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

boost tokenizer problem with stringstream
by Daryl Lilburn 07 Aug '08

07 Aug '08

Hi, A colleague has recently discovered a problem using tokenizerwith the output from a std::stringstream class in VC8. If we dothe following, we get an "ITERATOR LIST CORRUPTED" exceptionthrown: typedef boost::char_separator<char> Separators;typedef boost::tokenizer<Separators > Tokens; std::string str1("ABC");std::string str2("DEF");std::stringstream oss;oss << str1 << "," << str2; Tokens strList(oss.str(), Separators(",|"));Tokens::const_iterator strIter = strList.begin(); // exception thrown here I thought this might be because the string returned from oss.str() is temporaryso I tried the following change and it still failed: std::string tempStr(oss.str());Tokens strList(tempStr, Separators(",|"));Tokens::const_iterator strIter = strList.begin(); // exception thrown here Finally I tried this and it succeeded. std::string tempStr;oss >> tempStr;Tokens strList(tempStr, Separators(",|"));Tokens::const_iterator strIter = strList.begin(); // exception thrown here Can anyone tell me why the first two (in particular the second option) failed? RegardsDaryl Lilburn _________________________________________________________________ Are you paid what you're worth? Find out: SEEK Salary Centre http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau%2…

2 1

[graph] Is there a way to stop BFS when a certain vertex was visited?
by Martin Mann 07 Aug '08

07 Aug '08

Hi, I am using BFS to find the shortest path between two nodes as suggested in the documentation for unweighted graphs (or all weights == 1). I have written a visitor and can now access a shortest path from the start vertex to the vertex of interest. ==================================================================== My question is: HOW to stop further BFS traversal of the graph? ==================================================================== Is there a possibility e.g. by emptying the vertex queue (e.g. by the visitor)? Or some other, more elegant way to do? Thanks in advance for your help, Martin

2 1

[graph] Bug - add_edge (adjacency_list) does not report if edge already exists as stated in documentation
by Martin Mann 07 Aug '08

07 Aug '08

Hi, I encountered the following problem : a successive call with equal parameters u, v, and g of ============================================================ std::pair<edge_descriptor, bool> add_edge( vertex_descriptor u , vertex_descriptor v , adjacency_list& g ) ============================================================ yields both times "RESULT.second == true". But the documentation states, that only in case it is NOT PRESENT it should be true, which is obviously wrong for the second call. If I use the "edge(u,v,g)" method to check if the edge exists I get the correct result of "edge(u,v,g).second == false" before the first call and "edge(u,v,g).second == true" after the first and before the second call. So the problem is all about the add_edge function and I think it's more a bug than a feature. Because otherwise the bool-flag is obsolete or reports something totally different. Cheers, Martin ============================================================ The boost graph class I use ============================================================ //! The properties available for the nodes of a CoreGraph typedef boost::property< boost::vertex_index_t, size_t , boost::property< boost::vertex_name_t, std::string > > CoreGraph_NodeProperties; //! The properties available for the edges of a CoreGraph typedef boost::property< boost::edge_name_t, std::string > CoreGraph_EdgeProperties; //! The definition of a the internal graph representation typedef boost::adjacency_list< boost::vecS , boost::vecS , boost::undirectedS , CoreGraph_NodeProperties , CoreGraph_EdgeProperties > CoreGraph; ============================================================

2 4

boost::function to member function - "changing the instance of the class".
by Alejandro Martinez 07 Aug '08

07 Aug '08

After i define a boost::function using bind to a member function of some signature... Can i change the instance to have as "this" in such function for another of the same type? Maybe pass it as a parameter at call time? Why would i want that? Imagine i have some class A that dispatches calls for a big amount of methods. The requests for such calls come serialized, and the method is selected by its name (string). So there should be a std::map<std::string, boost::function<some_signature> > to be able to make the actual call after i know what the name is without doing a switch or a huge ammount of ifs. So A has that map, and has it populated with member functions of itself. But i may have many instances of A. (actually, one for each of many threads). So having that map in each is a waste of resources. So maybe i could declare the map as static, so that it is shared between all instances of A. But to do that, i would need to be able to tell the function in the map im about to invoke, what its "this" should be at this time. Did i explain myself clearly? Thanks a lot for your time :)

2 2

boost::asio non-blocking read
by Jean-Sebastien Stoezel 07 Aug '08

07 Aug '08

Hi, While I've been able to implement non-blocking reads on sockets, using boost::asio::read and boost::asio::socket_base::bytes_readable (only reading whern there's data available), I'm having troubles implementing non-blocking reads on serial_port stream object. Any idea how to do? Thanks! Jean

2 1

[spirit] skip parser problem
by piotr270977＠interia.pl 07 Aug '08

07 Aug '08

Hi All I have a problem with spirit skip parser. If parsing text is ending with skiped char (ie. space) the parsing failed returning parse_info.full as false. For example, let's examine code from spirit documentation: http://spirit.sourceforge.net/distrib/spirit_1_8_5/libs/spirit/example/fund… If we parse string like that "B 6,4 ,6, 5" everything is OK, but if we parse "B 6,4 ,6, 5 " (with space at the end) parsing fail. How are you resolving this problem ??? ---------------------------------------------------------------------- Tylko dla detektywow! Konkurs na Smaker.pl Kliknij >>> http://link.interia.pl/f1eb1

2 1

[proto] latest code and docs?
by Markus Werle 07 Aug '08

07 Aug '08

Hi! I am looking for the latest and greatest working version of proto and especially the proto docs in pdf and/or html. Where can I find it and when will it be included into boost? Btw.: Is there a canonical place where accepted libraries can be found *with generated docs*? Actually I have no svn access due to a firewall ... Thanks for any extra support. Markus

4 7

[Graph] depth_first_search and specifying the edge order
by David Walthall 07 Aug '08

07 Aug '08

Hello. I have a graph that I'd like to run a depth first search on. Is it possible to control the order that edges for each node are traversed? I read through the documentation, but I didn't see anything that would do this. Thanks, David

3 4

boost on apple gcc 4.2
by Daniel Oberhoff 07 Aug '08

07 Aug '08

Hi, Has anyone successfully used boost with the "new" apple gcc 4.2? I would like to since gcc42 has openmp, but I cant seem to be able to use boost. I compiled and installed boost with apple gcc 4.2 by directly using bjam with --toolset=darwin-4.2. When I try to use it I get an error (detailed output of gcc below) which hints towards problems with inline assembly. Actually, to get it compiled at all I had to set -fasm-blocks. Am I missing a flag of sorts? Or is it "just" another woe with apple's modified assembler? Daniel error message of gcc4.2 with boost::filesystem: In file included from /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base.hpp:32, from /opt/boost-gcc4.2/include/boost-1_35/boost/detail/shared_count.hpp:29, from /opt/boost-gcc4.2/include/boost-1_35/boost/shared_ptr.hpp:28, from /opt/boost-gcc4.2/include/boost-1_35/boost/filesystem/path.hpp:19, from /opt/boost-gcc4.2/include/boost-1_35/boost/filesystem/operations.hpp:17, from /opt/boost-gcc4.2/include/boost-1_35/boost/filesystem.hpp:16, from mmm_fileloader.hpp:12, from mmm_tool_gui.hpp:5, from mmm_tool_gui.cpp:1: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘int boost::detail::atomic_exchange_and_add(int*, int)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:46: error: expected `)' before ‘:’ token /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘void boost::detail::atomic_increment(int*)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:62: error: expected `)' before ‘:’ token /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘int boost::detail::atomic_conditional_increment(int*)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:88: error: expected `)' before ‘:’ token In file included from /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base.hpp:32, from /opt/boost-gcc4.2/include/boost-1_35/boost/detail/shared_count.hpp:29, from /opt/boost-gcc4.2/include/boost-1_35/boost/shared_ptr.hpp:28, from /opt/boost-gcc4.2/include/boost-1_35/boost/filesystem/path.hpp:19, from /opt/boost-gcc4.2/include/boost-1_35/boost/filesystem/operations.hpp:17, from /opt/boost-gcc4.2/include/boost-1_35/boost/filesystem.hpp:16, from mmm_fileloader.hpp:12, from mmm_tool_gui.hpp:5, from mmm_tool_2.cpp:3: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘int boost::detail::atomic_exchange_and_add(int*, int)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:46: error: expected `)' before ‘:’ token /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘void boost::detail::atomic_increment(int*)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:62: error: expected `)' before ‘:’ token /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘int boost::detail::atomic_conditional_increment(int*)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:88: error: expected `)' before ‘:’ token In file included from /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base.hpp:32, from /opt/boost-gcc4.2/include/boost-1_35/boost/detail/shared_count.hpp:29, from /opt/boost-gcc4.2/include/boost-1_35/boost/shared_ptr.hpp:28, from /opt/boost-gcc4.2/include/boost-1_35/boost/archive/shared_ptr_helper.hpp:23, from /opt/boost-gcc4.2/include/boost-1_35/boost/archive/polymorphic_iarchive.hpp:161, from mmm_tool_net.hpp:8, from mmm_tool_net.cpp:1: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘int boost::detail::atomic_exchange_and_add(int*, int)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:46: error: expected `)' before ‘:’ token /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘void boost::detail::atomic_increment(int*)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:62: error: expected `)' before ‘:’ token /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp: In function ‘int boost::detail::atomic_conditional_increment(int*)’: /opt/boost-gcc4.2/include/boost-1_35/boost/detail/sp_counted_base_gcc_x86.hpp:88: error: expected `)' before ‘:’ token

6 6

[boost-users][StateChart] deferring events
by Igor R 06 Aug '08

06 Aug '08

Hello, I noticed that if I declare an event as deferred in a "parent" state (unintentionally), and define the reaction in some "indirect" inner state, then this reaction has no chance to be called. Is this the correct behavior or a bug? Example (pseudo-code): struct Parent : simple_state<..., Inner1> { defer Ev1; //of course, this is unnecessary; but if this an error? if I leave this line, then when we come to Inner1_3, the Ev1 is *not* processed; if I remove this line, everything is ok! }; struct Inner1 { defer Ev1; transit<Ev1to2, Inner2> }; struct Inner1 { defer Ev1; transit<Ev2to3, Inner3> }; struct Inner1 { in_state_reaction<Ev1> }; Thank you.

2 3