 'boost::numeric::sqrt' [..snipped...]\boost_1_35_0\boost\numeric\ublas\traits.hpp(93) : while compiling class template member function 'double boost::numeric::ublas::scalar_traits<T>::type_sqrt(const double &)' with [ T=double ] [..snipped...]\boost_1_35_0\boost\numeric\ublas\traits.hpp(151) : see reference to class template instantiation 'boost::numeric::ublas::scalar_traits<T>' being compiled with [ T=double ] [..snipped ...]\boost_1_35_0\boost\numeric\ublas\traits.hpp(97) : fatal error C1903: unable to recover from previous error(s); stopping compilation

1 0

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