- Boost-users - lists.boost.org

incremental_components, component_index for listS, listS adjacency_list
by Sean Kelly 02 May '06

02 May '06

Hi All, I couldn't figure out how to use the component_index facility of the incremental components algorithm when operating on an listS, listS adjecency_list so I rolled my own as per below. Any suggestion on how to use the provided interface are appreciated. thanks Sean using namespace boost; typedef graph_traits<BlockDependencyGraph>::vertices_size_type size_type; typedef BlockDependencyGraph::vertex_descriptor vertex_descriptor; typedef BlockDependencyGraph::vertex_iterator vertex_iterator; BlockDependencyGraph& graph = _blk_dep_view->get_graph(); // Disjoint sets requires Rank to map vertex_descriptors onto integers typedef std::map<vertex_descriptor, int> rank_storage_t; rank_storage_t rank_map; typedef associative_property_map<rank_storage_t> rank_t; // Disjoint sets requires Parent to map vertex_descriptors onto vertex_descriptors typedef std::map<vertex_descriptor, vertex_descriptor> parent_storage_t; parent_storage_t parent_map; typedef associative_property_map<parent_storage_t> parent_t; disjoint_sets<rank_t, parent_t> component_set(make_assoc_property_map(rank_map), make_assoc_property_map(parent_map)); initialize_incremental_components(graph, component_set); incremental_components(graph, component_set); // Not at all clear how to extract components using component_index interface // when graph is listS, listS adjacency_list... so we roll our own. // // Index vertices by representative component in a multimap typedef std::map<vertex_descriptor, std::vector<vertex_descriptor> > components_t; components_t components; vertex_iterator curr(0), end(0); for(tie(curr, end) = vertices(graph); curr != end; ++curr) { components[component_set.find_set(*curr)].push_back(*curr); } // Print all components size_t n_c(0); for(components_t::const_iterator c = components.begin(); c != components.end(); ++c, ++n_c) { std::cout << "component " << n_c << std::endl; for(std::vector<vertex_descriptor>::const_iterator element = c->second.begin(); element != c->second.end(); ++element) { std::cout << graph[*element]->get_name() << std::endl; } }

2 1

Fusion. transform_view example
by John Christopher 02 May '06

02 May '06

Hello, The example at the bottom of: http://spirit.sourceforge.net/dl_more/fusion_v2/libs/fusion/doc/html/fusion… does not compile; I am using MinGw 3.4.2. #include <iostream> #include <boost/fusion/sequence.hpp> #include <boost/fusion/algorithm.hpp> using namespace boost::fusion; struct square { template <typename T> struct result { typedef int type; }; template <typename T> int operator()(T x) const { return x * x; } }; typedef vector<int, short, double> vector_type; int main (int argc, char ** argv) { vector_type vec(2, 5, 3.3); transform_view<vector_type, square> transform(vec, square()); std::cout << transform << std::endl; // return 0; } More specifically it is the line std::cout << transform << std::endl; that seems to create the problem. I was also wondering why operator() of square return int and not T. Thanks

2 3

Re: [Boost-users] Timer
by Andy Little 02 May '06

02 May '06

"Vivek Bhat" <bhat.vivek(a)gmail.com> wrote > Hello, I'm a Mechanical Engineer (see: programming-knowledge defficient) and > I need some code to output an elapsed time, accurate to within at least > hundredths of a second. Hi Vivek. Actually I think you are on the wrong boost list. The one you want for this sort of query is the boost user list gmane.comp.lib.boost.user. I have cross posted this item there and you should send any followups there only too. > It appears to me as though the 'timer' > library<http://www.boost.org/libs/timer/timer.htm>is ideal for this > application, but when I try to run the simple code given > in the documentation, I get errors regarding missing .hpp files. I'm > assuming that I have made errors in downloading and installing the boost > libraries. I tried to read the 'getting started' section, but I'm having > difficulty understanding. Are there any more simple 'getting started' > instructions available for programming dunces running Windows? As far as I know Boost.Timer is just a wrapper for the C standard timer functions in <ctime>. Anyway it would be helpful to provide more details of the compiler you are using and the actual error messages you are getting too. Hope this helps regards Andy Little

1 0

[program_options] Abort trap by store
by Pascal Vizeli 02 May '06

02 May '06

Hi, I habe a problem with store(parse_config_file(...), ...); I don't know why... That isn't my first program with program_options but it is my first program with parse_command_line and parse_config_file in one. Here is my code. First I parse the command line, second I parse the config file. Is it a very simple script. With my debuger (gdb and ddd) I can see, that the problem is here: bpo::store(bpo::parse_config_file(conf_file, desc_file), vm); (at the "end" from script). vm is global definiert. The Namspace bpo ist boost::program_options I use the gcc 3.3.5 and I compile static with: /usr/local/lib/libboost_program_options-gcc-1_33_1.a My OS is OpenBSD What I make false? My script is at the end. Sorry for my bad english. I thanks for Help, Dear, Pascal Vizeli [cpp] bpo::variables_map vm_cmd; bpo::options_description desc_file("config file"), desc_cmd("command line"); // Make options descriptions desc_cmd.add_options() ("config,c", bpo::value<std::string>(), "The Config file") ("help,h", "produce help message") ; desc_file.add_options() ("port", bpo::value<int>()->default_value(6667), "IRC Server Port") ("server", bpo::value<char *>(), "IRC Server IP") ("nick", bpo::value<std::string>(), "Bot Nick") ("user", bpo::value<std::string>(), "username hostname servername :realname") ("nickserv", bpo::value<std::string>(), "Nickserv Identify string") ("oper-pw", bpo::value<std::string>(), "Oper Password") ("mode", bpo::value<std::string>(), "Mode for Bot") ("logchan", bpo::value<std::string>(), "A #Chan for logs") ("logchan_status", bpo::value<int>(), "Logchan On or Off") ("invite-chan", bpo::value<int>(), "Invite for logchan?") ("invite-command", bpo::value<std::string>(), "Invite Command") ("re-chan", bpo::value<int>(), "Refresh Channellist until SEC") ("logfile", bpo::value<std::string>(), "File for logs") ("chroot", bpo::value<char *>(), "Set the chroot dir (path)") ; // Parse CMD bpo::store(bpo::parse_command_line(argc, argv, desc_cmd), vm_cmd); bpo::notify(vm_cmd); if(vm_cmd.count("help")){ std::cout << desc_cmd << std::endl; std::cout << desc_file << std::endl; exit(0); } if(!vm_cmd.count("config")){ printo_err("Config file not found! Use --help for Infos"); exit(0); } // Open Config file std::ifstream conf_file(vm_cmd["config"].as<std::string>().c_str()); if(conf_file.is_open()){ // Parse Config file bpo::store(bpo::parse_config_file(conf_file, desc_file), vm); // Here is the false say my debuger bpo::notify(vm); }else{ printo_err("can't open the conf file!"); exit(0); } [/cpp]

2 1

[BGL] parallel edges
by John Krajewski 02 May '06

02 May '06

How do I make a graph that will just ignore when I attempt to add a parallel edge? Is there a way to set this on the graph, or is it only determined by the type of storage structures I'm using with the graph? I want to use a vecS to store the edges, but I dont want to allow parallel edges. -John

2 1

Re: [Boost-users] adjacency_list directed to undirected conversion
by Sean Kelly 01 May '06

01 May '06

Doug, thanks -- that worked nicely. cheers Sean -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Douglas Gregor Sent: Wednesday, April 26, 2006 6:43 AM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] adjacency_list directed to undirected conversion On Apr 25, 2006, at 12:16 PM, Sean Kelly wrote: > I have a directed adjacency_list graph for which I would like to > identify connected components (ignoring the directed-ness). After > poking around the BGL docs it seems that my best option is to role my > own algo To do this, you could use the incremental connected components algorithms in the BGL. Those work equally well on directed and undirected graphs. Doug _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

[BGL] Inverting the map from strong_connected
by John Krajewski 01 May '06

01 May '06

Hi all, The function "strong_components" produces a mapping from vertex to component - but what I need is a mapping the other way, from component to vertex! Now of course I could go through the property_map, query each vertex, and create the new map myself, but theres got to be a better way. The map is just a writable_property_map concept, so theoretically I should be able to create my own somehow that will swap the key and value, but I'm not sure how I would do that. Any suggestions? I also want to create a graph of the strongly connected components, but I suppose theres no tricky way to do this - just go through each vertex in the returned map and add its connections to the strong component's connections (though there's probably something tricky I could do with iterators to make this automatic) Thanks for any help, my mind is still adjusting to being warped by this generic programming stuff :)

1 0

[serialization] My archive class + BOOST_CLASS_EXPORT
by Denis Kuznetsov 01 May '06

01 May '06

I have my own archive class (derived from binary_oarchive_impl). And I have a base abstract class and derived classes(non-abstract). I use BOOST_CLASS_EXPORT macro to register my derived classes. But it seems the macro works only for Boost archives, not for user defined archives. Is there any way to register derived classes to get the same functionality that BOOST_CLASS_EXPORT provides?

3 3

filesystem/signals/serialization dlls not linking
by Dave Compton 30 Apr '06

30 Apr '06

When I try to build boost 1_33_1 with mingw ( gcc 3.4.2 ) some libraries don't link properly. At the end of the build process, the following files are missing : boost_filesystem-mgw-1_33_1.dll boost_filesystem-mgw-1_33_1.lib boost_filesystem-mgw-d-1_33_1.dll boost_filesystem-mgw-d-1_33_1.lib boost_filesystem-mgw-d.lib boost_filesystem-mgw-mt-1_33_1.dll boost_filesystem-mgw-mt-1_33_1.lib boost_filesystem-mgw-mt-d-1_33_1.dll boost_filesystem-mgw-mt-d-1_33_1.lib boost_filesystem-mgw-mt-d.lib boost_filesystem-mgw-mt.lib boost_filesystem-mgw.lib boost_serialization-mgw-1_33_1.dll boost_serialization-mgw-1_33_1.lib boost_serialization-mgw-d-1_33_1.dll boost_serialization-mgw-d-1_33_1.lib boost_serialization-mgw-d.lib boost_serialization-mgw-mt-1_33_1.dll boost_serialization-mgw-mt-1_33_1.lib boost_serialization-mgw-mt-d-1_33_1.dll boost_serialization-mgw-mt-d-1_33_1.lib boost_serialization-mgw-mt-d.lib boost_serialization-mgw-mt.lib boost_serialization-mgw.lib boost_signals-mgw-1_33_1.dll boost_signals-mgw-1_33_1.lib boost_signals-mgw-d-1_33_1.dll boost_signals-mgw-d-1_33_1.lib boost_signals-mgw-d.lib boost_signals-mgw-mt-1_33_1.dll boost_signals-mgw-mt-1_33_1.lib boost_signals-mgw-mt-d-1_33_1.dll boost_signals-mgw-mt-d-1_33_1.lib boost_signals-mgw-mt-d.lib boost_signals-mgw-mt.lib boost_signals-mgw.lib I believe that all the other libraries are building successfully. Here are the first few errors that appear in my build log: gcc-Link-action bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\boost_filesystem-mgw-d-1_33_1.dll bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\boost_filesystem-mgw-d-1_33_1.lib Creating library file: bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\boost_filesystem-mgw-d-1_33_1.lib bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\operations_posix_windows.obj(.text+0x7ac): In function `ZN5boost10filesystem6detail12dir_itr_initERNS_10shared_ptrINS1_11dir_itr_impEEERKNS0_4pathE': c:/cygwin_home/dave/devl/thirdParty/Cygwin/boost/boost_1_33_1x/src/boost_1_33_1/libs/filesystem/build/../src/operations_posix_windows.cpp:308: undefined reference to `boost::filesystem::detail::system_error_code()' bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\operations_posix_windows.obj(.text+0x81d):c:/cygwin_home/dave/devl/thirdParty/Cygwin/boost/boost_1_33_1x/src/boost_1_33_1/libs/filesystem/build/../src/operations_posix_windows.cpp:308: undefined reference to `boost::filesystem::filesystem_error::filesystem_error(std::string const&, boost::filesystem::path const&, int)' bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\operations_posix_windows.obj(.text+0x8d7):c:/cygwin_home/dave/devl/thirdParty/Cygwin/boost/boost_1_33_1x/src/boost_1_33_1/libs/filesystem/build/../src/operations_posix_windows.cpp:308: undefined reference to `boost::filesystem::filesystem_error::~filesystem_error()' bin\boost\libs\filesystem\build\boost_filesystem.dll\mingw\debug\operations_posix_windows.obj(.text+0x8f3):c:/cygwin_home/dave/devl/thirdParty/Cygwin/boost/boost_1_33_1x/src/boost_1_33_1/libs/filesystem/build/../src/operations_posix_windows.cpp:308: undefined reference to `boost::filesystem::filesystem_error::~filesystem_error()' So far my work around has been to use 1_33_0 which builds 100% cleanly. Does anyone have any idea either what I might be doing wrong or what is going wrong otherwise. Thanks. - Dave

1 2

[Serialization] Difficulties with export
by Joseph Hosteny 30 Apr '06

30 Apr '06

Thanks for the response. I'd have continued the thread, but I didn't get the original emails from the list to reply to. I tried an explicit registration, with the same results. Also, I did step into the code where the EXPORT macros were being called, and I verified that the pertinent classes were registered in the boost::serialization::detail::ktmap, at least (not sure about the tkmap). It seems like it just doesn't find the class in the oserialization file where the exception is thrown. The serialization is invoked in a const public class method, via "oa << this;" Is there anything tricky you might be aware of that could cause the type comparisons to fail? Could I potentially be missing something in my class? Unfortunately, the project is quite large and it'd fairly difficult to strip out everything and add it back in incrementally.

2 4