October 2008 - Boost-users - lists.preview.boost.org

[multi_index] doing insert() within iteration over equal_range() result?
by Prometheus Prometheus 28 Oct '08

28 Oct '08

Hi all, i have the following problem: i use multi_index like this std::pair<Plugins::pEvents::index::type::iterator,Plugins::pEvents::index::type::iterator> range; range = this->events.get().equal_range(boost::make_tuple(event,condition)); while(range.first!=range.second){ // CALL an event handler! this->callEventHandler(range.first); // << this shows just how it works, its not exactly my code ++range.first; } within the callEventHandler and its called functions its possible that the event handler wants to register new events which results to this: this->events.insert(eventdata); The problem is OR can be, that the insert modifies the indices which modify the range.first iterators and thus a ++range.first results in an invalid "next" element this makes sense and i think most of you know this and had this problem already Questions: Does anyone have a good idea/solution how to solve this situation correctly so that the range i received will be valid? i thought about using a second multi_index which holds the newly inserted events and after the WHILE it gets moved to the original events index, but this solution only moves the problem from one index to the second which helps a bit but doesnt really solve the problem, since i could miss newly inserted events another solution i thought of is like this range = ....equal_range(...) ; while( range.first!=range.second ){ Plugins::Event &ev = (Plugins::Event&)*range.first; // get first event if(ev.alreadycalled == false){ ev.alreadycalled = true; // this element doesnt have an index and so it shouldnt invalidate the iterator/index (right?) callEventHandler(); range= ...equal_range() ; // reinitialize the range and begin to search again! } else{ ++range.first; } } but the above solution gets slower the more events are available thx all _________________________________________________________________ Explore the seven wonders of the world http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE

2 3

Problem calling bellman_ford_shortest_paths
by Tim Keitt 27 Oct '08

27 Oct '08

I'm prototyping some code that exposes an edge_list iterator interface to a database cursor. I've defined a forward_iterator using iterator_facade and can use that to create an edge_list graph. I'm running into a problem however calling bellman_ford_shortest_paths on the graph. I get boost_bellman_ford.cpp:18: error: no matching function for call to 'bellman_ford_shortest_paths(Graph&, position_t&, edge_weight_map, long long int*, double*)' The code leading up to that looks like: position_t N = getNumVertices(); edge_iterator iter; DEBUG_MESSAGE("Contructed iterator"); Graph g(iter, iter.end()); DEBUG_MESSAGE("Constructed edge list graph"); std::vector<double> distance(N, std::numeric_limits<double>::max()); distance[0] = 0.0; std::vector<position_t> parent(N); for ( position_t i = 0; i < N; ++i ) parent[i] = i; boost::bellman_ford_shortest_paths(g, N, edge_weight_map(), &parent[0], &distance[0]); The edge iterator is linked to the database cursor on construction. The problem I am guessing is in how I define the edge_weight_map. It looks like struct edge_weight_map { typedef cost_t value_type; typedef cost_t * reference; typedef edge_t key_type; typedef boost::readable_property_map_tag category; }; boost::property_traits<edge_weight_map>::value_type get(edge_weight_map map, boost::property_traits<edge_weight_map>::key_type key) { return key.cost; } It makes no sense for me to repeatedly query the database for the edge weight, so I attach it as a third field in the edge_descriptor. The get function merely returns the third component of the edge struct. I'm sort of assuming the way I am defining the edge_weight_map is the problem. Any insights? Also, I noticed the simply constructing the edge_list graph traverses the entire iterator range. Is there some copying going on? Can I override this behavior? THK -- Timothy H. Keitt http://www.keittlab.org/

1 0

Reusing IOService Object
by Tim Pynegar 27 Oct '08

27 Oct '08

Hi, I'm having an issue starting then stopping and trying to restart a socket or rather the IOService that's being used to service that socket. The IOService and Socket are member's of my class that has a run method itself that is on a separate thread. In my class constructor the Socket is passed into the IOService constructor. The sequence is as follow's Socket.Open Socket.Bind Socket.async_recieve_from Thread Start ------ IOService.Run Socket.Close IOService.Stop ( the thread terminates due to the IOService.Run method completing ) Socket.Open Socket.Bind Socket.async_recieve_from Thread Start ------ IOService.Run This then return's immediately due to a "stopped" long in IOService ( on windows ) being set to 1. So my question is really am i trying to use these in the wrong way as in should i construct a new Socket and IOService for each time i want to start them or is there some way to reset the IOService to it's original state beyond calling IOService.Stop. Thanks, Tim.

2 1

[test] testing protected/private methods out of test suits
by M. Siebeneicher 27 Oct '08

27 Oct '08

Hello boost-users, I tried to test protected/private class methods by declaring the test-case-class as a friend of the class-to-test. The test-case-class is defined by the macro BOOST_FIXTURE_TEST_CASE(T, F). That works fine as long as I encapsulate the test-case with BOOST_AUTO_TEST_SUITE(suite_name) and BOOST_AUTO_TEST_SUITE_END(). I'm not able to define the correct friend declaration, cause the compiler . Can someone give me hint, please? the code looks like this: // test-case-class.cpp namespace foo { namespace bar { BOOST_AUTO_TEST_SUITE(suite_name) struct F { class-to-test class_to_test_instance; } . . . BOOST_FIXTURE_TEST_CASE(test-case-class, F) { F.class_to_test_instance.i; . . . } . . . BOOST_AUTO_TEST_SUITE_END() }} class-to-test.cpp namespace foo { namespace bar { . . . class class-to-test { public: friend struct test-case-class; // that's wrong // No access to protected member i friend struct suite_name::test-case-class; // this is wrong, too: compiler error, can't find foo::bar::suite_name, .... protected: int i; . . . }; }} greetz, modbom -- GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen! Jetzt dabei sein: http://www.shortview.de/wasistshortview.php?mc=sv_ext_mf@gmx

7 8

asio:how to using iostream in blocking mode
by maosen.zhang 27 Oct '08

27 Oct '08

Hi all： In asio iostream client example: 27 tcp::iostream s(argv[1], "7777"); 28 std::string line; 29 std::getline(s, line); 30 std::cout << line << std::endl; Can I using iostream s in blocking mode? Which means std::getline will blocking to wait for data coming? Best wishes, maosen

1 0

Link error using thread library
by Phillip Jones 27 Oct '08

27 Oct '08

I'm trying to compile an inherited program, and a not so well documented one at that. Anyway, using CodeBlocks and MinGW on a Windows system, and I'm stuck on this error: ..\obj\Debug\src\GameVisualizer.o||In function `WinMain':| src\GameVisualizer.cpp|39|undefined reference to `boost::thread_group::thread_group()'| ..\obj\Debug\src\GameVisualizer.o||In function `GLOBAL__D_io_mutex1':| src\GameVisualizer.cpp|51|undefined reference to `boost::thread_group::create_thread(boost::function0<void, std::allocator<boost::function_base> > const&)'| src\GameVisualizer.cpp|51|undefined reference to `boost::thread_group::~thread_group()'| ..\obj\Debug\src\GameVisualizer.o||In function `ZN6PlayerC1Ev':| )]+0xd)||undefined reference to `boost::mutex::do_unlock()'| )]+0xd)||undefined reference to `boost::mutex::do_lock()'| (it keeps going) I assume this is a linking problem. Here's what I've done so far, I'm hoping an outside pair of eyes can see what I'm missing. Used bjam to compile the thread libraries. bjam --build-dir "D:\bin.v2" --with-thread --toolset=gcc --build-type=complete stage Then, added "boost_thread-mgw34-mt-d-1_36" to the link libraries. Verified that the linker can find the library. Yet, still getting these errors. Any pointers would be much appreciated.

1 0

[Boost-announce] [1.37.0 beta] Beta release available on SourceForge
by Olaf Peter 27 Oct '08

27 Oct '08

Hi, I've tested boost.program_options, where I get linker error: $ LANG=en g++ -DBOOST_ALL_NO_LIB -DDRIVER -O -I /usr/local/boost_1_37_0_beta1/include/boost-1_37/ cmdlineopts.cpp -o cmdlineopts -L /usr/local/boost_1_37_0_beta1/lib/ -l boost_program_options ... cmdlineopts.cpp:(.text+0x99): undefined reference to `boost::program_options::options_description::m_default_line_length' ... cmdlineopts.cpp:(.text+0x1ff): undefined reference to `boost::program_options::options_description::m_default_line_length' ... cmdlineopts.cpp:(.text+0x381): undefined reference to `boost::program_options::options_description::m_default_line_length' collect2: ld returned 1 exit status nevertheless compiled with -DBOOST_ALL_NO_LIB or -DBOOST_ALL_DYN_LINK Did I miss something? Regards, Olaf

2 1

Saving password encrypted
by Peter Barker 27 Oct '08

27 Oct '08

Hello, I can forsee a need for my application to store a password in a plain text file. Does Boost have a library to help me encrypt the password? Thank you, Pete

3 3

Is 1.36.0 already patched?
by Vincenzo Romano 27 Oct '08

27 Oct '08

Hello. I'd like to download and install 1.36.0 version but have also read about a "hot patch" that came later. Is the latest online version already patched? And, by the way, is there any up to date documentation in PDF format? Thanks in advance. -- Vincenzo Romano <Non quietis maribus nauta peritus>

2 2

Boost Graph Library - Using A*
by Daniel Esser 27 Oct '08

27 Oct '08

Hi! Does sombody know a good tutorial or code example to use the A* algorithm with simple arrays? I found an adapter in an pervious release of bgl but it seems to be outdated. Best regards Daniel

3 2