February 2013 - Boost-users

[parallel graph] error building parallel BGL
by The Maschine 25 Feb '13

25 Feb '13

Hi all, My boost.mpi test seems ok: ####################### ➜ Temp mpic++ -I/usr/local/include/boost/mpi/ mpitest2.cc - L/usr/local/lib/ -lboost_mpi-mt -lboost_serialization-mt ➜ Temp ./a.out I am process 0 of 1. ➜ Temp mpirun -n 6 ./a.out I am process 1 of 6. I am process 4 of 6. I am process 0 of 6. I am process 3 of 6. I am process 5 of 6. I am process 2 of 6. ###################### But when I try to build parallelBGL I get the following error: ###################### ➜ parallelBGL-Mac cmake ../parallelBGL/ -- The C compiler identification is Clang 4.2.0 -- The CXX compiler identification is Clang 4.2.0 -- Check for working C compiler: /usr/bin/cc -- Check for working C compiler: /usr/bin/cc -- works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working CXX compiler: /usr/bin/c++ -- Check for working CXX compiler: /usr/bin/c++ -- works -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Boost version: 1.53.0 -- Found the following Boost libraries: -- mpi -- serialization -- filesystem -- system -- thread -- Configuring done -- Generating done -- Build files have been written to: /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL-Mac ➜ parallelBGL-Mac make Scanning dependencies of target boost_graph_mpi [ 2%] Building CXX object libs/graph_parallel/src/CMakeFiles/boost_graph_mpi.dir/mpi_process_group.cpp.o Linking CXX static library ../../../lib/libboost_graph_mpi.a [ 5%] Built target boost_graph_mpi Scanning dependencies of target adjlist_build_test [ 8%] Building CXX object libs/graph_parallel/test/CMakeFiles/adjlist_build_test.dir/adjlist_build_test.cpp.o In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/libs/graph_parallel/test/adjlist_build_test.cpp:10: In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjacency_list.hpp:51: /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/named_graph.hpp:345:12: warning: reference 'self' is not yet bound to a value when used here [-Wuninitialized] : self(self), name(other.name), committed(other.committed) ^ In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/libs/graph_parallel/test/adjlist_build_test.cpp:10: In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjacency_list.hpp:3961: /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjlist/serialization.hpp:97:27: error: no viable conversion from 'boost::filesystem::path' to 'std::string' (aka 'basic_string<char>') std::string process_name = i->path().filename(); ^ ~~~~~~~~~~~~~~~~~~~~ /usr/include/c++/4.2.1/bits/basic_string.h:434:7: note: candidate constructor not viable: no known conversion from 'boost::filesystem::path' to 'const std::basic_string<char> &' for 1st argument basic_string(const basic_string& __str); ^ /usr/include/c++/4.2.1/bits/basic_string.h:469:7: note: candidate constructor not viable: no known conversion from 'boost::filesystem::path' to 'const char *' for 1st argument basic_string(const _CharT* __s, const _Alloc& __a = _Alloc()); ^ In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/libs/graph_parallel/test/adjlist_build_test.cpp:10: In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjacency_list.hpp:3961: /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjlist/serialization.hpp:841:37: error: no type named 'retag_property_list' in namespace 'boost::detail' typedef typename boost::detail::retag_property_list< ~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~ /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjlist/serialization.hpp:841:56: error: expected unqualified-id typedef typename boost::detail::retag_property_list< ^ /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjlist/serialization.hpp:876:25: error: unknown type name 'edge_property_type'; did you mean 'edge_property_tag'? , static_cast<edge_property_type const&>(get(edge_all_t(), g, e)) ^~~~~~~~~~~~~~~~~~ edge_property_tag /usr/local/include/boost/graph/properties.hpp:69:10: note: 'edge_property_tag' declared here struct edge_property_tag { }; ^ In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/libs/graph_parallel/test/adjlist_build_test.cpp:10: In file included from /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjacency_list.hpp:3961: /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjlist/serialization.hpp:955:37: error: no type named 'retag_property_list' in namespace 'boost::detail' typedef typename boost::detail::retag_property_list< ~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~ /Users/varoudis/Dropbox/Projects/Libraries/parallelBGL/boost/graph/distributed/adjlist/serialization.hpp:955:56: error: expected unqualified-id typedef typename boost::detail::retag_property_list< ^ 1 warning and 6 errors generated. make[2]: *** [libs/graph_parallel/test/CMakeFiles/adjlist_build_test.dir/adjlist_build_test.cpp.o] Error 1 make[1]: *** [libs/graph_parallel/test/CMakeFiles/adjlist_build_test.dir/all] Error 2 make: *** [all] Error 2 ########################## Best, Tasos

2 1

Re: [Boost-users] time out for r_c_shortest_paths
by Brammert Ottens 25 Feb '13

25 Feb '13

> Hi, > > I want to be able to put a time out on the r_c_shortest_paths, and still > be able to get the optimal paths that have been found so far. Currently, > I changed the REF such that it always returns false as soon as the time > out has reach, in effect emptying the unprocessed_labels queue. The > problem with this is that the algorithm still needs quite some time to > actually empty the queue. > > I tried to have the on_label_popped(cur_label) throw an exception when > the timeout has been reached. The timeout itself works like a charm, but > because the pareto_optimal_solutions container is only filled after the > while loop I can't access the already found optimal solutions. You might want to have on_label_popped(l, g) take l by reference and set l.b_is_dominated to true. You'll still need to empty the queue, but (if that approach works) it should be faster that way. -- Jeremiah Willcock Hi Jeremiah, Thanks for your suggestion, but if I understand correctly it would mean a change to the code. I do not want to change the code, because if ever there is an update to boost I just want to be able to replace the old libraries with the new ones. If I would change the code, I would keep the exception and put a try/catch statement around the while-loop. Then you go around the problem of having to empty the queue. Is there any change to get something like a graceful timeout supported in the boost graph algorithms? Brammert Ottens Hi, I've had a closer look and I was too quick in dismissing your solution. however, the on_label_popped(cur_label) gets a const reference to the label. I've used a const_cast to remove the const (see code below), but now the code crashes when cleaning up the labels using deallocate. Any ideas on what could be wrong? template<class Label, class Graph> void QRCSPPVisitor::on_label_popped( const Label& label, const Graph& graph ) { RCSPPVisitorData().IncrementLabelsPopped(); if ( TimedOut() || GetMillis(StartTime()) > MaxMillis() ) { this->SetTimedOut(); const_cast<Label&>(label).b_is_dominated = true; } } regards, Brammert Ottens

2 1

Re: [Boost-users] time out for r_c_shortest_paths
by Brammert Ottens 25 Feb '13

25 Feb '13

On Thu, 21 Feb 2013, Brammert Ottens wrote: > Hi, > > I want to be able to put a time out on the r_c_shortest_paths, and still > be able to get the optimal paths that have been found so far. Currently, > I changed the REF such that it always returns false as soon as the time > out has reach, in effect emptying the unprocessed_labels queue. The > problem with this is that the algorithm still needs quite some time to > actually empty the queue. > > I tried to have the on_label_popped(cur_label) throw an exception when > the timeout has been reached. The timeout itself works like a charm, but > because the pareto_optimal_solutions container is only filled after the > while loop I can't access the already found optimal solutions. You might want to have on_label_popped(l, g) take l by reference and set l.b_is_dominated to true. You'll still need to empty the queue, but (if that approach works) it should be faster that way. -- Jeremiah Willcock Hi Jeremiah, Thanks for your suggestion, but if I understand correctly it would mean a change to the code. I do not want to change the code, because if ever there is an update to boost I just want to be able to replace the old libraries with the new ones. If I would change the code, I would keep the exception and put a try/catch statement around the while-loop. Then you go around the problem of having to empty the queue. Is there any change to get something like a graceful timeout supported in the boost graph algorithms? Brammert Ottens

2 1

Boost.Geometry: please help; using buffer() extensions
by John Lilley 25 Feb '13

25 Feb '13

Something of a repost, but I've learned a bit more on the path to being truly confused. It appears that boost\geometry\extensions\algorithms\buffer\buffer_inserter.hpp contains some specializations for buffering other shape types? But I'm out of my depth here... is there an example? What do I need to do in order to buffer: a point? A linestring? A vector of polygons? A heterogeneous collections of shapes? Is this generalization of buffer also going to work as a "deflate" for polygons? If someone can tell me "include these headers, use these algorithms", it will help a lot! Thanks John

2 1

[Fusion] Associate arbitrary compile/runtime annotations with members of fusion sequence
by Kozlov Taras 24 Feb '13

24 Feb '13

Started to use Boost.Fusion in my project I quick realized that it would be very convenient if fusion adapt macroses like BOOST_FUSION_DEFINE_STRUCT would give ability to associate additional information with members. Second problem is that BOOST_FUSION_DEFINE_STRUCT and such doesn`t support inheritance. So I did a small utility library that allow to define fusion sequences with arbitrary list of annotations associated with each member, and want to share it with community :) http://tarasko.github.com/cppan/ Will appreciate for any comments

3 6

[MSM] How to access event-attributes through a "kleene"-event?
by Deniz Bahadir 24 Feb '13

24 Feb '13

Hello everyone, I am trying to use Boost.MSM with the eUML front-end and I must say: That is a pretty cool thing! :-) However, I stumbled across some points that raised several questions, which I hoped someone of you could answer. One of these points is the usage of the "kleene" event (with eUML). The questions that came to mind are the following: 1. Is there a way to access an attribute of an (original) event if the transition-table uses the "kleene" event? 2. If there is a way, can I access it directly from the guard or actions-list in the transition-table? 3. If there is not a way, is it planned to be implemented (if possible) and when? For better understanding I insert here a simple example which explains what I am trying to do: In my special case, all my events have an attribute "EventNo" which is a unique number for each event(-type). Now, I need a transition for the events with "EventNo" greater than a special number (e.g. 7). So I thought, I could do something like this: // The attribute: BOOST_MSM_EUML_DECLARE_ATTRIBUTE(int, EventNo) BOOST_MSM_EUML_ATTRIBUTES((attributes_ << EventNo), EventNoAttr) // All events that contain this attribute: BOOST_MSM_EUML_EVENT_WITH_ATTRIBUTES(event1, EventNoAttr) BOOST_MSM_EUML_EVENT_WITH_ATTRIBUTES(event2, EventNoAttr) /* ... */ BOOST_MSM_EUML_EVENT_WITH_ATTRIBUTES(eventN, EventNoAttr) // Entry-action of FSM which initializes these event-numbers: BOOST_MSM_EUML_ACTION(init_all_events) { template <class Evt, class Fsm, class State> void operator()(Evt const&, Fsm&, State&) { event1.get_attribute(EventNo) = 1; event2.get_attribute(EventNo) = 2; /* ... */ eventN.get_attribute(EventNo) = N; }; } // Some states: BOOST_MSM_EUML_STATE((), State1) BOOST_MSM_EUML_STATE((), State2) // The transition table which contains all the logic: BOOST_MSM_EUML_TRANSITION_TABLE(( State1 + kleene [event_(EventNo) > Int_<7>()] // <-- Not working! == State2, State1 + event1 == State3 /* event2 to event7 are ignored! */ ), stt) // The FSM front-end and back-end: BOOST_MSM_EUML_DECLARE_STATE_MACHINE(( stt, init_ << init_all_events ), FsmFrontend; typedef msm::back::state_machine<FsmFrontend> Fsm; However, the expression "event_(EventNo)" in the transition-guard fails to compile with the following error: error: no match for call to ‘(boost::msm::front::euml::GetEvent_<EventNo_>) (const boost::msm::front::euml::kleene_&, boost::msm::back::state_machine< boost::msm::front::euml::func_state_machine<FsmFrontendtag, ...’ I read in the documentation of the Functor front-end that I should use boost::any::type() to access the original event in transition actions, but I do not really know how to go on with it. I am also not sure if this applies to the eUML front-end at all (especially in the transition-guard). I hope someone can help me and answer these questions!? Thanks in advance, Deniz Bahadir

2 6

[asio] Adding methods directly to the IO_Service for it to call.
by Ryan 24 Feb '13

24 Feb '13

Boost Community, I've created a basic class that wrapped the boost asio framework receive/send functionality. My wrapper class gave the packet buffer to whatever processing function that it was created with and when the function was done sent any buffers it was given. This let me create classes that processed data without any knowledge of the transport layer. This concept has worked well since the asio sockets have an async_receive and async_send. I was able to setup all of my classes and then call io_service.run() to get everything going. I now have to receive data from a customer dll and they only support polling. Is there a way to add to the io_service a specific method to call once the io_service is running. I've included the ReceiveProcessSend.hpp with a comment above the two async_receive calls I need to replace. Also, below is a very basic example of how I setup classes that I pass into the ReceiveProcessSend. Any help on how to add calling methods to the io_service would be really appreciated. Ryan class DoSomeProcessing { private: typedef std::vector<unsigned char> t_ucvector; typedef boost::shared_ptr<t_ucvector> t_boostVector; typedef std::list<t_boostVector> t_boostList; public: t_boostList const processMessage(t_boostVector message, t_ucvector::size_type messageSize) { t_boostList outputList; //Do processing here and create output buffer. return outputList; } }; int main(int argc, char * argv[]) { DoSomeProcessing myProcess; ReceiveProcessSend::Arguments myArguments; boost::asio::io_service io_service; ReceiveProcessSend rps( io_service, boost::bind(&DoSomeProcessing::processMessage, &myProcess, _1, _2), myArguments); io_service.run(); return 0; }

1 0

Comparing two mpl::vector types
by Rodrigo Madera 23 Feb '13

23 Feb '13

Hello all, Can anyone give me a hint on how to call a functor on the differences of two same size mpl::vector types? For example, given: typedef boost::mpl::vector<int, char, short, double> v0_type; typedef boost::mpl::vector<int, int, short, std::string> v1_type; We can see that the second and third elements don't match, so I wish to call a functor twice, with the following signature: struct SomeFunctor { template <typename T, typename U> void operator()(T t, U u) const { // do something nice here } }; In this example, functor would be called twice: once with [T=char, U=int] and another with [T=short, U=std::string], along with their values. Thanks for any light, Rodrigo

2 2

Geometry buffer algorithm: complete or not?
by John Lilley 22 Feb '13

22 Feb '13

I was looking here: http://www.boost.org/doc/libs/1_52_0/libs/geometry/doc/html/geometry/refere… And saw this: "The current implemetation only enlarges a box, which is not defined by OGC. A next version of the library will contain a more complete implementation" It looks like Boost::Geometry is incapable of buffering an arbitrary polygon. But various forum discussions refer to the buffer() algorithm like it actually works for the general case. Which is correct? John Lilley

2 8

Building / installing only the geometry library
by John Lilley 22 Feb '13

22 Feb '13

Is there a recommended approach to building/installing only the geometry library and its required headers? I understand that geometry is header-only, so I don't really need the libraries at all, but the build process that I am seeing to move the headers to the target folder builds them anyway. I am trying to get a minimal checkin to our version control that contains only those headers needed for geometry. Thanks, john

2 2