December 2009 - Boost-users

Boost version of Loki StaticDispatcher
by Glenn Macgregor 16 Dec '09

16 Dec '09

Hi All, I am looking to replace Loki with Boost since we use Boost for other parts of projects. Example of current situation: template<typename T> struct param_traits {}; template<> struct param_traits<long> { typedef long param_type; static param_type init() { return 0; } static bool fromString(const string &str, param_type &value, int base) { // Do conversion } static string toString(param_type &value, const string &fmt) { // Do conversion } }; template<> struct param_traits<double> { typedef double param_type; static param_type init() { return 0.0; } static bool fromString(const string &str, param_type &value, int base) { // Do conversion } static string toString(param_type &value, const string &fmt) { // Do conversion } }; Class BaseParamValue { public: virtual ~BaseParamValue() {}; virtual bool fromString(const string &str, int base = 0) = 0; virtual string toString(const string &ftm = "") = 0; static BaseParamValue *create(const string &type) { if (type == string("LONG")) { return new ParamValue<long>(); } else if (type == string("DOUBLE")) { return new ParamValue<double>(); } return 0; } }; template<typename T> class ParamValue : public BaseParamValue { public: typedef param_traits<T> traits_type; typedef typename traits_type::param_type ParamType; ParamValue() : mValue() { } ~ParamValue() {}; bool fromString(const string &str, int base) { return traits_type::fromString(str, mValue, base); } string toString(const string &fmt) { return traits_type::toString(mValue, fmt); } private: ParamType mValue; }; class Param { public: Param() : mParamValue(0) { } Param(const string &type) : mParamValue(BaseParamValue::create(type)) { } ~Param() { delete mParamValue; } private: BaseParamValue *mParamValue; }; So as you can see I store a BaseParamValue in the param class instead of making the Param class itself a template, which is not an option at the moment. I am currently using loki StaticDispatcher to up cast (I guess) to the correct templatized instance in the operations. What I am trying to get done is the following: class TestVisit { public: TestVisit(); ~TestVisit(); void visit(const ParamValue<long> &value); void visit(const ParamValue<double> &value); }; Is there a way in Boost to go from a BaseParamValue * to the actual templatized ParamValue and call the correct visit using: testVisit.visit(*mParamValue). Currently I get a compiler error saying: cannot conver parameter 1 from 'BaseParamValue' to 'const ParamValue<T> &' Sorry for the long post, thanks in advance. Glenn

2 1

BOOST_FOREACH with Index
by The Dude 15 Dec '09

15 Dec '09

Hello, Boost::foreach is very useful for iterating over a sequence and doing something that depends only on the iterated element, e.g., <code> BOOST_FOREACH(const Foo &f, foos) cout << f.bar() << endl; </code> However, I often need to iterate over a sequence and do some operation that depends on both the iterated element and the iteration index. E.g., I would like something like <code> BOOST_FOREACH(size_t i, const Foo &f, foos) cout << "The bar of element " << i << " is " << f.bar() << endl; </code> Is there an easy way to do so? Thanks & Bye, TD

7 11

[program-options] c++/clr problem with variables_map
by Freeman, Ian J. 15 Dec '09

15 Dec '09

Greetings. This is my first use of Boost, despite knowing of its existence for quite some time. I've compiled Boost 1.41 with msvc and am using basically the tutorial code for program_options. When I compile it as native c++ with the msvc compiler, it works fine, however when I wrap the same code into a c++ dll to be called via .NET C# code, the variables_map always fails to locate the program options via .count("option_name"). I've stepped through the debugger and that doesn't really help because the debugger is unable to list the contents of variables that are outside of the current stack frame (according to Microsoft), even though when I try to access the contents of the variables the stack pointer is in their respective single-threaded assemblies. I'm hoping someone here has experience integrating Boost with .net via a c++/clr dll and can help point me to what the problem might be. Somewhere in the parse_command_line() call is where I lose track of the variable data. Before the call to po::store(po::parse_command_line(argc, cstr, desc), vm); the argc, argv, and desc variables contain the correct contents. The result of the below code is that "Compression level was not set." is always output. The non-working dll code: // mtest1.cpp : Defines the exported functions for the DLL application. // #include "stdafx.h" #include <tchar.h> #define WIN32_LEAN_AND_MEAN // Exclude rarely-used stuff from Windows headers #include <windows.h> #include <iostream> #include <msclr/marshal.h> #include <boost/program_options.hpp> #using<mscorlib.dll> using namespace System; using namespace std; using namespace msclr::interop; namespace po = boost::program_options; namespace boost_TestDLL { public ref class Boost_Test { public: static void Test(int argc, array<String^>^ argv) { //convert String array to char* array marshal_context ^ context = gcnew marshal_context(); char** cstr = new char*[argc]; for(int i = 0; i < argc; i++) { cstr[i] = (char*)context->marshal_as<const char*>(argv[i]); } po::options_description desc("Allowed options"); desc.add_options() ("help", "produce help message") ("compression", po::value<int>(), "set compression level") ; po::variables_map vm; po::store(po::parse_command_line(argc, cstr, desc), vm); po::notify(vm); if (vm.count("help")) { cout << desc << "\n"; } else if (vm.count("compression")) { cout << "Compression level was set to " << vm["compression"].as<int>() << ".\n"; } else { cout << "Compression level was not set.\n"; } delete context; } }; }

1 0

[asio] Error in async_accept
by Germán Diago 15 Dec '09

15 Dec '09

Hello. I have this code (the variables that end with _ are instance variables). The code works well as is, in a synchronous way. void tcpServer::startAccept() { tcpConnection::pointer new_connection = tcpConnection::create(acceptor_.io_service()); boost::system::error_code error; acceptor_.accept(new_connection->socket()); onAccept(new_connection, boost::system::error_code()); } void tcpServer::onAccept(tcpConnection::pointer new_connection, const boost::system::error_code & error) { if (!error) { if (currentconnections_ < maxconnections_) { boost::lock_guard<boost::mutex> l(mutexcurrconnections_); /*connectionsinfo_.push_back(ServerConnectionInfoPtr(new ServerConnectionInfo( new_connection->socket().remote_endpoint().address().to_string(), new_connection->socket().remote_endpoint().port(), new_connection))); */ currentconnections_++; } else throw std::runtime_error("The server is full. No more connections accepted"); /*boost::asio::async_read(new_connection->socket(), boost::asio::buffer(bufferread_), boost::bind(&tcpServer::onRead, this, _1, _2)); */ vector<char> vecrecv(12); std::size_t readsize = boost::asio::read(new_connection->socket(), boost::asio::buffer(vecrecv)); cout << "Server received " << readsize << " bytes: " << endl; copy(vecrecv.begin(), vecrecv.end(), ostream_iterator<char>(cout)); cout << endl; std::string helloclient = "hello client"; std::vector<char> sendvec(helloclient.begin(), helloclient.end()); //TODO:Cambiar por receiveAllData() //std::size_t writesize = new_connection->socket().send(boost::asio::buffer(sendvec, sendvec.size())); std::size_t writesize = boost::asio::write(new_connection->socket(), boost::asio::buffer(sendvec)); std::cout << "Written " << writesize << " bytes" << std::endl; } startAccept(); } But if I change the function startAccept so that instead of doing what it used to do, it does an asynchronous_accept: void tcpServer::startAccept() { tcpConnection::pointer new_connection = tcpConnection::create(acceptor_.io_service()); boost::system::error_code error; acceptor_.async_accept(new_connection->socket(), boost::bind(&tcpServer::onAccept, this, boost::ref(new_connection), boost::asio::placeholders::error)); } I get the following errors: terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::system::system_error> >' what(): Socket operation on non-socket terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::system::system_error> >' what(): Connection reset by peer Anyone can help? I have no clue of what's happening. The socket is a shared_ptr, so it should be there when I use onAccept. Don't know what happens. Thanks for your time.

1 1

Re: [Boost-users] [asio] Sending large packages
by Nicola Michael Gutberlet 15 Dec '09

15 Dec '09

Some testing showed that the maximum package (load) size of boost::asio::write/read() for a TCP socket seems to be 1007 kB _____ From: Nicola Michael Gutberlet [mailto:nicola.gutberlet@hhi.fraunhofer.de] Sent: 15 December 2009 17:11 To: 'boost-users(a)lists.boost.org' Subject: [asio] Sending large packages Dear all, Is there a well-defined way of sending large packages over TCP? Or do I have to split the packages on my own? For explanation: I want to send images (size around 6MB). Best regards, Nicola ***** Nicola Michael Gutberlet (Dipl.-Inf.) Image Processing Dept. Fraunhofer-Institut für Nachrichtentechnik Heinrich-Hertz-Institut Einsteinufer 37 10587 Berlin Phone: +49 (0) 30 / 31002-564 Room: AM 601

1 0

Re. [Boost Graph] newbie - dijkstra & external weightmap
by Fergal Dalton 15 Dec '09

15 Dec '09

Thanks again, > > > > But I'm talking about edge weights - why would you put in > > "get(vertex_index,g)" and not "get(edge_index,g)" or similar (I've tried this > > simple substitution and it doesn't work; so it's obviously more complex than > > that)? > OK. I misread that -- you do want edge_index. You will need to provide > them yourself, though. You must have some mapping from edges to numbers > already to index the vector; that needs to be turned into a property map. Exactly, that's the bit I can't understand. How do I do that? It's not a stupid question, I've been banging my head against it for 2 weeks but the documentation is full of examples far more complex than I need and I lose myself in the syntax (I'm a half-newbie to C++, which is not helping). > I do not know the exact cause of the operator[] error; you > removed too much of the error message to tell. So it's probably due to using "vertex_index" to get edge weights right? In any case, let's see if it still occurs when we figure out the right property_map. If so, then I can paste the entire error (it's very long and, to me, indecipherable). Fergal.

2 1

[asio] Sending large packages
by Nicola Michael Gutberlet 15 Dec '09

15 Dec '09

Dear all, Is there a well-defined way of sending large packages over TCP? Or do I have to split the packages on my own? For explanation: I want to send images (size around 6MB). Best regards, Nicola ***** Nicola Michael Gutberlet (Dipl.-Inf.) Image Processing Dept. Fraunhofer-Institut für Nachrichtentechnik Heinrich-Hertz-Institut Einsteinufer 37 10587 Berlin Phone: +49 (0) 30 / 31002-564 Room: AM 601

1 0

Re: [Boost-users] [Boost Graph] newbie - dijkstra & external weightmap
by Fergal Dalton 15 Dec '09

15 Dec '09

> Use iterator_property_map > (http://www.boost.org/doc/libs/1_41_0/libs/property_map/doc/iterator_propert…) > using syntax such as the following (replacing > "weight_map(agent_weightmap)" in the code just above): > > weight_map( > make_iterator_property_map( > agent_weightmap.begin(), > get(vertex_index, g))) > > You will need to include <boost/property_map/property_map.hpp> to get > iterator_property_map. > > -- Jeremiah Willcock > Thanks Jeremiah, But I'm talking about edge weights - why would you put in "get(vertex_index,g)" and not "get(edge_index,g)" or similar (I've tried this simple substitution and it doesn't work; so it's obviously more complex than that)? In any case I've tried more or less exactly what you suggested as follows but the compiler spits out a long error, reported below (libboost1.40): -------- #include <boost/property_map/property_map.hpp> class road; typedef adjacency_list<setS,vecS,undirectedS,no_property,road> Graph; typedef boost::vertex_descriptor vdest; vdest shortest_path(Graph& g, vdest src, vdest dst) { vector<vdest> parent(num_vertices(g)); vector<double> agent_weightmap(num_edges(g)); dijkstra_shortest_paths(g,src, //weight_map(get(&road::transit_time,g)) // ---> this weight_map works weight_map(make_iterator_property_map(agent_weightmap.begin(), get(vertex_index, g))) .predecessor_map(make_iterator_property_map(parent.begin(), get(vertex_index, g)))); vdest vd; for (vd=dst ; parent[vd]!=src ; vd=parent[vd]); return(vd); } -------- Compiler error: ------- /usr/include/boost/property_map/property_map.hpp: In function ‘Reference boost::get(... [lots of instantiated froms] ... instantiated from [dijkstra_shortest_paths(...)] /usr/include/boost/property_map/property_map.hpp:317: error: no match for ‘operator[]’ in ‘(const boost::iterator_property_map<__gnu_cxx::__normal_iterator<double*, .... /usr/include/boost/property_map/property_map.hpp:354: note: candidates are: R boost::iterator_property_map<RandomAccessIterator, IndexMap, ....... ------- I'm getting a backward warning on hash_set included from boost/graph/adjacency_list.hpp - could this be an issue? Thanks again, Fergal.

2 1

[type_traits] data pointers and function pointers
by Edward Diener 15 Dec '09

15 Dec '09

It is difficult in some places in the type_traits documentation to tell when the documentation mentions "pointer" whether this refers to data pointers, function pointers, member data pointers, and/or member function pointers. I assume that if the documentation just refers to "pointer" that it should mean all of the above. But here goes: 1) Does the 'remove_pointer' refer to all of the above ? The examples are only for a single one of the above, a data pointer. 2) The 'is_pointer' doc does say it excludes pointer to members by which I assume this means both member data and member functions. Is there any way to tell whether a pointer is a function pointer or a data pointer by using type_traits since 'is_pointer' includes both ?

3 3

[ANN] CLI 1.1.0 released
by Boris Kolpackov 15 Dec '09

15 Dec '09

Hi, Seeing that I've got a lot of useful feedback for the initial release, I am pleased to announce the availability of CLI 1.1.0. CLI is an open-source (MIT license), cross-platform command line interface compiler for C++. It allows you to specify the options that your program supports, their types, default values, and documentation. This specification can then be automatically translated to C++ classes that implement parsing of the command line arguments and provide a convenient and type-safe interface for accessing the extracted data. The automatic usage and documentation generation is by far the biggest new feature in this release. The usage information is formatted during compilation and the documentation can be generated in the HTML and man page formats. For an example of the HTML output, see the documentation for the CLI compiler itself (you may also want to check the cli.1 man page and the usage information printed by the compiler): http://www.codesynthesis.com/projects/cli/doc/cli.xhtml Other new features in this release: * Optional generation of modifiers in addition to accessors. * Support for erasing the parsed elements from the argv array. * Support for custom argument sources via an abstract scanner interface. * Support for reading command line arguments from the argv array as well as files specified with command line options. For a more detailed list of new features in this release see: http://www.codesynthesis.com/pipermail/cli-users/2009-December/000001.html More information, documentation, and source code distributions are available from the project's web page: http://www.codesynthesis.com/projects/cli/ Enjoy, Boris -- Boris Kolpackov, Code Synthesis http://codesynthesis.com/~boris/blog Open-source XML data binding for C++ http://codesynthesis.com/products/xsd XML data binding for embedded systems http://codesynthesis.com/products/xsde Command line interface to C++ compiler http://codesynthesis.com/projects/cli

1 0