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

Named semaphores in Windows Hi
by Kevin Williams 09 Dec '14

09 Dec '14

Hi I am looking at named semaphores in Windows. (Boost 1.57 with MSVC 2012) In a simple test, I create a named semaphore, and immediately do a "wait" after that, but the semaphore does not block. Am I missing something? Without a "post" surely it should block? Kind regards, Kevin

1 0

[Boost.Test] unexpected behavior after std::system(...)
by contini.mailing＠gmail.com 09 Dec '14

09 Dec '14

Hi, I'm using boost 1.57 with gcc 4.7.2 on a debian wheezy. I'm testing an external program that has a non-zero return code and i'm experiencing an unexpected behavior of Boost.Test library. If the test below is linked statically (commenting the first line of course) everything works fine and all the three tests are executed. If the test below is linked dynamically i get the following output: test0 test1 unknown location(0): fatal error in "test1": child has exited; pid: 32314; uid: 1000; exit value: 4 (here program hangs forever and the third test is never executed) Defining the environment variable BOOST_TEST_CATCH_SYSTEM_ERRORS=No circumvents the problem. But i don't want to define it on all of the team member's computers and test machines. This can be considered a boost bug? #define BOOST_TEST_DYN_LINK #define BOOST_TEST_MODULE test_module #define BOOST_TEST_IGNORE_NON_ZERO_CHILD_CODE #include <boost/test/unit_test.hpp> #include <stdlib.h> BOOST_AUTO_TEST_CASE( test0 ) { std::system("echo \"test0\";exit 0;"); } BOOST_AUTO_TEST_CASE( test1 ) { std::system("echo \"test1\";exit 1;"); } BOOST_AUTO_TEST_CASE( test2 ) { std::system("echo \"test2\";exit 2;"); } Any help is appreciated.

2 1

modifying brandes_betweenness_centrality_impl
by The Maschine 08 Dec '14

08 Dec '14

Hello all, Im trying to modify the brandes centrality in order to analyse only a subset of the graph based on a cutoff distance calculated by some other edge weight than the one used for the brandes call. What I have done so far is to add a shortest path search call before the main path search inside brandes_betweenness_centrality_impl. The first search uses the edge::m_metric_weight and constructs the distance map from 's' based on that weight. Then I feed the metric distance map in the 'brandes special visitor' and I check inside visitor's examine_vertex. Reminder that the brandes call uses the m_special_weight. While this works as a principle I have no idea how people with experience would do it! A problem with this is that while the metric cutoff can be small you still run both searches to the max (as if there is no cutoff). Any help on how to change the first 'dijkstra_shortest_paths' so that after reaching the cutoff the search stops gracefully and leaves the rest of the 'distance_map' to infinity? Thanks a lot and please let me know if something need more clarification. ############### struct EdgeProperties { double m_special_weight; double m_metric_distance_weight; }; typedef boost::adjacency_list<boost::vecS, boost::vecS, boost::undirectedS, VertexProperties, EdgeProperties> unGraph; // test my_brandes_dijkstra_visitor // template<typename Graph, typename WeightMap, typename IncomingMap, typename DistanceMap, typename PathCountMap> struct my_brandes_dijkstra_visitor : public bfs_visitor<> { typedef typename graph_traits<Graph>::vertex_descriptor vertex_descriptor; typedef typename graph_traits<Graph>::edge_descriptor edge_descriptor; my_brandes_dijkstra_visitor(std::stack<vertex_descriptor>& iordered_vertices, WeightMap iweight, IncomingMap iincoming, DistanceMap idistance, PathCountMap ipath_count, std::vector<float> c_dists, float coff) : ordered_vertices(iordered_vertices), weight(iweight), incoming(iincoming), distance(idistance), path_count(ipath_count), cutoff_dists(c_dists), cutoff(coff) { } /** * Whenever an edge e = (v, w) is relaxed, the incoming edge list * for w is set to {(v, w)} and the shortest path count of w is set to * the number of paths that reach {v}. */ void edge_relaxed(edge_descriptor e, const Graph& g) { vertex_descriptor v = source(e, g), w = target(e, g); incoming[w].clear(); incoming[w].push_back(e); put(path_count, w, get(path_count, v)); } /** * If an edge e = (v, w) was not relaxed, it may still be the case * that we've found more equally-short paths, so include {(v, w)} in the * incoming edges of w and add all of the shortest paths to v to the * shortest path count of w. */ void edge_not_relaxed(edge_descriptor e, const Graph& g) { typedef typename property_traits<WeightMap>::value_type weight_type; typedef typename property_traits<DistanceMap>::value_type distance_type; vertex_descriptor v = source(e, g), w = target(e, g); distance_type d_v = get(distance, v), d_w = get(distance, w); weight_type w_e = get(weight, e); closed_plus<distance_type> combine; if (d_w == combine(d_v, w_e)) { put(path_count, w, get(path_count, w) + get(path_count, v)); incoming[w].push_back(e); } } /// Keep track of vertices as they are reached void examine_vertex(vertex_descriptor w, const Graph&) { // add only if this is within the 'cutoff'. cutoff_dists holds // the metric distances from s. if( cutoff_dists[w] < cutoff ) ordered_vertices.push(w); } private: std::stack<vertex_descriptor>& ordered_vertices; WeightMap weight; IncomingMap incoming; DistanceMap distance; PathCountMap path_count; std::vector<float> cutoff_dists; float cutoff; }; //////////////////////////////////////////////////////////////////// template<typename WeightMap> struct my_brandes_dijkstra_shortest_paths { my_brandes_dijkstra_shortest_paths(WeightMap iweight_map) : weight_map(iweight_map) { } template<typename Graph, typename IncomingMap, typename DistanceMap, typename PathCountMap, typename VertexIndexMap> void operator()(Graph& g, typename graph_traits<Graph>::vertex_descriptor s, std::stack<typename graph_traits<Graph>::vertex_descriptor>& ov, IncomingMap incoming, DistanceMap distance, PathCountMap path_count, VertexIndexMap vertex_index , std::vector<float> c_dists, float coff) // modified cutoff vector and value { typedef my_brandes_dijkstra_visitor<Graph, WeightMap, IncomingMap, DistanceMap, PathCountMap> visitor_type; visitor_type visitor(ov, weight_map, incoming, distance, path_count,c_dists,coff); // modified cutoff vector and value dijkstra_shortest_paths(g, s, boost::weight_map(weight_map) .vertex_index_map(vertex_index) .distance_map(distance) .visitor(visitor)); } private: WeightMap weight_map; }; ////////////////////////////// template<typename Graph, typename CentralityMap, typename EdgeCentralityMap, typename IncomingMap, typename DistanceMap, typename DependencyMap, typename PathCountMap, typename VertexIndexMap, typename ShortestPaths> void my_brandes_betweenness_centrality_impl(const Graph& g, CentralityMap centrality, // C_B EdgeCentralityMap edge_centrality_map, IncomingMap, //incoming, // P // OpenMP mod DistanceMap, //distance,// d // OpenMP mod DependencyMap, //dependency,// delta // OpenMP mod PathCountMap, //path_count,// sigma // OpenMP mod VertexIndexMap vertex_index, ShortestPaths shortest_paths) { typedef typename graph_traits<Graph>::vertex_iterator vertex_iterator; typedef typename graph_traits<Graph>::edge_iterator edge_iterator; typedef typename graph_traits<Graph>::vertex_descriptor vertex_descriptor; // Initialize centrality init_centrality_map(vertices(g), centrality); init_centrality_map(edges(g), edge_centrality_map); int i, N = num_vertices(g); #pragma omp parallel for default(shared) private(i) schedule(dynamic) for (i = 0; i < N; ++i) { typename graph_traits<Graph>::vertex_descriptor s = vertex(i, g); if (s == graph_traits<Graph>::null_vertex()) continue; // First we build a map of all distances from 's' based // on m_metric_distance_weight. typedef typename boost::property_map < Graph, boost::vertex_index_t >::type CIndexMap; typedef typename boost::iterator_property_map < vertex_descriptor*, CIndexMap, vertex_descriptor, vertex_descriptor& > CPredecessorMap; typedef typename boost::iterator_property_map < float*, CIndexMap, float, float& > CDistanceMap; float co = 2200.0f; std::vector<vertex_descriptor> cut_predecessors(boost::num_vertices(g)); // To store parents std::vector<float> cut_distances(boost::num_vertices(g), std::numeric_limits<double>::max() ); // To store distances CIndexMap cut_indexMap = boost::get(boost::vertex_index, g); CPredecessorMap cut_predecessorMap(&cut_predecessors[0], cut_indexMap); CDistanceMap cut_distanceMap(&cut_distances[0], cut_indexMap); // boost::dijkstra_shortest_paths(g, s, boost::weight_map(get(&EdgeProperties::m_metric_distance_weight, g)).predecessor_map(cut_predecessorMap).distance_map(cut_distanceMap)); /////////// // Local storage for OpenMP std::stack<vertex_descriptor> ordered_vertices; vector_property_map<typename property_traits<IncomingMap>:: value_type,VertexIndexMap> incoming(vertex_index); vector_property_map<typename property_traits<DistanceMap>:: value_type,VertexIndexMap> distance(vertex_index); vector_property_map<typename property_traits<DependencyMap>:: value_type,VertexIndexMap> dependency(vertex_index); vector_property_map<typename property_traits<PathCountMap>:: value_type,VertexIndexMap> path_count(vertex_index); // Initialize for this iteration vertex_iterator w, w_end; for (tie(w, w_end) = vertices(g); w != w_end; ++w) { incoming[*w].clear(); put(path_count, *w, 0); put(dependency, *w, 0); } put(path_count, s, 1); // co - cutoff added. b_centrality is calculated based on // Edge::m_special_weight while we are testing // to be inside the 'cutoff' that is based on the // Edge::m_metric_weight. shortest_paths(g, s, ordered_vertices, incoming, distance, path_count, vertex_index, cut_distances, co); while (!ordered_vertices.empty()) { vertex_descriptor u = ordered_vertices.top(); ordered_vertices.pop(); typedef typename property_traits<IncomingMap>::value_type incoming_type; typedef typename incoming_type::iterator incoming_iterator; typedef typename property_traits<DependencyMap>::value_type dependency_type; for (incoming_iterator vw = incoming[u].begin(); vw != incoming[u].end(); ++vw) { vertex_descriptor v = source(*vw, g); dependency_type factor = dependency_type(get(path_count, v)) / dependency_type(get(path_count, u)); factor *= (dependency_type(1) + get(dependency, u)); put(dependency, v, get(dependency, v) + factor); update_centrality(edge_centrality_map, *vw, factor); } if (u != s) { #pragma omp critical(globalupdate) { update_centrality(centrality, u, get(dependency, u)); } } } } typedef typename graph_traits<Graph>::directed_category directed_category; const bool is_undirected = is_convertible<directed_category*, undirected_tag*>::value; if (is_undirected) { divide_centrality_by_two(vertices(g), centrality); divide_centrality_by_two(edges(g), edge_centrality_map); } } //

1 1

errors while compiling the boost-libraries
by steffen.wittek＠posteo.de 08 Dec '14

08 Dec '14

I was trying to compile the boost-libraries in order to use them in dolfin (1.4.0), as I got an error message in Dolfin as follows: > ImportError: /home/phoenix/Work/FEniCS/lib/libdolfin.so.1.4: > undefined symbol: _ZN5boost6thread4joinEv I thought that the undefined symbol pointed towards the boost libraries and so I tried to recompile them by including them into the dorsal script. A compilation of the latter stopped with another error however. The ..Work/FEniCS/src/boost_1_55_0/dorsal_build.log ends with > ...failed updating 45 targets... > ...skipped 50 targets... > ...updated 564 targets... and can be seen in full glory here: https://www.dropbox.com/s/529m6pb4ygcsxxv/dorsal_build.log?dl=0 Does anyone know how I can fix this problem?

1 0

[ASIO][Windows] async_resolve() Blocking on Reverse Lookup?
by Kyle Ketterer 06 Dec '14

06 Dec '14

Hello, I am working on a simple utility which scans an IP range and performs a reverse lookup. Using ASIO, I would like to use async_resolve to accomplish this. However, it seems that when using async_resolve() with an ::endpoint overload, the call is blocking. I am wrapping all async functions in a strand and calling io_service::run() from multiple threads. Here is an example: ------------------------------------------------------------------- class client { public: client ( boost::asio::io_service & io_service ) : io_service_(io_service), resolver_(io_service), strand_(io_service) { /* */ } void resolve_host ( const std::string & host ) { strand_.post( boost::bind(&client::start_resolve, this, host ) ); } void start_resolve ( const std::string & host_ ) { //boost::asio::ip::tcp::resolver::query query( tcp::v4(), host_, "" ); tcp::endpoint endpoint ( boost::asio::ip::address::from_string( host_) , 0 ); resolver_.async_resolve( endpoint, strand_.wrap( boost::bind( &client::handle_resolve, this, boost::asio::placeholders::error, boost::asio::placeholders::iterator ) ) ); } void handle_resolve ( const boost::system::error_code& err, tcp::resolver::iterator endpoint_iterator ) { if ( !err ) { /* attempt to connect */ tcp::endpoint endpoint = *endpoint_iterator; cout << endpoint.address() << "' >> " << (*endpoint_iterator).host_name() << endl; } } private: tcp::resolver resolver_; boost::asio::strand strand_; boost::asio::io_service & io_service_; }; int main( int argc, char* argv[] ) { /* io service */ boost::asio::io_service io_service; /* ip range */ std::string m_addr = "10.1.10."; /* client object */ boost::shared_ptr<client> m_client ( new client ( io_service ) ); /* add the entire subnet */ for ( int i=1; i<255; i++ ) { m_client->resolve_host ( m_addr + boost::to_string(i) ); } /* create threads and run io_service */ std::vector < boost::shared_ptr<boost::thread> > m_threads; for ( int i=0; i < CPU_CORES; i++ ) { m_threads.push_back ( boost::shared_ptr<boost::thread> ( new boost::thread(boost::bind(&boost::asio::io_service::run, &io_service ) ) ) ); } for ( int i=0; i < CPU_CORES; i++ ) { m_threads[i]->join(); } return 0; } ------------------------------------------------------ I am looking for suggestions on how I can take a multi-threaded approach to this. I am fairly new to ASIO so I appreciate any advice! Thanks -- Kyle Ketterer <reminisc3(a)gmail.com>

1 0

phoenix bind template parameter
by Philipp Kraus 06 Dec '14

06 Dec '14

Hello, I try to use phoenix bind to bind std::pow for double values, but I get errors, because the template parameters not set: phoenix::bind(&std::pow, _val, _1) I would like to compile it with C++11. How can I setup the template parameter on the phoenix bind? At the moment I have defined a function double mypow(double a, double b) { return std::pow(a,b); } but can I do this directly with std::pow in the bind call? Thanks a lot Phil

2 1

[Boost.SIMD] Interest in library
by dariomt＠gmail.com 05 Dec '14

05 Dec '14

Hi, I'm interested in trying out Boost.SIMD. Where can I download the latest stable version? Any idea when it might get proposed for Boost review? I think it would be beneficial if it is added to the Boost Incubator until then. Also, in http://meetingcpp.com/tl_files/mcpp/slides/12/simd.pdf slide 36 I see mention of a vectorized Thomas Tri-diagonal Solver. I thought the data dependency made the algorithm unfeasible for vectorization, so I'm *very* interested to learn how Boost.SIMD helps in this case. Is there more information available about this? Thanks for working in a library that shows a lot of potential.

4 3

Boost Serialization - Overriding builtin implementation for std::wstring
by cpp_learner 04 Dec '14

04 Dec '14

Hi: The current serialization implementation for std::wstring is not portable across platforms. To make it cross platform, we would like to customize the serialization implementation for std::wstring (by splitting "serialize" and providing our own implementation for "save/load"). Is it possible to achieve this? One way to get around this is to provide our own implementation, and not include the header boost/serialization/string.hpp. This is not an option to us as there is no way for us to enforce clients to not include this header at all either directly or indirectly. And unfortunately, the compiler does not complain if the same compilation unit has two different serialization mechanism for std::wstring (the builtin and customized serialization implementation, in which case the builtin serialization implementation is picked, which is not portable). Many Thanks AB

1 0

Fusion Move Semantics
by Ovanes Markarian 04 Dec '14

04 Dec '14

Hi! I would like to ask does fusion supports move semantics, or are all the values going to get copied. I use C++11 (clang with libstdcxx) and boost 1.54 (but can upgrade). In my use case I have boost::fusion::map and was wondering, that in case of moving of that map, is it going to forward the move call to contained values or not. Many thanks for help, Ovanes

3 2

Boost BGL question regarding depth_first_visit / undirected_depth_first_visit
by Mark Fanara 03 Dec '14

03 Dec '14

I am looking to add a TerminatorFunc to undirected_depth_first_visit similar to what is in depth_first_visit. I think I have successfully made the changes through mimicry (I am an old school C++ guy and still don't understand generic programming) My problem is that I don't know how to declare/define my own terminate function. Nor do I know how to call undirected_depth_first_visit with the new argument. I have tried searching for examples of calling depth_first_visit with a TerminatorFunc argument, but can't find anything helpful. Appreciate any suggestions you might have.

1 0

Marriage of boost::any_range and boost::fusion?
by Ernest Zaslavsky 03 Dec '14

03 Dec '14

I'm trying to achieve the following. I have an any_range of const items, in runtime I need dynamically calculate and store values for each item, lets say two additional integers, what I expect to get in the end is an any_range with something like joint view of my const item and two integers. Is it possible? [Read More]<http://feeds.feedburner.com/~r/sizmek-blog/~6/1> [http://www.sizmek.com/Sizmek.png]<http://www.sizmek.com/>

1 0

Re: [Boost-users] how to lisk unique keys in boost multi_index_container
by Bill Tian 03 Dec '14

03 Dec '14

thanks for JoaquÃn M LÃ³pez MuÃ±oz 's reply. i have tried your code, it works as i need exactly. for(auto it=data.get<1>().begin(),it_end=data.get<1>().end(); it!=it_end; it=data.get<1>().upper_bound(it->m2)){ std::cout<<it->m2<<"\n"; // or whatever you need to do with the key } i'll use this method in reality, thanks! Bill

1 0

how to lisk unique keys in boost multi_index_container
by Bill Tian 02 Dec '14

02 Dec '14

hello,every one now i have a struct like this struct a { a(const std::string &m1_, const std::string &m2_, boost::int16_t m3_) : m1(m1_),m2(m2_), m3(m3_) {} std::string m1; std::string m2; boost::int16_t m3; }; typedef boost::multi_index::multi_index_container < a, boost::multi_index::indexed_by< boost::multi_index::hashed_unique< BOOST_MULTI_INDEX_MEMBER(a, std::string, m1) >, boost::multi_index::ordered_non_unique< BOOST_MULTI_INDEX_MEMBER(a, std::string, m2) >, boost::multi_index::ordered_non_unique < BOOST_MULTI_INDEX_MEMBER(a, boost::uint16_t, m3) > > > my_data; i want to list all unique keys of m2 and m3 after the container has been filled with a lot of records, is there a example ? my_data data; data.insert(a("a1","k1", 1)); data.insert(a("a2","k1", 2)); data.insert(a("a3","k2", 3)); ..... data.insert(a("a10000","k1", 1)); i want to get all data that m2 equal to "k1", when this data has been processed, get all data that m2 equal to "k2" ... but i do not how many unique keys in the container, am i need to store another set to store that when i fill the container? i need a function to get the keys like the SQL:"select distinct(m2) from data". thanks a lot! best regards. Bill

2 1

Error "Used Uninitialized" in boyer_myrvold_impl.hpp
by wanglurg 02 Dec '14

02 Dec '14

Hi, guys, I'm a new boost graph library user. I met an error when using boyer_myrvold_planar_test() function in Boost_1_55_0 to generate the kuratowski subgraph. Error message is listed as follows: ./boost/graph/planar_detail/boyer_myrvold_impl.hpp:1545:11 error: 'z' may be used uninitialized in this function. I added a "Wno-maybe-uninitialized" flag to Makefile to remove this error temporarily, but it's not a good method. I'm looking for a better way to fix this. Does anyone encounter the same case? Best regards, Lu Wang -- View this message in context: http://boost.2283326.n4.nabble.com/Error-Used-Uninitialized-in-boyer-myrvol… Sent from the Boost - Users mailing list archive at Nabble.com.

2 2

Does Boost 1.56 boost::async support auto scheduling (launch deferred)?
by Adi Shavit 02 Dec '14

02 Dec '14

Hi, Onc of the design choices for async() default launch policy was to allow the OS scheduler to decide whether or not to launch a new thread or remain on the same thread (presumably based on available system resources). Does Boost 1.56 or 1.57 support this default auto scheduling yet? It seems to be launching as many threads as tasks. Thanks, Adi

2 2

Boost build and first program
by Great Avenger Singh 01 Dec '14

01 Dec '14

Hi I am first time boost user. I build it using following procedure on Ubuntu Linux 14.04. cd /boost-src/ ./bootstrap.sh and followed instructions. after the completion it gave me following suggestion: The Boost libraries were successfully built. Following directory should be included to compiler include paths; /home/metal-machine/boost-src Following directory should be included to linker library paths; /home/metal-machine/boost-src/stage/lib Now I try to compile and program with following options from command line: $ g++ -I "/home/metal-machine/boost-src/" -L "/home/metal-machine/boost-src/stage/lib/" boostlist.cpp -lboost_system -lboost_filesystem -o honey $ ./honey ./honey: error while loading shared libraries: libboost_system.so.1.57.0: cannot open shared object file: No such file or directory But on the other hand if I compile and run as follows: $ g++ boostlist.cpp -lboost_system -lboost_filesystem -o honey $./honey ->it works fine I am using thiscode: https://gist.github.com/vivithemage/9517678 Where I am doing wrong? -- Thanks Arshpreet Singh http://arshpreetsingh.wordpress.com/ “If we all did the things we are really capable of doing, we would literally astound ourselves …” T. E. ------------------------------------------------------------------------------ Slashdot TV. Video for Nerds. Stuff that matters. http://tv.slashdot.org/

3 2

fatal error: 'boost/utility/detail/result_of_iterate.hpp' file not found
by Ted Middleton 01 Dec '14

01 Dec '14

I'm trying to build boost 1.56.0 for OSX 10.10 with Xcode 5.1.1 (clang-503.0.40). Xcode 5.1.1 Build version 5B1008 Apple LLVM version 5.1 (clang-503.0.40) (based on LLVM 3.4svn) Target: x86_64-apple-darwin14.0.0 Thread model: posix When I try to build boost thread, the compile fails building thread.cpp on the command: "g++" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -pedantic -gdwarf-2 -fexceptions -Wno-long-long -Wextra -Wno-long-long -Wno-variadic-macros -Wunused-function -fpermissive -pedantic -DBOOST_ALL_NO_LIB=1 -DBOOST_ATOMIC_STATIC_LINK=1 -DBOOST_SYSTEM_STATIC_LINK=1 -DBOOST_THREAD_BUILD_LIB=1 -DBOOST_THREAD_DONT_USE_CHRONO -DBOOST_THREAD_POSIX -DNDEBUG -I"." -c -o "bin.v2/libs/thread/build/darwin-4.2.1/release/link-static/threading-multi/pthread/thread.o" "libs/thread/src/pthread/thread.cpp" The error is: In file included from libs/thread/src/pthread/thread.cpp:19: In file included from ./boost/thread/future.hpp:22: In file included from ./boost/thread/detail/invoker.hpp:28: In file included from ./boost/utility/result_of.hpp:202: ./boost/preprocessor/iteration/detail/iter/forward1.hpp:47:18: fatal error: 'boost/utility/detail/result_of_iterate.hpp' file not found ./boost/preprocessor/iteration/detail/iter/forward1.hpp:27:56: note: expanded from macro 'BOOST_PP_FILENAME_1' # define BOOST_PP_FILENAME_1 BOOST_PP_ARRAY_ELEM(2, BOOST_PP_ITERATION_PARAMS_1) ^ ./boost/utility/result_of.hpp:201:68: note: expanded from macro 'BOOST_PP_ITERATION_PARAMS_1' #define BOOST_PP_ITERATION_PARAMS_1 (3,(0,BOOST_RESULT_OF_NUM_ARGS,<boost/utility/detail/result_of_iterate.hpp>)) ^ ./boost/preprocessor/array/elem.hpp:23:114: note: expanded from macro 'BOOST_PP_ARRAY_ELEM' # define BOOST_PP_ARRAY_ELEM(i, array) BOOST_PP_TUPLE_ELEM(BOOST_PP_ARRAY_SIZE(array), i, BOOST_PP_ARRAY_DATA(array)) ^ note: (skipping 10 expansions in backtrace; use -fmacro-backtrace-limit=0 to see all) ./boost/preprocessor/tuple/rem.hpp:23:31: note: expanded from macro 'BOOST_PP_REM' # define BOOST_PP_REM(...) __VA_ARGS__ ^ ./boost/preprocessor/variadic/elem.hpp:26:89: note: expanded from macro 'BOOST_PP_VARIADIC_ELEM' # define BOOST_PP_VARIADIC_ELEM(n, ...) BOOST_PP_CAT(BOOST_PP_VARIADIC_ELEM_, n)(__VA_ARGS__,) ^ ./boost/preprocessor/variadic/elem.hpp:30:55: note: expanded from macro 'BOOST_PP_VARIADIC_ELEM_2' # define BOOST_PP_VARIADIC_ELEM_2(e0, e1, e2, ...) e2 Sure enough, boost/utility/detail/result_of_iterate.hpp doesn't exist from the base of the project. It does if you start at libs/utlility/include/boost/utility/detail, though. Most of what's in boost/ is a symlink to directories in libs (I guess this is how modulization works), and if boost/utility were a symlink libs/utility/include/boost/utility, I think this code in forward1.hpp would work. But boost/utility isn't a symlink - boost/utility is an actual directory with a few symlinks to files in it, but none of them leads to result_of_iterate.hpp. There's an email (http://lists.boost.org/Archives/boost/2014/07/215233.php) to the developers list from back in July discussing this build break in the 1.56.0 beta 1, but I'm seeing this at the tag boost-1.56.0 for the 1.56.0 release in git.I don't seem to be able to build 1.57.0 either. Boost 1.56.0 wouldn't have shipped with a broken boost thread compile on OSX 64-bit, so I'm certain that its something bad that I've done. Can anyone figure out what I'm doing wrong from that command line or the error printout?

1 1

[Boost-users][Spirit] Semantic Action for a map of a map
by Ryan 01 Dec '14

01 Dec '14

I posted this question to the Boost Spirit mailing list but there doesn't seem to be much traffic there. I'm hoping that the main mailing list also has a few Spirit users as well and can possible receive some help. I'm wanting to add a semantic action for a complex rule. I'm not entirely sure this is the correct design. I need to parse the following EBNF grammar. EBNF value = (a-zA-Z_0-9)* tag = (a-zA-Z_) value tagValue = tag '=' value list = "@list=" tag '=' tag ',' (tagValue ',')* "@end" info = tag (',' (tagValue | list) )* I thought if I defined the basic rules I could then created a custom instance of that rule with a semantic action. I did that for the first "tag" in the info definition and it seem to work well. I was also able to create this custom instance for the "tagValue" rule within the info definition. The problem is how to create the custom instance for list. The list result goes into a map of a map that I've created within my class. I'm not sure how to treat the tag=tag as the key and let the tagValue be the value. Any help with the semantic action or changing the design would be appreciated. This is my first time using Spirit and the learning curve is proving to be difficult. Ryan #include <string> #include <map> #include <utility> #include <boost/spirit/include/phoenix.hpp> #include <boost/spirit/include/qi.hpp> #include <boost/spirit/include/std_pair.hpp> class Bob { typedef std::pair<std::string, std::string> infoPair; typedef std::map<std::string, std::string> infoMap typedef std::map<std::string, infoMap> infoMap2; std::string m_Name; infoMap m_Arguments; infoMap2 m_ArgumentCollection; public: bool const parseMessage(std::string const& message) { namespace qi = boost::spirit::qi; namespace phx = boost::phoenix; typedef std::string::const_iterator iter; qi::rule<iter, std::string()> value = *qi::char_("a-zA-Z_0-9"); qi::rule<iter, std::string()> tag = qi::char_("a-zA-Z_") >> value; qi::rule<iter, infoPair()> tagValue = tag >> '=' >> value; qi::rule<iter, std::string()> name = tag[ phx::bind(&Bob::m_Name, this) = qi::_1 ]; qi::rule<iter, infoMap()> tagValueMap = tagValue[ [=](infoPair& p, qi::unused_type, qi::unused_type) { m_Arguments[p.first] = p.second; } ]; //The compiler wants an infoPair. Not really sure that is what I need. qi::rule<iter, infoPair()> list = "@list=" >> tag >> '=' >> tag >> ',' >> *(tagValue >> ',') >> "@end"; //How can I create a custom version of list (like I did for tag) to assign a semantic action. //The action needs to allow tag=tag be the key to m_ArgumentCollection and place all //the repeating tagValue items into its value. auto first(message.begin()), last(message.end()); bool results = qi::parse(first, last, name >> (*(',' >> tagValueMap))); //Something like this as the final form to the grammar. //bool results = qi::parse(first, last, name >> (*(',' >> (tagValueMap | customList)))); if (!results || first != last) return false; return true; } }; int main(void) { Bob bob; bob.parseMessage("hello,this=great,@list=alice=here,wish=fish,@end,look=good"); return 0; }

1 0

LibBoost - Asio - SerialPort - set_option throw an exception
by Mikael Plouhinec 01 Dec '14

01 Dec '14

Hello, I have some problem using the libboost asio to communicate with serial port. I want to open my serial port with : m_Balance.open(BalancePort, (unsigned int)9600, boost::asio::serial_port_base::parity(boost::asio::serial_port_base::parity::even), boost::asio::serial_port_base::character_size((unsigned int)7), boost::asio::serial_port_base::flow_control( boost::asio::serial_port_base::flow_control::hardware), boost::asio::serial_port_base::stop_bits(boost::asio::serial_port_base::stop_bits::one)); but I have an exception with the character_size : Invalid Argument I put some log in the open method and I have found that my problem is on the set_option of the character size. Can someone help me? Thanks a lot. Mike Libboost is on Debian Wheezy : version 1.49 I use the following Serial Class (cpp) : /* * File: Serial.cpp * Author: Terraneo Federico * Distributed under the Boost Software License, Version 1.0. * Created on September 12, 2009, 3:47 PM * * v1.05: Fixed a bug regarding reading after a timeout (again). * * v1.04: Fixed bug with timeout set to zero * * v1.03: Fix for Mac OS X, now fully working on Mac. * * v1.02: Code cleanup, speed improvements, bug fixes. * * v1.01: Fixed a bug that caused errors while reading after a timeout. * * v1.00: First release. */ #include "Serial.h" #include <string> #include <algorithm> #include <iostream> #include <boost/bind.hpp> using namespace std; using namespace boost; CSerial::CSerial(): io(), port(io), timer(io), timeout(posix_time::seconds(0)) {} CSerial::CSerial(const std::string& devname, unsigned int baud_rate, asio::serial_port_base::parity opt_parity, asio::serial_port_base::character_size opt_csize, asio::serial_port_base::flow_control opt_flow, asio::serial_port_base::stop_bits opt_stop) : io(), port(io), timer(io), timeout(posix_time::seconds(0)) { open(devname,baud_rate,opt_parity,opt_csize,opt_flow,opt_stop); } void CSerial::open(const std::string& devname, unsigned int baud_rate, asio::serial_port_base::parity opt_parity, asio::serial_port_base::character_size opt_csize, asio::serial_port_base::flow_control opt_flow, asio::serial_port_base::stop_bits opt_stop) { if(isOpen()) close(); port.open(devname); Log.AddLine("set 1", LOG_DEBUG); port.set_option(asio::serial_port_base::baud_rate(baud_rate)); Log.AddLine("set 2", LOG_DEBUG); port.set_option(opt_parity); Log.AddLine("set 3", LOG_DEBUG); port.set_option(opt_csize); Log.AddLine("set 4", LOG_DEBUG); port.set_option(opt_flow); Log.AddLine("set 5", LOG_DEBUG); port.set_option(opt_stop); Log.AddLine("set 6", LOG_DEBUG); } bool CSerial::isOpen() const { return port.is_open(); } void CSerial::close() { if(isOpen()==false) return; port.close(); } void CSerial::setTimeout(const posix_time::time_duration& t) { timeout=t; } void CSerial::write(const char *data, size_t size) { asio::write(port,asio::buffer(data,size)); } void CSerial::write(const std::vector<char>& data) { asio::write(port,asio::buffer(&data[0],data.size())); } void CSerial::writeString(const std::string& s) { asio::write(port,asio::buffer(s.c_str(),s.size())); } void CSerial::read(char *data, size_t size) { if(readData.size()>0)//If there is some data from a previous read { istream is(&readData); size_t toRead=min(readData.size(),size);//How many bytes to read? is.read(data,toRead); data+=toRead; size-=toRead; if(size==0) return;//If read data was enough, just return } setupParameters=ReadSetupParameters(data,size); performReadSetup(setupParameters); //For this code to work, there should always be a timeout, so the //request for no timeout is translated into a very long timeout if(timeout!=posix_time::seconds(0)) timer.expires_from_now(timeout); else timer.expires_from_now(posix_time::hours(100000)); timer.async_wait(boost::bind(&CSerial::timeoutExpired,this, asio::placeholders::error)); result=resultInProgress; bytesTransferred=0; for(;;) { io.run_one(); switch(result) { case resultSuccess: timer.cancel(); return; case resultTimeoutExpired: port.cancel(); throw(timeout_exception("Timeout expired")); case resultError: timer.cancel(); port.cancel(); throw(boost::system::system_error(boost::system::error_code(), "Error while reading")); //if resultInProgress remain in the loop } } } std::vector<char> CSerial::read(size_t size) { vector<char> result(size,'\0');//Allocate a vector with the desired size read(&result[0],size);//Fill it with values return result; } std::string CSerial::readString(size_t size) { string result(size,'\0');//Allocate a string with the desired size read(&result[0],size);//Fill it with values return result; } std::string CSerial::readStringUntil(const std::string& delim) { // Note: if readData contains some previously read data, the call to // async_read_until (which is done in performReadSetup) correctly handles // it. If the data is enough it will also immediately call readCompleted() setupParameters=ReadSetupParameters(delim); performReadSetup(setupParameters); //For this code to work, there should always be a timeout, so the //request for no timeout is translated into a very long timeout if(timeout!=posix_time::seconds(0)) timer.expires_from_now(timeout); else timer.expires_from_now(posix_time::hours(100000)); timer.async_wait(boost::bind(&CSerial::timeoutExpired,this, asio::placeholders::error)); result=resultInProgress; bytesTransferred=0; for(;;) { io.run_one(); switch(result) { case resultSuccess: { timer.cancel(); bytesTransferred-=delim.size();//Don't count delim istream is(&readData); string result(bytesTransferred,'\0');//Alloc string is.read(&result[0],bytesTransferred);//Fill values is.ignore(delim.size());//Remove delimiter from stream return result; } case resultTimeoutExpired: port.cancel(); throw(timeout_exception("Timeout expired")); case resultError: timer.cancel(); port.cancel(); throw(boost::system::system_error(boost::system::error_code(), "Error while reading")); //if resultInProgress remain in the loop } } } CSerial::~CSerial() {} void CSerial::performReadSetup(const ReadSetupParameters& param) { if(param.fixedSize) { asio::async_read(port,asio::buffer(param.data,param.size),boost::bind( &CSerial::readCompleted,this,asio::placeholders::error, asio::placeholders::bytes_transferred)); } else { asio::async_read_until(port,readData,param.delim,boost::bind( &CSerial::readCompleted,this,asio::placeholders::error, asio::placeholders::bytes_transferred)); } } void CSerial::timeoutExpired(const boost::system::error_code& error) { if(!error && result==resultInProgress) result=resultTimeoutExpired; } void CSerial::readCompleted(const boost::system::error_code& error, const size_t bytesTransferred) { if(!error) { result=resultSuccess; this->bytesTransferred=bytesTransferred; return; } //In case a asynchronous operation is cancelled due to a timeout, //each OS seems to have its way to react. #ifdef _WIN32 if(error.value()==995) return; //Windows spits out error 995 #elif defined(__APPLE__) if(error.value()==45) { //Bug on OS X, it might be necessary to repeat the setup //http://osdir.com/ml/lib.boost.asio.user/2008-08/msg00004.html performReadSetup(setupParameters); return; } #else //Linux if(error.value()==125) return; //Linux outputs error 125 #endif result=resultError; } Serial.h /* * File: Serial.h * Author: Terraneo Federico * Distributed under the Boost Software License, Version 1.0. * * Created on September 12, 2009, 3:47 PM */ #ifndef Serial_H #define Serial_H #include <stdexcept> #include <boost/utility.hpp> #include <boost/asio.hpp> #include "Log.h" /** * Thrown if timeout occurs */ class timeout_exception: public std::runtime_error { public: timeout_exception(const std::string& arg): runtime_error(arg) {} }; /** * Serial port class, with timeout on read operations. */ class CSerial: private boost::noncopyable { public: CSerial(); /** * Opens a serial device. By default timeout is disabled. * \param devname serial device name, example "/dev/ttyS0" or "COM1" * \param baud_rate serial baud rate * \param opt_parity serial parity, default none * \param opt_csize serial character size, default 8bit * \param opt_flow serial flow control, default none * \param opt_stop serial stop bits, default 1 * \throws boost::system::system_error if cannot open the * serial device */ CSerial(const std::string& devname, unsigned int baud_rate, boost::asio::serial_port_base::parity opt_parity= boost::asio::serial_port_base::parity( boost::asio::serial_port_base::parity::none), boost::asio::serial_port_base::character_size opt_csize= boost::asio::serial_port_base::character_size(8), boost::asio::serial_port_base::flow_control opt_flow= boost::asio::serial_port_base::flow_control( boost::asio::serial_port_base::flow_control::none), boost::asio::serial_port_base::stop_bits opt_stop= boost::asio::serial_port_base::stop_bits( boost::asio::serial_port_base::stop_bits::one)); /** * Opens a serial device. * \param devname serial device name, example "/dev/ttyS0" or "COM1" * \param baud_rate serial baud rate * \param opt_parity serial parity, default none * \param opt_csize serial character size, default 8bit * \param opt_flow serial flow control, default none * \param opt_stop serial stop bits, default 1 * \throws boost::system::system_error if cannot open the * serial device */ void open(const std::string& devname, unsigned int baud_rate, boost::asio::serial_port_base::parity opt_parity= boost::asio::serial_port_base::parity( boost::asio::serial_port_base::parity::none), boost::asio::serial_port_base::character_size opt_csize= boost::asio::serial_port_base::character_size(8), boost::asio::serial_port_base::flow_control opt_flow= boost::asio::serial_port_base::flow_control( boost::asio::serial_port_base::flow_control::none), boost::asio::serial_port_base::stop_bits opt_stop= boost::asio::serial_port_base::stop_bits( boost::asio::serial_port_base::stop_bits::one)); /** * \return true if serial device is open */ bool isOpen() const; /** * Close the serial device * \throws boost::system::system_error if any error */ void close(); /** * Set the timeout on read/write operations. * To disable the timeout, call setTimeout(boost::posix_time::seconds(0)); */ void setTimeout(const boost::posix_time::time_duration& t); /** * Write data * \param data array of char to be sent through the serial device * \param size array size * \throws boost::system::system_error if any error */ void write(const char *data, size_t size); /** * Write data * \param data to be sent through the serial device * \throws boost::system::system_error if any error */ void write(const std::vector<char>& data); /** * Write a string. Can be used to send ASCII data to the serial device. * To send binary data, use write() * \param s string to send * \throws boost::system::system_error if any error */ void writeString(const std::string& s); /** * Read some data, blocking * \param data array of char to be read through the serial device * \param size array size * \return numbr of character actually read 0<=return<=size * \throws boost::system::system_error if any error * \throws timeout_exception in case of timeout */ void read(char *data, size_t size); /** * Read some data, blocking * \param size how much data to read * \return the receive buffer. It iempty if no data is available * \throws boost::system::system_error if any error * \throws timeout_exception in case of timeout */ std::vector<char> read(size_t size); /** * Read a string, blocking * Can only be used if the user is sure that the serial device will not * send binary data. For binary data read, use read() * The returned string is empty if no data has arrived * \param size hw much data to read * \return a string with the received data. * \throws boost::system::system_error if any error * \throws timeout_exception in case of timeout */ std::string readString(size_t size); /** * Read a line, blocking * Can only be used if the user is sure that the serial device will not * send binary data. For binary data read, use read() * The returned string is empty if the line delimiter has not yet arrived. * \param delimiter line delimiter, default="\n" * \return a string with the received data. The delimiter is removed from * the string. * \throws boost::system::system_error if any error * \throws timeout_exception in case of timeout */ std::string readStringUntil(const std::string& delim="\n"); ~CSerial(); private: /** * Parameters of performReadSetup. * Just wrapper class, no encapsulation provided */ class ReadSetupParameters { public: ReadSetupParameters(): fixedSize(false), delim(""), data(0), size(0) {} explicit ReadSetupParameters(const std::string& delim): fixedSize(false), delim(delim), data(0), size(0) { } ReadSetupParameters(char *data, size_t size): fixedSize(true), delim(""), data(data), size(size) { } //Using default copy constructor, operator= bool fixedSize; ///< True if need to read a fixed number of parameters std::string delim; ///< String end delimiter (valid if fixedSize=false) char *data; ///< Pointer to data array (valid if fixedSize=true) size_t size; ///< Array size (valid if fixedSize=true) }; /** * This member function sets up a read operation, both reading a specified * number of characters and reading until a delimiter string. */ void performReadSetup(const ReadSetupParameters& param); /** * Callack called either when the read timeout is expired or canceled. * If called because timeout expired, sets result to resultTimeoutExpired */ void timeoutExpired(const boost::system::error_code& error); /** * Callback called either if a read complete or read error occurs * If called because of read complete, sets result to resultSuccess * If called because read error, sets result to resultError */ void readCompleted(const boost::system::error_code& error, const size_t bytesTransferred); /** * Possible outcome of a read. Set by callbacks, read from main code */ enum ReadResult { resultInProgress, resultSuccess, resultError, resultTimeoutExpired }; boost::asio::io_service io; ///< Io service object boost::asio::serial_port port; ///< Serial port object boost::asio::deadline_timer timer; ///< Timer for timeout boost::posix_time::time_duration timeout; ///< Read/write timeout boost::asio::streambuf readData; ///< Holds eventual read but not consumed enum ReadResult result; ///< Used by read with timeout size_t bytesTransferred; ///< Used by async read callback ReadSetupParameters setupParameters; ///< Global because used in the OSX fix }; #endif //CSerial_H -- [logo Aston] MIKAEL PLOUHINEC miplouhinec(a)sqli.com T�l : +33(0) 2 51 79 77 40 - Mob : +33(0) 6 80 23 21 28 www.sqli.com - www.aston-ecole.com

1 0

[boost::iterators] Missing typedefs in v1.57
by news.gmane.org 01 Dec '14

01 Dec '14

The typedefs for difference_type and reference used to be in scope for classes derived from iterator_facade. Since v1.57 they have to be declared thus: typedef typename iterator_facade_::difference_type difference_type; typedef typename iterator_facade_::reference reference; This seems out of place for a class that was designed to reduce boiler plate code. However, if the change was intentional, please document it. Thanks, Keith MacDonald

1 0

Reverse for loop with boost::adaptors::reverse crashes
by Filip Konvička 28 Nov '14

28 Nov '14

Hi, I was trying to replace BOOST_REVERSE_FOREACH with the standard range-based for loop. #include <boost/range/adaptor/reversed.hpp> #include <boost/foreach.hpp> #include <vector> #include <iostream> std::vector<int> getv() { return std::vector<int>{1,2,3}; } int main(int argc, char* argv[]) { // works fine: BOOST_REVERSE_FOREACH(int i, getv()) std::cout << i << std::endl; // crashes: for(int i : boost::adaptors::reverse(getv())) std::cout << i << std::endl; return 0; } Maybe this is basic C++ I should understand; it looks like the temporary vector returned by getv() gets destroyed in the 2nd loop too soon. I was wondering whether the adaptor could warn me that I'm using it in a wrong way (the code looks so nice and compiles cleanly...). Thanks, Filip

5 9

Boost.Log: default sink auto flush
by Mathias Gaunard 27 Nov '14

27 Nov '14

It would be very useful if the default sink could be made to auto flush. Is there any way to do this? I do not want to add a new sink, since the default sink cannot be replaced.

1 0

Unable to build universal binary on Mac OS X
by Kevin Molcard 27 Nov '14

27 Nov '14

Hi all, I am trying to compile boost on Mac OS X 10.9.5 for 32 and 64 bit. I am using the following command: ./b2 link=static threading=multi architecture=x86 address-model=32_64 I get the following error message: darwin.compile.c++.pch bin.v2/libs/math/build/darwin-4.2.1/release/address-model-32_64/architecture-x86/link-static/threading-multi/../src/tr1/pch.hpp.gch clang: error: cannot use 'precompiled-header' output with multiple -arch options "g++" -x c++-header -O3 -finline-functions -Wno-inline -Wall -gdwarf-2 -fexceptions -arch i386 -arch x86_64 -DBOOST_ALL_NO_LIB=1 -DBOOST_BUILD_PCH_ENABLED -DNDEBUG -I"." -I"libs/math/src/tr1" -c -o "bin.v2/libs/math/build/darwin-4.2.1/release/address-model-32_64/architecture-x86/link-static/threading-multi/../src/tr1/pch.hpp.gch" "libs/math/build/../src/tr1/pch.hpp" ...failed darwin.compile.c++.pch bin.v2/libs/math/build/darwin-4.2.1/release/address-model-32_64/architecture-x86/link-static/threading-multi/../src/tr1/pch.hpp.gch... Then 144 targets are skipped. Does anyone knows how to fix this? Thanks in advance Kevin

2 2

[review] [sort] Sort library review manager results
by Edward Diener 27 Nov '14

27 Nov '14

This is the results from the recent review of the Sort library of Steven Ross. First I would like to thank all those who made comments during the review, whether or not they officially gave a final Yes or No vote to whether the Sort library should be accepted as a Boost library. This list includes: Niall Douglas, Julian Gonggrijp, Phil Endecott, Vladimir Prus, Mathias Gaunard, Jeremy Murphy, Peter Dimov, Robert Ramey, Adam Walling, Anthony Polukhin, Phil Endecott, Paul Bristow, Thijs (M.A.) van den Berg Dupuis Etienne, and Frank Gennari If I have missed anyone I do apologize. Secondly I would like to thank Steven Ross for patiently answering all of the review comments to the best of his ability. My tally of Yes and No votes for acceptance are: Yes votes (5) : Niall Douglas ( conditional ), Julian Gonggrijp ( conditional ), Frank Gennari, Phil Endecott, Paul Bristow. No votes (3) : Vladimir Prus, Adam Walling ( conditional ), Anthony Polukhin I believe the condtional Yes vote from Julian Gonggrijp was completely met in the discussions of issues with Steven Ross and the conditional Yes vote from Niall Douglas was almost completely met in the discussion with Steven Ross about the issues mentioned, and enough to maintain his Yes vote. I want to also mention that the conditional No vote by Adam Walling has an implied Yes vote to it if the Sort library were one among other sort implementations. Since my final decision is not entirely based on the Yes and No votes I want to adumbrate some of the major issues brought up by the review without necessarily focussing on every one of the people who brought them up initially, as well as my own reactions to them as Review Manager. 1) The first major issue was whether a library whose basic merit lies in its algorithm and its speed/space constraints needs better theoretical backing. A number of reviewers discussed this after it was brought up. I tend to agree with reviewers that while the best theoretical basis is always desirable, it is not necessary for a Boost library whose empirical evidence can be and has been measured by its implementor and can be measured by any user. Furthermore Steven Ross has provided an extensive discussion in his documentation, as well as his original paper, about the theoretical merits of his technique. While much of this discussion is probably beyond the understanding of any but sorting experts and afficionados enough of it adequately explains the basic ideas behind SpreadSort for those who understand and have knowledge of basic sorting ideas and popular sorting techniques. 2) An important issue is the need for the Sorting library to provide extensive timing charts/tables comparing SpreadSort to at least std::sort ( and possibly other popular mainstream sorts ) with both different numbers of sort keys and different initial unsorted distributions. Steven Ross has been providing these charts/tables in the 'develop' branch of the library. 3) Connected to the previous issue is that a number of reviewers asked for better documentation on the conditions in which SpreadSort will show a marked improvement in speed compared to std::sort, as well as any condition(s) where SpreadSort performs worse than std::sort. While Steven Ross has explained that generally SpreadSort will fall back to using std::sort for some cases ( low N ), I too feel that the documentation needs to be more extensive for the average user in attempting to delineate the conditions where SpreadSort works best. 4) A really important issue discussed by a number of reviewers involves the creation of a synthetic radix, where most users of sorting algorithm are used to comparison based sorting functors. While the documentation provides information on the various individual sorts, as well as examples, I also feel that a discussion in the documentation of how to use SpreadSort when sorting structs/classes with multiple keys of different types would be invaluable. In other words, what and if are the common techniques for using SpreadSort with complicated keys which, when using comparison-based sorting, can be represented fairly easily be a functor. 5) Niall Douglas brought up the issue of exception guarantees when using SpreadSort, and I concur. The documentation should be stronger specifying what happens during the sorting algorithms if an exception occurs and/or does the sorting algorithm itself ever throw any exceptions and, if so, under what conditions. Steven Ross mentioned that he is adding exception guarantee information to his docs. 6) Many reviewers did not like the idea that the library should be called Sort when it is basically implementing a single sorting algorithm called SpreadSort. Alternate suggestions are: a) The library should be called SpreadSort. b) The library can be called Sort only if other sorting algorithms are added to it. c) The library can be called Sort with the proviso that it serves as a general library for sorting algorithms potentially added to it in the future. d) The library should be part of the Boost Algorithm library. Steven Ross mentioned other potential sorting implementations as part of the Sort library, under that general name, but all providing a good deal of extra work. These included a copy-based stable sorting algorithm, an LSD radix sorting, and an OpenMP based parallel implementation. Other reviewers mentioned some of the same types of sorting implementations and a few reviewers remarked that they also have created sorting implementations of their own. While I will give my own view at the end, I would like to mention here that I strongly believe that each sorting implementation needs to be reviewed separately for acceptance into Boost and never automatically added to a general library, whether it be called Sort or Algorithm or whatever. 7) Random issues brought up in comments I will mention, which have importance but which I do not believe are absolutely intrinsic yet still might be considered. a) C++11 techniques in general and inline namespaces in particular. b) A version that can work with less than random access. c) Optimizations for small sets of data ( Steven Ross has already addressed this to an extent in the 'developer' branch of the library ). d) Possible documentation of using SpreadSort as a subsort for parallel sorting techniques and possible discusion in the docs about SpreadSort and parallel sorts. e) General O() information in docs. f) SpreadSort examples hoisted directly into docs using Quickbook technique and updating the Spreadort docs to use Quickbook. Review Result: Based on the voting for acceptance or not, based on the discussions during the review period, and based on Steven Ross's responses, I have decided that Steven Ross's library should be accepted into Boost. My main reason for my decision, aside from the votes and comments and specific issues discussed, is twofold: 1) There is a general consensus that Steven Ross has an excellent understanding of sorting techniques and algorithms and will support and refine his library for the benefit of programmers. 2) The empirical tests show that SpreadSort offers an improvement over std::sort, significantly in the area of large data sets, and this could be a major benefit to programmers. I do not know if the review manager has a say in this but based on the remarks of the reviewers I would like to see the library as accepted be named 'SpreadSort' rather than just 'Sort'. I do think that Boost can have a library called 'Sort' but I agree with the general consensus that a 'Sort' library needs more than one type of sorting algorithm. I would like to see other people, who mentioned in the reviews/comments that they have their own sorting implementation, also submit their own implementations to Boost and, if this happens and they are accepted, I can see combining them with 'SpreadSort' into a general Boost sorting library called 'Sort' in the future.

1 0

Boost not compiling with arm-cross-compiler
by ustulation 27 Nov '14

27 Nov '14

{Ubuntu 14.04 64 bits, arm-g++ compiler 4.9 one that comes with google's ndk r10c) (The compiler has --sysroot pointing to platform android-14, and -march=armv7-a) on using arm cross-compiler for compiling boost I get the following error in "*socket_ops.ipp*": *<1> for all function (templates) of this type:* template <typename SockLenType> inline int call_getsockname(SockLenType msghdr::*, socket_type s, socket_addr_type* addr, std::size_t* addrlen) { SockLenType tmp_addrlen = (SockLenType)*addrlen; int result = ::getsockname(s, addr, &tmp_addrlen); *addrlen = (std::size_t)tmp_addrlen; return result; } /some-path/thirdparty/boost/boost_1_55_0_Android/boost/asio/detail/impl/socket_ops.ipp:1639: error: invalid conversion from 'int*' to 'socklen_t* {aka unsigned int*}' [-fpermissive] int result = ::getsockname(s, addr, &tmp_addrlen); ^ On changing the line to: int result = ::getsockname(s, addr, (socklen_t*)tmp_addrlen); it compiles fine ofcourse but i don't know if this is what is to be done. *<2> And in "mapped_region.hpp": /some-path/thirdparty/boost/boost_1_55_0_Android/boost/interprocess/mapped_region.hpp:49: error: sys/shm.h: No such file or directory # include <sys/shm.h> //System V shared memory... ^ * and points to: # if defined(BOOST_INTERPROCESS_XSI_SHARED_MEMORY_OBJECTS) # include <sys/shm.h> //System V shared memory... # endif in that file. Here are the flags in my user-config.jam file: using gcc : android : @AndroidNDKSrcDir@/bin/arm-linux-androideabi-g++ : <archiver>@AndroidNDKSrcDir@/bin/arm-linux-androideabi-ar <compileflags>-fexceptions <compileflags>-frtti <compileflags>-fPIC <compileflags>-std=c++11 <compileflags>-ffunction-sections <compileflags>-funwind-tables <compileflags>-D__ARM_ARCH_5__ <compileflags>-D__ARM_ARCH_5T__ <compileflags>-D__ARM_ARCH_5E__ <compileflags>-D__ARM_ARCH_5TE__ <compileflags>-Wno-psabi <compileflags>-march=armv5te <compileflags>-mtune=xscale <compileflags>-msoft-float <compileflags>-mthumb <compileflags>-Os <compileflags>-fomit-frame-pointer <compileflags>-fno-strict-aliasing <compileflags>-finline-limit=64 <compileflags>-I@AndroidNDKSrcDir@/include/c++/4.9 <compileflags>-Wa,--noexecstack <compileflags>-DANDROID <compileflags>-D__ANDROID__ <compileflags>-DNDEBUG <compileflags>-O2 <compileflags>-g <compileflags>-I@AndroidNDKSrcDir@/sysroot/usr/include <architecture>arm <compileflags>-fvisibility=hidden <compileflags>-fvisibility-inlines-hidden <compileflags>-fdata-sections <cxxflags>-D__arm__ <cxxflags>-D_REENTRANT <cxxflags>-D_GLIBCXX__PTHREADS ; *What is the way forward for those two errors - <1> and <2> above ?* -- View this message in context: http://boost.2283326.n4.nabble.com/Boost-not-compiling-with-arm-cross-compi… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1