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

[Wave] Modifying the way function-like macros are interpreted
by fractured＠modularsystems.sl 04 Jan '10

04 Jan '10

I am trying to use boost::wave as a preprocessor for LSL, a language used in the virtual world Second Life. There is a data type there for collections of data, but it can only be accessed through functions such as llList2String(list,iterator); I wanted to define macros that would allow me to do something like #define x[y] llList2String(x,y), however, wave as it stands does not seem to have any immediately obvious way to do this.

4 6

[test] 64bit test case deletion errors again
by Alastair Rankine 04 Jan '10

04 Jan '10

This issue has been discussed before on this list, see the thread at gmane.comp.lib.boost.user/51628 I can confirm that the fix in 1.41 does indeed work for most regular test cases. However I believe that there is still an issue when using fixtures. Below is a sample app which crashes for me when compiled with gcc on MacOS X 10.6.2. If someone could confirm that I'm not doing anything particularly stupid, that would be appreciated. #define BOOST_TEST_MODULE autoreg test #include <boost/test/unit_test.hpp> struct Fixture { Fixture() : n(1) {} unsigned n; }; BOOST_AUTO_TEST_SUITE(AutoregTest) BOOST_FIXTURE_TEST_CASE(Test1, Fixture) { BOOST_CHECK_EQUAL(n, 1); } BOOST_FIXTURE_TEST_CASE(Test2, Fixture) { BOOST_CHECK_EQUAL(n, 1); } BOOST_AUTO_TEST_SUITE_END() % ./autoregtest Running 2 test cases... *** No errors detected autoregtest(52796) malloc: *** error for object 0x30001002006d0: pointer being freed was not allocated *** set a breakpoint in malloc_error_break to debug Breakpoint 1, 0x00007fff818f5951 in malloc_error_break () (gdb) bt #0 0x00007fff818f5951 in malloc_error_break () #1 0x00007fff8181f143 in free () #2 0x000000010006dc62 in ~framework_impl [inlined] () at /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_ports_devel_boost/work/boost_1_41_0/boost/test/impl/framework.ipp:133 #3 0x000000010006dc62 in __tcf_1 () at framework.ipp:225 #4 0x00007fff8182b334 in __cxa_finalize () #5 0x00007fff8182b24c in exit () #6 0x00000001000031e7 in start ()

1 0

asio - interrupting a long process running in server
by Lloyd 04 Jan '10

04 Jan '10

Hi, I am using ASIO to create a server. The clients sends command to the server, the server executes it in it and sends back the result to the client (here the client remotely controls the server). Lets assume that one command implementation at the server side contains only an infinite loop, as this wastes the server resource, the client need to "interrupt" or "terminate" the execution of the infinite loop in the server. Is there any mechanism available in ASIO to accomplish this? What I am doing now is, this long job is created as a thread, collect the thread's id and store it in a container. Then this id is send to the client, if the client wants to interrupt the thread, it sends another command (STOP_THREAD ) with this thread id, the server looks for the thread running with this id, and interrupts it. At the end of execution thread, the thread id has to be removed from the container (thread id,pointer to thread). But this makes it very complicated, and it is not working right. So I am looking for a solution available with ASIO... Please share your ideas... Thanks, Lloyd ___________________ Happy New Year 2010

1 0

[BGL] Free graph - double free or corruption
by Matthieu BOUSSARD 04 Jan '10

04 Jan '10

Hello, I'm using BGL to construct an exploration graph. At each decision step I have to create a new graph. It seems that I'm not freeing correctly the memory, causing a memory leak. The name of a node in the graph is a pointer to a structure, typedef adjacency_list <listS,vecS, directedS, property<vertex_name_t, state_t *, property<vertex_prop_t,s_prop> >, property<edge_weight_t, float, property<edge_name_t,int> > >Graph; I also use a map to find the vertex according to a given state typedef std::tr1::unordered_map<state_t *, vertex_t> state_vertex_map_t; In the destructor of my planner object, every time I try to free anything (from the map or from the vertices of the graph) I'v got an "double free or corruption" error. Could you help me to find the way to free all the memory ? Thank you. Best regards Matthieu

2 4

Re: [Boost-users] Defining Statements with Proto
by David A. Greene 04 Jan '10

04 Jan '10

On Sunday 03 January 2010 18:55:36 you wrote: > On Sunday 03 January 2010 12:15:59 you wrote: > > > There were a few other small problems, but on the whole, you were > > > pretty close. > > You moved the code around a bit so the diff isn't trivial. What were the > other small problems? Ah, I found one. Duplicated template parameters for the expression rules. -Dave

1 0

[test] Unexpected Exceptions
by Florian Winter 04 Jan '10

04 Jan '10

Hi, What is the intended behavior of Boost.Test if an unexpected exception is thrown? Example: void do_not_throw() throw() { throw "unexpected"; } Some of my applications using Boost.Test seem exit with an error code of 0 (successful) if an unexpected exception is thrown. Is this the intended behavior? Best regards, fw

3 5

[UTF] Accessing global fixture members within tests
by Simon Pickles 04 Jan '10

04 Jan '10

Hi I am using the Unit Test Framework to test my use of boost.python. I am using a global fixture to initialize the python intepreter a single time (rather than before each test). My problem is I would like to use the Python initialising class in my tests, and I can't figure out how to do this. I don't want to use normal fixtures since each test would call initialize the interpreter, causing problems. Here is what I want to do: ////////////////////////////////////////////////////// #include <boost//python.hpp> #include <boost//test//unit_test.hpp> #include "..//pythonManager//pythonManager.h" using namespace boost; // First make a global fixture which sets up python interpreter struct PythonManagerFixture { PythonManagerFixture() { pm.Init(); } PythonManager pm; }; // Set up suite BOOST_AUTO_TEST_SUITE(TestPythonManager) // Only init python interpreter once BOOST_GLOBAL_FIXTURE(PythonManagerFixture); BOOST_AUTO_TEST_CASE(TestInit) { // Get python globals dict python::dict globalDict = python::extract<python::dict>(pm.m_pyGlobals); // COMPILE ERROR 'pm' : undeclared identifier // Has logging module been imported? BOOST_REQUIRE(globalDict.has_key("logging")); } BOOST_FIXTURE_TEST_CASE(TestLogCallback, PythonManagerFixture) // RUNTIME ERROR pm.Init called a second time - bad { function<void(const string&, const string&)> f = pm.GetLogFunction(); // pm is in scope though.... BOOST_REQUIRE(f); } // End the suite BOOST_AUTO_TEST_SUITE_END() /////////////////////////////////////////////////////////////////// I need a way to supply the PythonManager instance, pm, to the test cases. Thanks for any advice Simon

2 3

alignment question
by Slawomir Lisznianski 03 Jan '10

03 Jan '10

I'm having trouble implementing properly aligned storage. The following example shows conceptually what I'm trying to do, but it's NOT a correct way of doing it. template <typename T> struct envelope { size_t len = sizeof(T); // <--- compiler may pad here, hence this doesn't work... T body; }; So, given an arbitrary type T, we want to create an envelope that will have the size of T immediately before it, with no padding between len and body. The user would then be able to: envelope<MyStruct> env; env.body.x = ... send(env); I'm curious if anyone has run into something similar and have a clean solution. Our code currently has to defer to temporary buffers and memcpy'ing data back to it. Thanks, Slaw

3 2

Re: [Boost-users] regex
by ninti＠internode.on.net 03 Jan '10

03 Jan '10

Thanks for your help, linking was exactly the problem. As you can no doubt tell, I'm new not only to Boost but C++ in general. If you have the patience for one more newbie question ... I have the following code (in a FastCGI++ web application). The intention is to sanitise user input by allowing only alphanumeric characters, spaces, full stops and hyphens, simply replacing everything else with underscores: std::string input_name = environment.posts["name"].value; std::string filler = "_"; boost::regex re("[^a-zA-Z0-9\s.-]"); std::string output_name = boost::regex_replace(input_name, re, filler); Compilation fails with the following error: contact.cpp: In member function ‘virtual bool Contact::response()’: contact.cpp:48: error: unknown escape sequence '\s' The regex works fine with the \s removed (but of course replaces all spaces with underscores). These alternatives compile OK but don't work (program still replaces all spaces with underscores): [^a-zA-Z0-9\\s.-] [^a-zA-Z0-9[:space:].-] I know this is going to be something pretty basic, but how can I include whitespace - non-breaking, breaking, tabs, newlines, the lot - in the regex above as symbols to NOT be replaced? Thanks again Mike

2 1

regex
by ninti＠internode.on.net 03 Jan '10

03 Jan '10

Boost 1.40 on CentOS 5.3 64bit: I have #included <boost/regex.hpp> in a file but get the following compile time error: /tmp/ccXM0HkR.o: In function `boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::assign(char const*, char const*, unsigned int)': contact.cpp:(.text._ZN5boost11basic_regexIcNS_12regex_traitsIcNS_16cpp_regex_traitsIcEEEEE6assignEPKcS7_j[boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::assign(char const*, char const*, unsigned int)]+0x27): undefined reference to `boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::do_assign(char const*, char const*, unsigned int)' collect2: ld returned 1 exit status This file definitely exists at /usr/include/boost/regex.hpp. Another Boost #include in the same file (<boost/date_time/posix_time/posix_time.hpp>) produces no error. 'ldconfig -p' lists libboost_regex.so. What am I doing wrong? Thanks Mike

2 1

regex again
by ninti＠internode.on.net 03 Jan '10

03 Jan '10

Sorry if this is double-posted, but my webmail has stripped angle brackets and their contents out of my previous email ... Boost 1.40 on CentOS 5.3 64bit: I have #included boost/regex.hpp in a file but get the following compile time error (mangled by webmail): /tmp/ccXM0HkR.o: In function `boost::basic_regex boost::cpp_regex_traits > >::assign(char const*, char const*, unsigned int)': contact.cpp:(.text._ZN5boost11basic_regexIcNS_12regex_traitsIcNS_16cpp_regex_traitsIcEEEEE6assignEPKcS7_j[boost::basic_regex boost::regex_traits > >::assign(char const*, char const*, unsigned int)]+0x27): undefined reference to `boost::basic_regex > >::do_assign(char const*, char const*, unsigned int)' collect2: ld returned 1 exit status The important bit appears to be 'undefined reference to boost__basic_regex' etc This file definitely exists at /usr/include/boost/regex.hpp. Another Boost #include in the same file (boost/date_time/posix_time/posix_time.hpp) produces no error. 'ldconfig -p' lists libboost_regex.so. What am I doing wrong? Thanks Mike

1 0

[Accumulators] Evaluation of Expression Tree
by Horst S 02 Jan '10

02 Jan '10

Hello! Given an expression tree over mathematical expressions like 'sum', 'min', 'product', etc. and a fixed number of variables X1..Xn like the following, would it be possible to use BOOST Accumulators to re-evaluate the root evaluation function based on changes in the values of the variables efficiently? sum / \ / \ min product / \ / / \ X1 X2 X3 X4 Note that variables appear only as leaf nodes. Obviously, one can flatten the tree to the following expression: RootValue = sum( min(X1,X2), product(X2,X3,X4) ) Assuming that the current values of the variables are X1=3, X2=4, X3=2, X4=2 then RootValue evaluates to 19. Now setting X2=2 results in 10. The main objective is to minimize the cost of computing the value of the root evaluation function when changing the value of a subset of variables in X1..Xn. One can assume that the subset of variables that changes each time is rather small (typically only one) and the size of the expression tree is rather large. Are BOOST Accumulators a suitable way to model such kind of expressions and the corresponding incremental evaluation? Thanks!

2 1

[boost] [program options] Recommended way to specify a list of strings
by Sanjit Jhala 01 Jan '10

01 Jan '10

Say I want to specify a list of servers via the boost program_options library. Something like: --total-servers N --server srv1 --server srv2 ... --server srvN How would I do this? It doesn't look like multitoken would work for this case, so would it be better for me to just have something like: --total-servers N --serverlist srv1 srv2 ... srvN and use a single string to store the serverlist and parse it myself? -Sanjit

4 6

Strange issue with Serialization of BiMaps
by Mangal 01 Jan '10

01 Jan '10

Hi I am using Boost 1.40.0 and stlport 5.0 on Visual Studio 2008. The Bimap deserialization succeeds for 24 elements but starts failing after that. Here is the file with all the code #include "stdafx.h" #include <boost/archive/binary_oarchive.hpp> #include <boost/archive/binary_iarchive.hpp> #include <boost/functional/hash.hpp> #include <boost/bimap/bimap.hpp> #include <fstream> #include <exception> #include <iostream> using namespace boost; using namespace boost::bimaps; typedef unsigned int uint; struct GeoNode { public: GeoNode(){} GeoNode(uint x, uint y ): x_geocode(x), y_geocode(y) {} friend bool operator==(GeoNode const& a, GeoNode const& b) { return a.x_geocode == b.x_geocode && a.y_geocode == b.y_geocode; } friend std::size_t hash_value(GeoNode const& g) { std::size_t seed = 0; boost::hash_combine(seed, g.x_geocode); boost::hash_combine(seed, g.y_geocode); return seed; } friend bool operator < (GeoNode const& a, GeoNode const& b) { if (a.x_geocode != b.x_geocode) return a.x_geocode < b.x_geocode; return a.y_geocode < b.y_geocode; } friend std::ostream& operator << (std::ostream& os, const GeoNode& g) { os<<"X: "<<g.x_geocode<<" Y: "<<g.y_geocode<<std::endl; return os; } private: friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & x_geocode; ar & y_geocode; } public: uint x_geocode; uint y_geocode; }; template< class MapType > void print_map(const MapType & m) { typedef typename MapType::const_iterator const_iterator; for( const_iterator iter = m.begin(), iend = m.end(); iter != iend; ++iter ) { std::cout << iter->first << "-->" << iter->second << std::endl; } } int _tmain(int argc, _TCHAR* argv[]) { typedef bimap<uint, GeoNode> Store_type; const char* fileName = "Nodes.dat"; Store_type storeToWrite; Store_type storeThatsRead; int countThatWorks = 24; int countThatFails = 25; //insert Data for ( int i =0;i<countThatFails;++i) { storeToWrite.insert(Store_type::value_type(i, GeoNode(i+1, i+2))); } try { //write to file { std::ofstream ofs(fileName); boost::archive::binary_oarchive oa(ofs); oa << const_cast<const Store_type&> (storeToWrite); } //read from File { std::ifstream ifs(fileName); boost::archive::binary_iarchive ia(ifs); ia >> storeThatsRead; } print_map(storeThatsRead.left); } catch (std::exception e) { std::cout<<e.what(); } return 0; } An unknown exception is thrown when more than 24 elements are added. Tried to debug but right now can't make much sense. Any help is appreciated Thanks -- View this message in context: http://old.nabble.com/Strange-issue-with-Serialization-of-BiMaps-tp26983930… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

Boost unit test framework and the "system" system call under Linux.
by Avila Mathieu 31 Dec '09

31 Dec '09

Hello, I'm a bit new to Boost, so please don't throw rocks at me if this sounds stupid. I have a simple module that allows to wrap an external command execution using "system()". It returns a result that depends on the exit code. I want to unit test this with boost test, as I do for other modules. Unfortunately, it seems that the framework catches the signal of the child exit if it exits with a non-zero exit code. So that the unit test is aborted, which is not what I want. It seems the Execution Monitor is the solution to my problem. Is it ? How can I use it to prevent Boost from catching every signal ? Thanks in advance, -- Mathieu

2 1

[boost::python] How can I link python library against certain version?
by blp330 31 Dec '09

31 Dec '09

Hi, I try to upgrade my boost from 1.36 to 1.41, but it will link against python26.lib. My own python version is 2.5, How can I tell boost to link with python25.lib? Thanks. -- View this message in context: http://old.nabble.com/-boost%3A%3Apython--How-can-I-link-python-library-aga… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

[interprocess] mutex in shared memory, crash recovery
by Jerry Hicks 31 Dec '09

31 Dec '09

I know the correct answer is to avoid crashing while holding a mutex in shared memory ;-) But since I'm handing a system off to other developers I'd like to understand if there is a most correct way to 'scrub' a mutex in shared memory clean if a thread crashed while holding the mutex. Assume that at the point I would like to do this that there is only one thread attached to the shared memory and no possibility of any thread attempting to lock the (locked stagnant) mutex.

2 1

How to use boost thread & asio library without exception
by n179911 31 Dec '09

31 Dec '09

Hi, I am using boost's thread and asio library: #include <boost/asio.hpp> #include <boost/thread.hpp> But when I compile it, I get this error: /media/data/bin/boost/include/boost/exception_ptr.hpp:11:2: error: #error This header requires exception handling to be enabled. In file included from /media/data/bin/boost/include/boost/asio/detail/winsock_init.hpp:22, /media/data/bin/boost/include/boost/function/function_base.hpp:386: error: cannot use typeid with -fno-rtti Is there a way to use boost's thread and asio library without exception and rtti? Thank you.

2 2

Need help in setting up boost library in ubuntu 9.10
by n179911 31 Dec '09

31 Dec '09

I have compiled and installed my boost library in '/media/data/bin' in ubuntu 9.10. And I have setup the INCLUDE_PATH, LIBRARY_PATH env: $ echo $INCLUDE_PATH /media/data/bin/boost/include: $ echo $LIBRARY_PATH /media/data/bin/boost/lib: But when I compile the asio example, I get the following error: $ g++ blocking_tcp_echo_server.cpp blocking_tcp_echo_server.cpp:13:26: error: boost/bind.hpp: No such file or directory blocking_tcp_echo_server.cpp:14:31: error: boost/smart_ptr.hpp: No such file or directory blocking_tcp_echo_server.cpp:15:26: error: boost/asio.hpp: No such file or directory blocking_tcp_echo_server.cpp:16:28: error: boost/thread.hpp: No such file or directory blocking_tcp_echo_server.cpp:18: error: ‘boost’ has not been declared blocking_tcp_echo_server.cpp:22: error: ‘boost’ has not been declared blocking_tcp_echo_server.cpp:22: error: expected initializer before ‘<’ token blocking_tcp_echo_server.cpp:24: error: variable or field ‘session’ declared void blocking_tcp_echo_server.cpp:24: error: ‘socket_ptr’ was not declared in this sco $ g++ -I/media/data/bin/boost/include -L/media/data/bin/boost/lib blocking_tcp_echo_server.cpp /usr/bin/ld: /tmp/ccBXjXeh.o: in function server(boost::asio::io_service&, short):blocking_tcp_echo_server.cpp(.text+0x39e): error: undefined reference to 'boost::thread::~thread()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function server(boost::asio::io_service&, short):blocking_tcp_echo_server.cpp(.text+0x3c7): error: undefined reference to 'boost::thread::~thread()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function server(boost::asio::io_service&, short):blocking_tcp_echo_server.cpp(.text+0x3d8): error: undefined reference to 'boost::thread::~thread()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function __static_initialization_and_destruction_0(int, int):blocking_tcp_echo_server.cpp(.text+0x5ea): error: undefined reference to 'boost::system::get_system_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function __static_initialization_and_destruction_0(int, int):blocking_tcp_echo_server.cpp(.text+0x5f4): error: undefined reference to 'boost::system::get_generic_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function __static_initialization_and_destruction_0(int, int):blocking_tcp_echo_server.cpp(.text+0x5fe): error: undefined reference to 'boost::system::get_generic_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function __static_initialization_and_destruction_0(int, int):blocking_tcp_echo_server.cpp(.text+0x608): error: undefined reference to 'boost::system::get_generic_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function __static_initialization_and_destruction_0(int, int):blocking_tcp_echo_server.cpp(.text+0x612): error: undefined reference to 'boost::system::get_system_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::system::error_code::error_code():blocking_tcp_echo_server.cpp(.text._ZN5boost6system10error_codeC1Ev+0x10): error: undefined reference to 'boost::system::get_system_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::asio::error::get_system_category():blocking_tcp_echo_server.cpp(.text._ZN5boost4asio5error19get_system_categoryEv+0x7): error: undefined reference to 'boost::system::get_system_category()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::asio::detail::posix_thread::~posix_thread():blocking_tcp_echo_server.cpp(.text._ZN5boost4asio6detail12posix_threadD1Ev+0x1d): error: undefined reference to 'pthread_detach' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::asio::detail::posix_thread::join():blocking_tcp_echo_server.cpp(.text._ZN5boost4asio6detail12posix_thread4joinEv+0x25): error: undefined reference to 'pthread_join' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::detail::thread_data_base::thread_data_base():blocking_tcp_echo_server.cpp(.text._ZN5boost6detail16thread_data_baseC2Ev+0x1b): error: undefined reference to 'vtable for boost::detail::thread_data_base' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::thread::thread<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > >(boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > >, boost::disable_if<boost::is_convertible<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > >&, boost::detail::thread_move_t<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > > >, boost::thread::dummy*>::type):blocking_tcp_echo_server.cpp(.text._ZN5boost6threadC1INS_3_bi6bind_tIvPFvNS_10shared_ptrINS_4asio19basic_stream_socketINS5_2ip3tcpENS5_21stream_socket_serviceIS8_EEEEEEENS2_5list1INS2_5valueISC_EEEEEEEET_NS_10disable_ifINS_14is_convertibleIRSK_NS_6detail13thread_move_tISK_EEEEPNS0_5dummyEE4typeE+0x66): error: undefined reference to 'boost::thread::start_thread()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::detail::thread_data<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > >::~thread_data():blocking_tcp_echo_server.cpp(.text._ZN5boost6detail11thread_dataINS_3_bi6bind_tIvPFvNS_10shared_ptrINS_4asio19basic_stream_socketINS5_2ip3tcpENS5_21stream_socket_serviceIS8_EEEEEEENS2_5list1INS2_5valueISC_EEEEEEED1Ev+0x28): error: undefined reference to 'boost::detail::thread_data_base::~thread_data_base()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::detail::thread_data<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > >::~thread_data():blocking_tcp_echo_server.cpp(.text._ZN5boost6detail11thread_dataINS_3_bi6bind_tIvPFvNS_10shared_ptrINS_4asio19basic_stream_socketINS5_2ip3tcpENS5_21stream_socket_serviceIS8_EEEEEEENS2_5list1INS2_5valueISC_EEEEEEED1Ev+0x42): error: undefined reference to 'boost::detail::thread_data_base::~thread_data_base()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::detail::thread_data<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > >::~thread_data():blocking_tcp_echo_server.cpp(.text._ZN5boost6detail11thread_dataINS_3_bi6bind_tIvPFvNS_10shared_ptrINS_4asio19basic_stream_socketINS5_2ip3tcpENS5_21stream_socket_serviceIS8_EEEEEEENS2_5list1INS2_5valueISC_EEEEEEED0Ev+0x28): error: undefined reference to 'boost::detail::thread_data_base::~thread_data_base()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::detail::thread_data<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > >::~thread_data():blocking_tcp_echo_server.cpp(.text._ZN5boost6detail11thread_dataINS_3_bi6bind_tIvPFvNS_10shared_ptrINS_4asio19basic_stream_socketINS5_2ip3tcpENS5_21stream_socket_serviceIS8_EEEEEEENS2_5list1INS2_5valueISC_EEEEEEED0Ev+0x42): error: undefined reference to 'boost::detail::thread_data_base::~thread_data_base()' /usr/bin/ld: /tmp/ccBXjXeh.o: in function typeinfo for boost::detail::thread_data<boost::_bi::bind_t<void, void (*)(boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >), boost::_bi::list1<boost::_bi::value<boost::shared_ptr<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > > > >:blocking_tcp_echo_server.cpp(.rodata._ZTIN5boost6detail11thread_dataINS_3_bi6bind_tIvPFvNS_10shared_ptrINS_4asio19basic_stream_socketINS5_2ip3tcpENS5_21stream_socket_serviceIS8_EEEEEEENS2_5list1INS2_5valueISC_EEEEEEEE+0x8): error: undefined reference to 'typeinfo for boost::detail::thread_data_base' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::asio::detail::posix_tss_ptr<boost::asio::detail::call_stack<boost::asio::detail::task_io_service<boost::asio::detail::epoll_reactor<false> > >::context>::~posix_tss_ptr():blocking_tcp_echo_server.cpp(.text._ZN5boost4asio6detail13posix_tss_ptrINS1_10call_stackINS1_15task_io_serviceINS1_13epoll_reactorILb0EEEEEE7contextEED2Ev+0xf): error: undefined reference to 'pthread_key_delete' /usr/bin/ld: /tmp/ccBXjXeh.o: in function boost::asio::detail::posix_tss_ptr<boost::asio::detail::call_stack<boost::asio::detail::task_io_service<boost::asio::detail::epoll_reactor<false> > >::context>::posix_tss_ptr():blocking_tcp_echo_server.cpp(.text._ZN5boost4asio6detail13posix_tss_ptrINS1_10call_stackINS1_15task_io_serviceINS1_13epoll_reactorILb0EEEEEE7contextEEC2Ev+0x22): error: undefined reference to 'pthread_key_create' collect2: ld returned 1 exit status Thank you.

2 1

Proto Help
by David Greene 31 Dec '09

31 Dec '09

I am resurrecting some ancient code written back before Proto was an official Boost library. Things have changed quite a bit since then so I need to port the code. The code was never complete in the first place but I think much of it still applies. In any case, my goal is to take a Proto expression tree and transform it into another kind of tree using custom classes. Essentially it's an IR translation problem. Let's take an add node for example. Here's what I have so far: struct Base { ... }; struct Add : Base { ... }; // Custom expression nodes. // Transform a two-operand node. template<typename NodeType, typename Dummy = boost::proto::callable> struct ConstructBinary : boost::proto::callable { typedef boost:shared_ptr<NodeType> result_type; result_type operator()(boost::shared_ptr<Base> left, boost::shared_ptr<Base> right) { return result_type(new NodeType(left, right)); } }; struct ConstructExpressionGrammar; struct AddRule : boost::proto::plus<ConstructExpressionGrammar, ConstructExpressionGrammar> {}; // Want to use switch because of the large expression space. struct ConstructExpressionGrammarCases { // The primary template matches nothing: template<typename Tag> struct case_ : proto::not_<_> {}; }; // Match and construct an Add tree. template<> struct ConstructExpressionGrammarCases::case_<proto::tag::plus> : boost::proto::when<AddRule, // ?? ConstructBinary<Add>( ConstructExpressionGrammmar( boost::proto::_left), ConstructExpressionGrammmar( boost::proto::_right))> {}; struct ConstructExpressionGrammar : boost::proto::switch_<ConstructExpressionGrammarCases> {}; That's the basic outline. I don't think this is correct, however. In the grammar rule when<>, I'm not sure I can make a nested function type like that. I'm trying to tell the grammar to match the children and transform them, passing them to the ConstructBinary transformation to build the Add subtree. Any help is greatly appreciated. Thanks! -Dave

2 2

[lambda] Boost.Bind works, labmda bind fails to compile
by Igor R 30 Dec '09

30 Dec '09

Hello, Could someone please explain the following behavior: #include <boost/lambda/bind.hpp> #include <boost/bind.hpp> struct Functor { void operator()() const {} template<class T> void func(T t) {} }; int main() { Functor f; boost::lambda::bind(&Functor::func<Functor>, &f, f); // doesn't compile boost::bind(&Functor::func<Functor>, &f, f); // compiles well } Thanks!

4 8

c++: boost::multi_array - initialize a multi_array
by Hannan Sadar 30 Dec '09

30 Dec '09

Hi all, i am trying to initialize all the elements in boost::multi_array type to zero, so far, with out success. At tried using std::fill for 1-dim boost::multi_array and it worked. Once i tried std:fill on 2-dim boost::multi_array it haven't worked any more. I saw boost have #include <boost/assign/std/vector.hpp> // for 'operator+=()' #include <boost/assert.hpp>; libraries but i don't know if this is the correct way to initialize a 2-dim multi_array. I would like to know how to initialize a boost::multi_array in an elegant way. here is an example code of my multi_array (it pass compilation, but return an error at run time): #include <iomanip> #include "boost/multi_array.hpp" #include <iostream> namespace vec { typedef boost::multi_array<unsigned int, 1> uint_1d_vec_t; typedef boost::multi_array<unsigned int, 2> uint_2d_vec_t; typedef uint_1d_vec_t::index index_1d_t; typedef uint_2d_vec_t::index index_2d_t; } int main( ) { unsigned int num_elements, num_bits, max_runs, m; num_bits = 12; max_runs = 5000; m = 2; num_elements = ( 1 << num_bits ); int kappa = 79; vec::uint_1d_vec_t foo( boost::extents[ static_cast< vec::index_1d_t >(num_elements) ] ); vec::uint_2d_vec_t boo( boost::extents[ static_cast< vec::index_2d_t >(num_elements) ][ static_cast< vec::index_2d_t >(kappa) ] ); std::fill( foo.begin() , foo.end() , 0); std::fill( boo.begin()->begin() , boo.end()->end() , 0); std::cout << "Done" << std::endl; return EXIT_SUCCESS; } any suggestion is welcome regards

2 1

Fwd: [fusion] Using boost::mpl::fold with a fusion sequence
by Roberto Giménez 30 Dec '09

30 Dec '09

Hi. When I try to compile this: #include <boost/fusion/container/map.hpp> #include <boost/mpl/fold.hpp> int main(int argc, char** argv) { typedef boost::fusion::map < boost::fusion::pair<int, const char*>, boost::fusion::pair<long, char> > FuMap; FuMap fuMap("hello", 'w'); unsigned val = boost::mpl::fold < FuMap, boost::mpl::int_<0>, boost::mpl::next<boost::mpl::_1> >::type::value; } I get the following error: "...boost/mpl/begin_end.hpp", line 35: Error, nofieldfnd: apply is not a member of boost::mpl::begin_impl<boost::fusion::fusion_sequence_tag>. >From the fusion documentation: "Fusion provides full round compatibility with MPL. Fusion sequences are fully conforming MPL sequences and MPL sequences are fully compatible with Fusion. You can work with Fusion sequences on MPL if you wish to work solely on types." When I pass a boost::mpl::map it works. Any clues?

3 3

[Fwd: boost::bind vs. boost::lambda::bind - bug]
by Chris Hite 30 Dec '09

30 Dec '09

The code below fails to compile on g++ (GCC) 3.4.3: lambda_bind_test.cxx:9: error: `_1' undeclared (first use this function) lambda_bind_test.cxx:9: error: (Each undeclared identifier is reported only once for each function it appears in.) Commenting the include of boost::bind fixes it. How can the second header hide types defined in the first header? Shall I report a bug in trac or reopen this one?: https://svn.boost.org/trac/boost/ticket/3075 Is there any way in the long run the two libraries could be unified? All the asio examples use boost::bind? Should developers starting out a new code base use just one? Which one? #include <boost/lambda/bind.hpp> #include <boost/bind.hpp> //comment this line out and it compiles int foo(const char*); void test(){ using namespace boost::lambda; bind(foo,_1); }

3 4

[serialization] Can't build first example from docs.
by Patrick Horgan 29 Dec '09

29 Dec '09

Help! And thanks:) Same error as on our code. I thought I'd go through all the documentation (BTW do you want notes on typos?) and come up to speed so I could figure out why our code (that someone else wrote), had this error. To my surprise, the first example I tried had the same no get_debug_info member error. What am I doing wrong? Using trunk Using gcc 4.4.1 or 4.5 Linux dell 2.6.31-17-generic #54-Ubuntu SMP Thu Dec 10 16:20:31 UTC 2009 i686 GNU/Linux make example1 g++ -ggdb -ansi -pedantic -Wextra -Wwrite-strings -Wformat=2 -Wall -fstrict-aliasing -O3 -Wstrict-aliasing=3 -std=c++0x example1.cpp -o example1 In file included from /usr/local/include/boost/archive/detail/abi_prefix.hpp:11, from /usr/local/include/boost/archive/shared_ptr_helper.hpp:35, from /usr/local/include/boost/archive/text_iarchive.hpp:100, from example1.cpp:5: /usr/local/include/boost/config/abi_prefix.hpp:12:3: error: #error double inclusion of header boost/config/abi_prefix.hpp is an error In file included from /usr/local/include/boost/archive/detail/abi_suffix.hpp:14, from /usr/local/include/boost/archive/text_iarchive.hpp:122, from example1.cpp:5: /usr/local/include/boost/config/abi_suffix.hpp:13:3: error: #error Header boost/config/abi_suffix.hpp must only be used after boost/config/abi_prefix.hpp In file included from /usr/local/include/boost/archive/detail/interface_oarchive.hpp:24, from /usr/local/include/boost/archive/detail/common_oarchive.hpp:20, from /usr/local/include/boost/archive/basic_text_oarchive.hpp:32, from /usr/local/include/boost/archive/text_oarchive.hpp:31, from example1.cpp:4: /usr/local/include/boost/archive/detail/oserializer.hpp: In static member function ‘static void boost::archive::detail::save_pointer_type<Archive>::polymorphic::save(Archive&, T&)’: /usr/local/include/boost/archive/detail/oserializer.hpp:441: error: ‘const class boost::archive::detail::basic_pointer_oserializer’ has no member named ‘get_debug_info’ In file included from /usr/local/include/boost/archive/detail/interface_iarchive.hpp:23, from /usr/local/include/boost/archive/detail/common_iarchive.hpp:23, from /usr/local/include/boost/archive/basic_text_iarchive.hpp:31, from /usr/local/include/boost/archive/text_iarchive.hpp:24, from example1.cpp:5: /usr/local/include/boost/archive/detail/iserializer.hpp: In member function ‘virtual void boost::archive::detail::iserializer<Archive, T>::load_object_data(boost::archive::detail::basic_iarchive&, void*, unsigned int) const’: /usr/local/include/boost/archive/detail/iserializer.hpp:176: error: there are no arguments to ‘get_debug_info’ that depend on a template parameter, so a declaration of ‘get_debug_info’ must be available /usr/local/include/boost/archive/detail/iserializer.hpp:176: note: (if you use ‘-fpermissive’, G++ will accept your code, but allowing the use of an undeclared name is deprecated) /usr/local/include/boost/archive/detail/iserializer.hpp: In member function ‘void boost::archive::detail::iserializer<Archive, T>::load_object_data(boost::archive::detail::basic_iarchive&, void*, unsigned int) const [with Archive = boost::archive::text_iarchive, T = gps_position]’: example1.cpp:62: instantiated from here /usr/local/include/boost/archive/detail/iserializer.hpp:173: error: ‘get_debug_info’ was not declared in this scope example1.cpp: At global scope: example1.cpp: In instantiation of ‘void gps_position::serialize(Archive&, unsigned int) [with Archive = boost::archive::text_iarchive]’: /usr/local/include/boost/serialization/access.hpp:118: instantiated from ‘static void boost::serialization::access::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::text_iarchive, T = gps_position]’ /usr/local/include/boost/serialization/serialization.hpp:74: instantiated from ‘void boost::serialization::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::text_iarchive, T = gps_position]’ /usr/local/include/boost/serialization/serialization.hpp:133: instantiated from ‘void boost::serialization::serialize_adl(Archive&, T&, unsigned int) [with Archive = boost::archive::text_iarchive, T = gps_position]’ /usr/local/include/boost/archive/detail/iserializer.hpp:182: instantiated from ‘void boost::archive::detail::iserializer<Archive, T>::load_object_data(boost::archive::detail::basic_iarchive&, void*, unsigned int) const [with Archive = boost::archive::text_iarchive, T = gps_position]’ example1.cpp:62: instantiated from here example1.cpp:20: warning: unused parameter ‘version’ example1.cpp: In instantiation of ‘void gps_position::serialize(Archive&, unsigned int) [with Archive = boost::archive::text_oarchive]’: /usr/local/include/boost/serialization/access.hpp:118: instantiated from ‘static void boost::serialization::access::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::text_oarchive, T = gps_position]’ /usr/local/include/boost/serialization/serialization.hpp:74: instantiated from ‘void boost::serialization::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::text_oarchive, T = gps_position]’ /usr/local/include/boost/serialization/serialization.hpp:133: instantiated from ‘void boost::serialization::serialize_adl(Archive&, T&, unsigned int) [with Archive = boost::archive::text_oarchive, T = gps_position]’ /usr/local/include/boost/archive/detail/oserializer.hpp:148: instantiated from ‘void boost::archive::detail::oserializer<Archive, T>::save_object_data(boost::archive::detail::basic_oarchive&, const void*) const [with Archive = boost::archive::text_oarchive, T = gps_position]’ example1.cpp:62: instantiated from here example1.cpp:20: warning: unused parameter ‘version’ make: *** [example1] Error 1 example.cpp is the first example in the serialization docs, reproduced here for completeness: #include <fstream> // include headers that implement a archive in simple text format #include <boost/archive/text_oarchive.hpp> #include <boost/archive/text_iarchive.hpp> ///////////////////////////////////////////////////////////// // gps coordinate // // illustrates serialization for a simple type // class gps_position { private: friend class boost::serialization::access; // When the class Archive corresponds to an output archive, the // & operator is defined similar to <<. Likewise, when the class Archive // is a type of input archive the & operator is defined similar to >>. template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & degrees; ar & minutes; ar & seconds; } int degrees; int minutes; float seconds; public: gps_position(){}; gps_position(int d, int m, float s) : degrees(d), minutes(m), seconds(s) {} }; int main() { // create and open a character archive for output std::ofstream ofs("filename"); // create class instance const gps_position g(35, 59, 24.567f); // save data to archive { boost::archive::text_oarchive oa(ofs); // write class instance to archive oa << g; // archive and stream closed when destructors are called } // ... some time later restore the class instance to its orginal state gps_position newg; { // create and open an archive for input std::ifstream ifs("filename"); boost::archive::text_iarchive ia(ifs); // read class state from archive ia >> newg; // archive and stream closed when destructors are called } return 0; }

1 1

Boost.Python and _GLIBCXX_DEBUG
by Christoph Groth 29 Dec '09

29 Dec '09

Using boost 1.40 and gcc 4.3.4 (same problem occurs with gcc 4.4.2) I have the following problem. After compiling the attached python module with the command g++ -shared -fPIC -o kernel.so python.cc box.cc -lboost_python -I/usr/include/python2.5 and executing python run.py I get the error message *** glibc detected *** python: free(): invalid next size (fast): 0x0000000001a32a30 *** This error does not occur when the macro _GLIBCXX_DEBUG is not being defined. It also does not occur when the contents of python.cc and box.cc are put into one file. In more complicated programs defining _GLIBCXX_DEBUG can lead to a "Segmentation fault". Not defining _GLIBCXX_DEBUG also helps there. Is there a way to deal with this short of not defining _GLIBCXX_DEBUG? Could it be a bug in Boost.Python? Thanks in advance, Christoph here are the files: file python.cc: **************************************************************** #include <boost/python.hpp> #define _GLIBCXX_DEBUG #include <vector> #include "box.hh" using namespace boost::python; BOOST_PYTHON_MODULE(kernel) { class_<Box>("Box", init<> ()) ; } **************************************************************** file box.cc **************************************************************** #define _GLIBCXX_DEBUG #include "box.hh" Box::Box() { } **************************************************************** file box.hh **************************************************************** #ifndef BOX_HH #define BOX_HH #include <vector> class Box { public: Box(); private: std::vector<int> blocks; }; #endif // BOX_HH **************************************************************** file run.py **************************************************************** #!/usr/bin/env python import kernel sys = kernel.Box() ****************************************************************

2 1

[property tree] Property Tree equivalent for matching nodes with specific child data?
by Jesse Perla 29 Dec '09

29 Dec '09

The property tree library looks like a very good way for me to remove the tinyxpath and tinyxml dependencies in my project. However, I am confused by one usage pattern. I have a lot of properties like the following: <top> <struct1 name="bla"> <stuff> data specific to struct1 </stuff> <struct1/> <struct1 name="bla2"> <stuff> data specific to struct1 </stuff> <struct1/> <struct2 name="bla3"> <stuff2>data specific to struct2 </stuff2> <struct1/> </top> I need to be able to get the tree given the name. In xpath I would go something like: ""/top/struct1[@name='bla']/stuff" to get at that information. I realize I could embed a <name> underneath the struct1 in xpath as well, but how do I get back to the siblings? Any ideas on how best to use property tree for this kind of pattern? Thanks, Jesse

1 0

[asio] posting a function to an io_service object
by Vilhelm Sundberg 29 Dec '09

29 Dec '09

Hi, I am trying to call a function by posting it to an io_service object running in a different thread from the main thread of the application. The problem is the function always gets called from the main thread. Basically I would like to do this with asio: #include <iostream> #include <boost/thread.hpp> using namespace std; void f() { cout << boost::this_thread::get_id() << endl; } int main(int argc, char *argv[]) { cout << "main thread: " << boost::this_thread::get_id() << endl; boost::thread t1(f); boost::thread t2(f); t1.join(); t2.join(); return 0; } The output is: main thread: 0x9739628 0x97392e8 0x9739488 Using Boost.Asio, f() always gets called from the main thread: #include <iostream> #include <boost/bind.hpp> #include <boost/thread.hpp> #include <boost/asio.hpp> using namespace std; using namespace boost::asio; void f() { cout << boost::this_thread::get_id() << endl; } int main(int argc, char *argv[]) { io_service ios1; io_service ios2; io_service::work w1(ios1); io_service::work w2(ios2); io_service::strand s1(ios1); io_service::strand s2(ios2); boost::thread t1(boost::bind(&io_service::run, &ios1)); boost::thread t2(boost::bind(&io_service::run, &ios2)); cout << "main thread: " << boost::this_thread::get_id() << endl; //s1.post(f); //s2.post(f); ios1.post(s1.wrap(f)); ios2.post(s2.wrap(f)); while (ios1.poll() > 0) ; while (ios2.poll() > 0) ; ios1.stop(); ios2.stop(); t1.join(); t2.join(); return 0; } The output is: main thread: 0x9739628 0x9739628 0x9739628 What is the correct way of posting/dispatching a function to an io_service running in a different thread, so that function is called from that thread? Thanks Vil

3 5

Thread function deletion
by Alexander Gutenev 29 Dec '09

29 Dec '09

Why does boost::thread destroys its thread function only on boost::thread object destruction and not just after thread function returns?

1 0

Compiling version 1.41 with /clr
by Dominique Aigroz 29 Dec '09

29 Dec '09

Dear all, I've upgraded from 1.39 to 1.41 and now I've got an error when compiling. error C2711: 'boost::detail::future_waiter::all_futures_lock::all_futures_lock' : this functon cannot be compiled as managed, consider using #pragma unmanaged 1> #pragma unmanaged is in effect 1> C:\Kadeo\Dev9\Thirdparty\boost_1_41_0\boost/thread/ future.hpp(411) : see source of the previous compiler diagnostic 1>C:\Kadeo\Dev9\Thirdparty\boost_1_41_0\boost/thread/future.hpp (416) : error C3821: 'boost::detail::future_waiter::all_futures_lock::all_futures_lock (std::vector<_Ty> &)': managed type or function cannot be used in an unmanaged function 1> with 1> [ 1> _Ty=boost::detail::future_waiter::registered_waiter 1> ] My dll is compiled with /clr flag, but i put #pragma unmanaged before include the boost header files. Does anyone have a solution for that problem. Regards, Dominique

2 2

[asio] Echo server demo
by Michael Fowler 29 Dec '09

29 Dec '09

Hello. I tried using the code provided here: http://www.boost.org/doc/libs/1_41_0/doc/html/boost_asio/example/echo/block… to echo text back to a client but I haven't been successful. It seems to hang at > size_t length = sock->read_some(boost::asio::buffer(data), error); when the client sends a message. What is wrong? Thanks.

1 0

Boost.Range fancify please
by anony 29 Dec '09

29 Dec '09

I have this variable definition: float vertices[][3] = { {-1, -1, 1}, {1, -1, 1}, {1, 1, 1}, {-1, 1, 1}, {-1, -1, -1}, {1, -1, -1}, {1, 1, -1}, {-1, 1, -1} }; I want to scale all elements of vertices, by some constant, for now I do it like this: BOOST_FOREACH(float& f, boost::make_iterator_range( boost::begin(vertices[0]), boost::end(vertices[7]))) { f *= scale; } But maybe something more beautiful could be done with Boost.Range?

4 11

[Iterator] Interest in bzip2 compress/decompress iterators
by Jason Felice 28 Dec '09

28 Dec '09

I've written them for a work project. It would need to be essentially rewritten for Boost (different coding standards), but I'm willing to do it if there's interest and it would likely be accepted. It would depend on bzlib (part of bzip2) being installed. I noticed there's a gzip iostream, but no gzip iterator. This is a kind of precedent for the external dependency aspect. I've also written a class which produces a tar file in a streaming fashion to an output iterator. File contents are streamed in from input iterators. I'm not sure if this is useful or where it would go.

2 2

Re: [Boost-users] asio: sys/epoll.h
by Pascal Kesseli 28 Dec '09

28 Dec '09

Ah, that could indeed do the trick! I will give it a shot immediately and report back here. Thanks a lot and best regards Pascal

1 0

Problem with boost::any_cast on SUN CC 5.10
by Alexander Vinokur 28 Dec '09

28 Dec '09

Hi, I have the following problem. Here is some program. File myFile1.cpp #include <boost/any.hpp> int main() { long myLongValue = 123; long long myLongLongValue = 123; boost::any_cast<long>(myLongValue); boost::any_cast<long long>(myLongLongValue); return 0; } Compilation on machine1 machine1> uname -a SunOS machine1 5.10 Generic_127111-11 sun4v sparc SUNW,SPARC-Enterprise-T5220 machine1> CC -V CC: Sun C++ 5.9 SunOS_sparc Patch 124863-01 2007/07/25 machine1> CC -m64 -I/local64/boost_1_34_0 myFile1.cpp machine1> // No errors Compilation on machine2 machine2> uname -a SunOS machine2 5.10 Generic_139555-08 sun4u sparc SUNW,SPARC-Enterprise machine2> CC -V CC: Sun C++ 5.10 SunOS_sparc 128228-03 2009/11/05 Usage: CC [ options ] files. Use 'CC -flags' for details machine2> CC -m64 -I/local64/boost_1_34_0 myFile1.cpp "myFile1.cpp", line 7: Error: Could not find a match for boost::any_cast<boost::ValueType>(long) needed in main(). "myFile1.cpp", line 8: Error: Could not find a match for boost::any_cast<boost::ValueType>(long long) needed in main(). 2 Error(s) detected. Any suggestions? Thanks, Alex Vinokur AMDOCS > CUSTOMER EXPERIENCE SYSTEMS INNOVATION This message and the information contained herein is proprietary and confidential and subject to the Amdocs policy statement, you may review at http://www.amdocs.com/email_disclaimer.asp

1 0

[GIL] planar view with different bit depth
by Eloi Du Bois 28 Dec '09

28 Dec '09

Hi all, I have a question. I am using gil to read exr image files (movie standard) and I would like to know can I make a planar view with n channels of different pixel bit depth... That sound weird but the exr standard allow users to put channels of different types ! For example: R = unsigned int 32 G = half float B = float A = byte Thank you, Eloi.

3 3

[BGL] A* queue size not mutable
by Arnaud Masserann 26 Dec '09

26 Dec '09

Hi, I'm using the BGL to explore an implicit graph. At the beginning, I only have one node, and no arc. And I instanciate new nodes and their corresponding arc on-the-fly in the A* visitor ( in examine_vertex ). However, the A* algorithm internally uses this queue : MutableQueue Q(num_vertices(g), icmp, index_map); which means that I'm supposed to : - know how many vertices I'm gonna need during the search, which I don't - Instanciate them all... Any hints here ? Notes : I use astar_search_no_init, and property maps based on std::map so that put() can index it in a not-known-yet location. Btw, my color map type is std::map<NodeID,boost::default_color_type > (plus the wrapper around it). I feel that using "boost::default_color_type" is kind of a hack. Is it ? Thanks Arnaud Arnaud

2 3

Re: [Boost-users] Dynamically creating new units
by Brandon Olivares 26 Dec '09

26 Dec '09

Hi, Oh, sorry, I forgot to specify that I was talking about the Boost.Units library. Brandon On 2009-12-25, Brandon Olivares wrote: > Hi, > > I posted before about having a problem with having so many units that it > takes forever to install. I was thinking about trying to perhaps > dynamically create only the ones I need, because I can't know that at > compile-time. > > Mostly it would be scaled units. For instance I might have the unit > meter, but might want to create a scaled unit, say like kilometer or > millimeter. > > Also though I wouldn't mind being able to create some of the > astronomical units when necessary, such as parsec or light year. But > in all cases, the base units already exist. > > What would be the best way to do this? > > Thanks, > Brandon > > -- > www.creatorspiritus.com > Blog and forum discussing living a Christ-centered life Follow me on > Twitter: http://twitter.com/devbanana Follow creatorspiritus.com on > Twitter: http://twitter.com/creatorspiritus -- Brandon www.creatorspiritus.com Blog and forum discussing living a Christ-centered life Follow me on Twitter: http://twitter.com/devbanana

4 13

[Serialization] Unregistered_class exception
by elizabeta petreska 25 Dec '09

25 Dec '09

Hello And I am sorry for bothering again The last couple of weeks I am trying to implement saving projects future in my program with boost serialization. However it seems that it is more hard that I think :) Now I am getting unregistered_class exception all over the place First some info about what I am trying to do : I am developing a plugin dll , which is loaded in third party main program. My plugin , say MyPlugin.dll , contains serialization code ( non intrusive save/load pair ) for classes which are defined in third party dll , say A.dll. A.dll is suppiled with my third party main program. Both A.dll and MyPlugin.dll are laoded at runtime in my main program. Only MyPlugin.dll ( and not A.dll and main program ) links with boost 1.41.0 staticly, if this is relevant at all. And I am using Visual Studio 2005. I think that I am getting this unregistered_class exception because somehow I am getting weird behavior from typeid operator ( maybe some bug), for particular class form A.dll. Note that the class have virtual methods. That is if I say : std::string s1=typeid(ON_Brep).raw_name(); ON_Brep someBrepObj; std:string s2=typeid(someTypeObj).raw_name(); s1 and s2 are not the same string. I am doing the following : /////////////////////MyPlugin.cpp //all archive classes are included here namespace boost { namespace serialization { template<class Archive> inline void save(Archive & ar,const ON_Brep& t,const unsigned int version) { boost::serialization::void_cast_register<ON_Brep, ON_Geometry>( static_cast<ON_Brep*>(NULL), static_cast<ON_Geometry *>(NULL)); //save ON_Brep here } template<class Archive> inline void load(Archive & ar,ON_Brep& t,const unsigned int version) { boost::serialization::void_cast_register<ON_Brep, ON_Geometry>( static_cast<ON_Brep*>(NULL), static_cast<ON_Geometry *>(NULL)); //load ON_Brep here } } } BOOST_SERIALIZATION_SPLIT_FREE(ON_Brep) BOOST_CLASS_EXPORT_GUID(ON_Brep, "ON_Brep") void CommandFunction() { ON_Brep* brep=GetBrepObject(); boost::archive::xml_oarchive oa(ofs); oa & BOOST_SERIALIZATION_NVP(brep); // fails to save with uregistered_class exception } The unregistered_exception is thrown before control gets to my nonintrusive save function for my ON_Brep Actually the exception is thrown in oserializer.hpp in line 376 : const boost::serialization::extended_type_info * true_type = i.get_derived_extended_type_info(t); // here throws if(NULL == true_type){ boost::serialization::throw_exception( archive_exception( archive_exception::unregistered_class, true_type->get_debug_info() ) ); } get_derived_extended_type_info is trying to get the eti record for ON_Brep, but I think that that it gets it wrong, because of the weird typeid behavior. get_derived_extended_type_info(const T & t) const { // note: this implementation - based on usage of typeid (rtti) // only does something if the class has at least one virtual function. BOOST_STATIC_WARNING(boost::is_polymorphic<T>::value); return typeid_system::extended_type_info_typeid_0::get_extended_type_info( typeid(t) ); /////// Here the typeid(t) differs from typeid(ON_Brep) <------------------1 } The thing is when I export ON_Brep with BOOST_CLASS_EXPORT_GUID(ON_Brep,"ON_Brep") extended_type_info entry is inserted in the global map, just typeid(ON_Brep) is inserted in the global map But when control gets to <-----1 the typeid(t) is different than the typeid that is entered in the global map,and just looking in the global map fails, becuase I have different behavior when I apply typeid(ON_Brep).raw_name();//this gives me ".?AVON_Brep@@" ON_Brep someBrepObj; typeid(someBrepObj).raw_name();// this gives me ".?.?AVTL_Brep@@" Anyone have seen such a thing ? Thanks for any help

3 4

help me with xpressive
by haccel 25 Dec '09

25 Dec '09

Cant compile in mingw ( QtCreator , Windows ) code with using boost xpressive. It talks " undeclared symbols - wsregex " . But with sregex ang cregex it works fine. Visual Studio compile wsregex fine , too. But GCC not. Found nex code in \boost\xpressive\xpressive_fwd.hpp - #if defined(BOOST_NO_CWCHAR) | defined(BOOST_NO_CWCTYPE) | defined(BOOST_NO_STD_WSTRING) # ifndef BOOST_XPRESSIVE_NO_WREGEX # define BOOST_XPRESSIVE_NO_WREGEX # endif #endif and later : #ifndef BOOST_XPRESSIVE_NO_WREGEX typedef basic_regex<std::wstring::const_iterator> wsregex; typedef basic_regex<wchar_t const *> wcregex; #endif What can i do? Thanx. -- View this message in context: http://old.nabble.com/help-me-with-xpressive-tp26842492p26842492.html Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

Dynamically creating new units
by Brandon Olivares 25 Dec '09

25 Dec '09

Hi, I posted before about having a problem with having so many units that it takes forever to install. I was thinking about trying to perhaps dynamically create only the ones I need, because I can't know that at compile-time. Mostly it would be scaled units. For instance I might have the unit meter, but might want to create a scaled unit, say like kilometer or millimeter. Also though I wouldn't mind being able to create some of the astronomical units when necessary, such as parsec or light year. But in all cases, the base units already exist. What would be the best way to do this? Thanks, Brandon -- www.creatorspiritus.com Blog and forum discussing living a Christ-centered life Follow me on Twitter: http://twitter.com/devbanana Follow creatorspiritus.com on Twitter: http://twitter.com/creatorspiritus

1 0

Project using Boost.Units compiling slowly
by Brandon Olivares 25 Dec '09

25 Dec '09

Hi, I finally figured out how to use Boost.Units. But now it is compiling very slowly. It compiles in about 30 minutes on my local Windows machine using Cygwin, and in about 15 minutes on my remote CentOS server. There is very little code, but I am using a lot of units. Is there any way to speed it up? I don't know exactly how many units I'm using, because I have several macros for more easily defining them, but I know that there are many. Right now it is just a test program for testing out the units, but obviously I'll be expanding it soon and I'm worried that it is already taking so long to compile. Thanks, Brandon -- www.creatorspiritus.com Blog and forum discussing living a Christ-centered life Follow me on Twitter: http://twitter.com/devbanana Follow creatorspiritus.com on Twitter: http://twitter.com/creatorspiritus

2 3

[serialization] version handling in {save|load}_construct_data
by JOAQUIN M. LOPEZ MUÑOZ 24 Dec '09

24 Dec '09

Hello, I've taken a look at how the version number is dealt with in Boost.Serialization collection serialization code: * The relevant code at saving time is: if(3 < ar.get_library_version()){ const unsigned int item_version = version< BOOST_DEDUCED_TYPENAME Container::value_type >::value; ar << BOOST_SERIALIZATION_NVP(item_version); } * At loading time, the equivalent code is: unsigned int item_version; ar >> BOOST_SERIALIZATION_NVP(count); if(3 < ar.get_library_version()) ar >> BOOST_SERIALIZATION_NVP(item_version); else item_version = 0; Can someone (aka Robert :-) ) shed some light on this bit? Why the version number has to be given such manual processing when this is not needed if, for instance, serializating values via the simple ar&value? I'm trying to do the equivalent thing in Boost.MultiIndex. Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

1 0

Binary (non-character-based) file iterators
by Jason Felice 24 Dec '09

24 Dec '09

I'm looking for binary (value_type = unsigned char) input and output iterators for files. This is as opposed to a character base iterator (which rules out ostream_iterator<>). I don't need anything fancy (in fact, I'm just copying whole files to and from archives). I know that Sprit has an input iterator which is a forward iterator, and I could use that. I've already started to write them with Boost.Iterator; but this seems like something you should expect to find somewhere in boost already.

4 5

[serialization] extended_type_info has no member named get_debug_info
by Patrick Horgan 24 Dec '09

24 Dec '09

Building my own code against trunk I started getting this error. Anyone know off the top of your head what it is before I dig it out? archive/detail/oserializer.hpp:398:36: error: ‘const class boost::serialization::extended_type_info’ has no member named ‘get_debug_info’

2 1

[Iterators] Filtering and conversion
by Ruediger Berlich 24 Dec '09

24 Dec '09

Hi there, I'm playing with the iterators library at the moment. Say I have a std::vector<boost::shared_ptr<base> > , with base being the base class of derived1 and derived2. Pointers to derived1 and derived2 are stored in arbitrary order in the vector. Would it be possible to create an iterator that filters out the derived1 objects (including implicit conversion and possibly dereferencing of the resulting shared_ptr<derivedx>) ? So what I'm looking for is something like this: /*******************************************************/ main() { vector<base> x; for (int i=0; i<10; i++) { if(i%2==0) x.push_back(shared_ptr<derived1>(new derived1(i))); else x.push_back(shared_ptr<derived2>(new derived2(i))); } conversion_iterator<derived1> it(x.end()); // ++it will skip over derived2 objects and set itself // to x.end() if no derived1 is remaining for(it=x.begin(); it!=x.end(); ++it) { // Implies implicit conversion and dereferencing // of shared_ptr<base> it->someDerived1SpecificCall(); } } /*******************************************************/ The filter_iterator already does part of what I'd want here, but cannot help with the conversion part. I'm not asking for a solution to the above problem, just whether you think this can be done at all. I had the impression that the iterators library assumes that the types of the sequence being iterated over and the type being returned are identical. Also, while it=x.begin() can be handled by a conversion constructor. it!=x.end() does not appear to be accessible with the predefined comparison operators. In any case thanks and have a nice Christmas time, Ruediger

2 1

Re: [Boost-users] mpl/fusion: metafunction to 'rangify'
by Hicham Mouline 23 Dec '09

23 Dec '09

>> typedef >> boost::fusion::result_of::as_vector<boost::fusion::result_of::transform< >> params,make_fixed_or_range>::type>::type >> params_fixed_or_range; >> >> -Christopher >It does. >Of course, the return type of fusion::transform contains the compile-time >information required. >Thanks very much, >Now I'll look at spirit (and some magic code to generate the Qi parser for >this new sequence) to fill it up with values from an input stream. I have simplified the code to http://codepad.org/NDM2Ub7z I have a couple of needs now: 1. attempt to transform pfr back to params. This should succeed if all elements of the fusion sequence are fixed values and throw an exception if any of them is a range. Should I write this as a free floating function, as a conversion operator? 2. there will be many params types, params_1 .... params_n, each of those structs will yield after a fusion::transform the sequences params_fixed_of_range_1... params_fixed_of_range_n. I need the following behavior (I would write for this a free floating template function, templated on the sequence type : one of the params_fixed_of_range_i): template <typename params_fixed_of_range> void do_work(const params_fixed_of_range& seq_i) { /// /// generate a inner for loop for all the ranges in seq_i /// for e.g., if seq_i contains /// { d1= 5.; d2= 4.0..5.0..0.1; d3=-1.5..1.5..0.6; i4=4..26..1; i5=4; i6=0; i7=0; } /// then I would want to generate the following code, or if impossible /// maybe by runtime recursion /// for (double d2=4.0; d2<=5.0; d2+=0.1) for (double d3=-1.5; d3<=1.5; d3+=0.6) for (unsigned int i4=4; i4<=26; i4+=1) { // I know d1=5. i5=4 i6=0 i7=0 // construct a params frm current d1,d2,d3,i4,i5,i6,i7 // do some work based on params // return value, this will be used to plot graphs } }; This seems very doable with boost.fusion, and extremely complicated and not flexible without it. I realize there are a lot of questions into 1 email but appreciate the help you can give me, Best regards,

1 0

Possibly Memory Leaostks in Boost.Date_Time (Boost 1.41) ?
by Daniel Vallejos 23 Dec '09

23 Dec '09

Hello! Memory Leaks in Boost.Date_Time (Boost 1.41) ? Code: #include <boost/date_time.hpp> #include <iostream> using namespace boost::posix_time; using namespace std; int main() { time_duration interval = seconds(120); cout << "interval = " << interval << endl; } Valgrind profile: $ valgrind ./date_time_test ==8666== Memcheck, a memory error detector ==8666== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. ==8666== Using Valgrind-3.5.0-Debian and LibVEX; rerun with -h for copyright info ==8666== Command: ./date_time_test ==8666== interval = 00:02:00 ==8666== ==8666== HEAP SUMMARY: ==8666== in use at exit: 1,595 bytes in 27 blocks ==8666== total heap usage: 48 allocs, 21 frees, 5,650 bytes allocated ==8666== ==8666== LEAK SUMMARY: ==8666== definitely lost: 0 bytes in 0 blocks ==8666== indirectly lost: 0 bytes in 0 blocks ==8666== possibly lost: 619 bytes in 20 blocks ==8666== still reachable: 976 bytes in 7 blocks ==8666== suppressed: 0 bytes in 0 blocks Environment: $ uname -a Linux electroforce 2.6.31-15-generic #50-Ubuntu SMP Tue Nov 10 14:53:52 UTC 2009 x86_64 GNU/Linux $ g++ --version g++ (Ubuntu 4.4.2-5ubuntu1) 4.4.2

3 2

Boost.Units - Can't figure out how to use it
by Brandon Olivares 23 Dec '09

23 Dec '09

Hi, I've been reading the documentation for Boost.Units. I understand most of what it is saying, but am still confused on how to do many things with this library. Most of the examples are using that test_system.hpp. Does this mean that you actually have to explicitly define the units you want to use? I am confused because I see the header files for the units I want to use, e.g., light_year or astronomical_unit, but I can't figure out how to actually use them. If I use boost::units::astronomical::light_second, it cannot find it. I am able to use the SI units just fine, such as: Quantity<si::length> l = 4.0 * si::meters; But beyond that, I'm lost. I also don't know which headers I should include. I don't know if there's one or two headers that include all the basics I would need, or if I have to include each of them individually. Here's a simple program I was trying to write just to test light_second, but do not know how to make it work or which headers to include: int main() { using namespace boost::units; using namespace boost::units::astronomical; using namespace std; quantity<light_second> l1 = 4.0 * light_second; cout << "l1 = " << l1 << endl; return 0; } I'm using Boost 1.41 and I'm on Windows using Cygwin. Thanks, Brandon -- www.creatorspiritus.com Blog and forum discussing living a Christ-centered life Follow me on Twitter: http://twitter.com/devbanana Follow creatorspiritus.com on Twitter: http://twitter.com/creatorspiritus

2 5

bjam? cmake?
by Roman Neuhauser 23 Dec '09

23 Dec '09

Hello, can someone explain to me what's been going on with Boost.Build? I cannot find a conclusive reference in the mailing lists, but it looks like bjam is being replaced with cmake. Is that right? If it is, how far has the effort gotten? I have a few tiny utilities written in C++ currently built and installed with GNU make, which I'd like to replace with something more portable, and I don't want to learn bjam and bbv2 if their replacement is around the corner. TIA -- Roman Neuhauser

3 4

[date_time] Delimiters are ignored
by Vadim Guchenko 23 Dec '09

23 Dec '09

Hello, All! Why the following code works while input string has wrong delimiters and too large value of seconds? #include <boost/date_time/posix_time/posix_time.hpp> #include <iostream> #include <sstream> #include <string> #include <locale> int main() { std::string str("08a02b61 15c10d2008"); std::istringstream is(str); is.imbue(std::locale(is.getloc(), new boost::posix_time::time_input_facet( "%H:%M:%S %d/%m/%Y"))); boost::posix_time::ptime ptime; is >> ptime; if (is) std::cout << ptime << "\n"; else std::cout << "Error\n"; } It outputs: 2008-Oct-15 08:03:01 -- Best regards, Vadim Guchenko [yhw at relost dot net].

1 0

iterator adaptor (complex->real)
by Neal Becker 23 Dec '09

23 Dec '09

I've been using an iterator adaptor to reinterpret between vectors of scalar and vectors of complex. Problem is, it relies on real(z) returning an lvalue. This is true in current gnu libstdc++, but looks like it's not in the current standard, and appears to be going away (real(z) will return rvalue). I need some help fixing this code in my complex_to_real iterator adaptor: template<typename BaseIterator> class const_c_to_r_adapt : public boost::iterator_adaptor< const_c_to_r_adapt<BaseIterator>, BaseIterator, typename std::iterator_traits<BaseIterator>::value_type::value_type, use_default, typename std::iterator_traits<BaseIterator>::value_type::value_type ... scalar_type & dereference() const { return (cnt % 2 == 0) ? real (const_cast<complex_type&>(*this- >base_reference())) : imag (const_cast<complex_type&>(*this- >base_reference())); } Changing it to return an rvalue is not acceptable. I need the functionality of returning an lvalue. Any ideas?

1 0

mpl/fusion: metafunction to 'rangify'
by Hicham Mouline 22 Dec '09

22 Dec '09

I am making small progress in my attempt to generate a new mpl sequence that I will use to read from a stream. The following code fails to compile: #include <boost/tuple/tuple.hpp> #include <boost/mpl/vector_c.hpp> #include <boost/mpl/transform.hpp> #include <boost/fusion/adapted/struct/adapt_struct.hpp> #include <boost/fusion/include/adapt_struct.hpp> #include <boost/fusion/mpl.hpp> // makes params a MPL sequence #include <boost/fusion/support/is_sequence.hpp> #include <boost/fusion/include/is_sequence.hpp> struct params { double d1; double d2; double d3; unsigned int i4; unsigned int i5; unsigned int i6; unsigned int i7; }; BOOST_FUSION_ADAPT_STRUCT( params, (double, d1) (double, d2) (double, d3) (unsigned int, i4) (unsigned int, i5) (unsigned int, i6) (unsigned int, i7) ) BOOST_MPL_ASSERT(( boost::fusion::traits::is_sequence<params> )); // passes template <typename T1> struct rangify { typedef boost::tuple<T1, T1, T1> type; /// min max incr }; template <typename T> struct fixed_or_range { typedef struct { union { T fixed; typename rangify<T>::type range; }; bool is_fixed; } type; }; typedef boost::mpl::transform< params, fixed_or_range<boost::mpl::_> >::type params_fixed_or_range; I understand that after the macro BOOST_FUSION_ADAPT_STRUCT , params is a fusion sequence. Including <boost/fusion/mpl.hpp> made params a MPL sequence, and therefore I can call boost::mpl::transform<> on it. The compilation error is c:\program files (x86)\boost_1_41_0\boost\mpl\clear.hpp(30) : error C2903: 'apply' : symbol is neither a class template nor a function template 1> c:\program files (x86)\boost_1_41_0\boost\mpl\transform.hpp(113) : see reference to class template instantiation 'boost::mpl::clear<Sequence>' being compiled 1> with 1> [ 1> Sequence=params 1> ] 1> c:\program files (x86)\boost_1_41_0\boost\mpl\eval_if.hpp(41) : see reference to class template instantiation 'boost::mpl::transform1<P1,P2,P3>' being compiled 1> with 1> [ 1> P1=params, 1> P2=fixed_or_range<boost::mpl::_1>, 1> P3=boost::mpl::na 1> ] 1> c:\program files (x86)\boost_1_41_0\boost\mpl\transform.hpp(138) : see reference to class template instantiation 'boost::mpl::eval_if<C,F1,F2>' being compiled 1> with 1> [ 1> C=boost::mpl::or_<boost::mpl::is_na<boost::mpl::na>,boost::mpl::is_lambda_ex pression<fixed_or_range<boost::mpl::_1>>,boost::mpl::not_<boost::mpl::is_seq uence<fixed_or_range<boost::mpl::_1>>>>, 1> F1=boost::mpl::transform1<params,fixed_or_range<boost::mpl::_1>,boost::mpl:: na>, 1> F2=boost::mpl::transform2<params,fixed_or_range<boost::mpl::_1>,boost::mpl:: na,boost::mpl::na> 1> ] 1> c:\users\hich\documents\visual studio 2008\projects\test\test1\main.cpp(51) : see reference to class template instantiation 'boost::mpl::transform<Seq1,Seq2OrOperation>' being compiled 1> with 1> [ 1> Seq1=params, 1> Seq2OrOperation=fixed_or_range<boost::mpl::_1> 1> ] Regards,

2 4

[Asio] io_service and queuing priority
by Dylan Klomparens 22 Dec '09

22 Dec '09

I have a program that uses an io_service to queue and execute function calls sequentially within its own thread. Other threads call io_service::post to queue small tasks, but every so often a large task is posted to the io_service. In my case, this large task is time-sensitive and must take priority over all other tasks. Is there a way to push that large task to the front of the io_service's queue, so I can guarantee it gets completed as soon as possible? A simpler way of asking my question: is it possible to prioritize an io_service queue? How? -- Dylan

2 1

[boost-graph]remove_vertex gives segmentation fault
by Suresh Kumar 22 Dec '09

22 Dec '09

Hi, I get segmentation fault when I call remove_vertex in boost graph. The graph is defined as adjacency_list<listS, listS, undirectedS, Cluster, Cluster> Graph; clear_vertex(leaf,g); remove_vertex(leaf,g); I call clear_vertex() and then remove_vertex(). remove_vertex gives a segmentation violation. How do I go about in debugging this? Thanks suresh -- R. Suresh Kumar PhD Student, Vislab #216, Engineering II Bldg University of California, Riverside, CA 92521

1 1

Re: [Boost-users] mpl/fusion: metafunction to 'rangify'
by Hicham Mouline 22 Dec '09

22 Dec '09

Prettified source code at http://codepad.org/knN4xmIR

1 0

BOOST_THROW_EXCEPTION old style cast
by Sam Miller 22 Dec '09

22 Dec '09

I am trying to build code using the BOOST_THROW_EXCEPTION macro with the gcc warning flags -Werror -Wold-style-cast. The macro appears to cast __LINE__ into an integer using an old-style cast instead of a static_cast. Any idea why? Sam

2 1

Compile error with boost_1_41
by Steve Ong 22 Dec '09

22 Dec '09

Hi, First of all, I don't know whether this is the right group to submit this particular issue; if not, please direct me to the right one... ...The issue: I just downloaded the source from the current Boost release (i.e., boost_1_41_0). I compiled this load into a static library and a dynamic libray (on both Solaris 10 and CYGWIN). After that I I try to compile the "basic_client" and "basic_server" (Source from the boost.asio project). Previously, these compiled just fine under boost_1_40_0, but now I running into the following compile errors: g++ -o chat_client chat_client.cpp -I/home/song/MyProjects/LibDist/boostS_1_41/include -L/home/song/MyProjects/LibDist/boostS_1_41/lib -lposix4 -lboost_thread -lboost_system -pthreads -lsocket In file included from /home/song/MyProjects/LibDist/boostS_1_41/include/boost/thread/future.hpp:12, from /home/song/MyProjects/LibDist/boostS_1_41/include/boost/thread.hpp:24, from chat_client.cpp:16: /home/song/MyProjects/LibDist/boostS_1_41/include/boost/exception_ptr.hpp:43: error: looser throw specifier for `virtual boost::exception_ptr::~exception_ptr()' /home/song/MyProjects/LibDist/boostS_1_41/include/boost/exception/detail/exception_ptr_base.hpp:27: error: overriding `virtual boost::exception_detail::exception_ptr_base::~exception_ptr_base() throw ()' make: *** [chat_client_] Error 1 ...Any advice to fix this would be appreciated. Steven

2 2

[typeof] Determining if a type has been registered or not at compile time
by Edward Diener 22 Dec '09

22 Dec '09

Is there any way one can determine whether a type has been registered or not with Boost.Typeof at compile time so that the information can be used in TMP ? I am thinking, of course, of the use of typeof in a template, where a template parameter may or may not be aregistered type when the template has been instantiated. I would like to provide my own TMP fallback code if the type has not been registered rather than see an immediate compiler error generated. Something like: boost::mpl::eval_if < boost::typeof::registered<T>, // this would be nice to have boost::mpl::identity<BOOST_TYPEOF_TPL(some_expression)>, boost::mpl::identity</* use another facility to get the type I want into a metafunction*/> >

2 2

Possibly Memory Leaks in Boost.Date_Time ?
by Daniel Vallejos 22 Dec '09

22 Dec '09

Hello! Memory Leaks in Boost.Date_Time ? Code: #include <boost/date_time.hpp> #include <iostream> using namespace boost::posix_time; using namespace std; int main() { time_duration interval = seconds(120); cout << "interval = " << interval << endl; } Valgrind profile: $ valgrind ./date_time_test ==8666== Memcheck, a memory error detector ==8666== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. ==8666== Using Valgrind-3.5.0-Debian and LibVEX; rerun with -h for copyright info ==8666== Command: ./date_time_test ==8666== interval = 00:02:00 ==8666== ==8666== HEAP SUMMARY: ==8666== in use at exit: 1,595 bytes in 27 blocks ==8666== total heap usage: 48 allocs, 21 frees, 5,650 bytes allocated ==8666== ==8666== LEAK SUMMARY: ==8666== definitely lost: 0 bytes in 0 blocks ==8666== indirectly lost: 0 bytes in 0 blocks ==8666== possibly lost: 619 bytes in 20 blocks ==8666== still reachable: 976 bytes in 7 blocks ==8666== suppressed: 0 bytes in 0 blocks Environment: $ uname -a Linux electroforce 2.6.31-15-generic #50-Ubuntu SMP Tue Nov 10 14:53:52 UTC 2009 x86_64 GNU/Linux $ g++ --version g++ (Ubuntu 4.4.2-5ubuntu1) 4.4.2

1 0

Program executable won't run using boost 1.40
by Carles Fernandez 22 Dec '09

22 Dec '09

Hi everyone, I'm new to boost and I've followed the getting started guide from boost to set it up into my computer. The first example<http://www.boost.org/doc/libs/1_40_0/more/getting_started/windows.html#buil…>compiles and executes correctly. But the second one<http://www.boost.org/doc/libs/1_40_0/more/getting_started/windows.html#link…> fails to execute (but it does compile and link correctly). When I try to execute the program from a command line (like the tutorial indicates) I get the following error: *The system cannot execute the specified program.* * * And if i try to execute it directly from Visual Studio it says: *Unable to start program ...../..../example.exe* *This application has failed to start because the application configuration is incorrect. Review the manifest file for possible errors. Reinstalling the application may fix this problem. For more details, please see the application event log.* I am using Visual Studio 2005 and I have downloaded all the VC8.0 (visual studio 2005) boost 1.40 precompiled libraries using the boostpro installer. Can someone help me please? I'm sure I am doing something wrong, but I don't know what..

2 1

boost::serialization::null_deleter
by Andreas Ronnquist 22 Dec '09

22 Dec '09

Hi! I am using boost::serialization::detail::null_deleter in combination with shared_ptr (on boost 1.34/35) - and I am having a user of my program compiling my program (on boost 1.41) who says he has to change that line to just boost::serialization::null_deleter to get it to work. Should I work around this with pre-processor checks, and in that case - what versions should I look for? Or is there a simpler fix? thanks in advance -- Andreas Ronnquist <andreas.ronnquist(a)meritkonsult.se>

5 4

Re: [Boost-users] storing data
by pete＠pcbartlett.com 22 Dec '09

22 Dec '09

On Tue 22/12/09 05:08 , "Sean Farrow" <sean.farrow(a)seanfarrow.co.uk> wrote: > Hi: > > I am in need to store data. Basically I have an application, when I > have 8 collumns that are named, however thee are 38 possible rows with > differing nameson. I also need to store the screen co-ordinates for > each column name on each row. So for example: > > Column 1, name Item, row 1, x 66, y 85, row 2 x 88 y 121. > > What is the best way of being able to store this data with boost and > get access to the item name, and co-ordinates given a row and column > number? Hi Sean, There are several boost libraries that may be of help - I am thinking of boost.array, boost.multi_array and possibly even boost.multi_index. This is in addition to standard containers such as map and vector. However it is difficult to give very specific advice as I find your mail unclear. The best I can suggest is to think through your problem again, thinking how to make it easier for others to understand - in my experience the very process of thinking how to explain a problem clearly to others may help /you/ solve the problem by yourself. Your mail would also benefit from being passed through a spelling and grammar checker prior to posting. I hope this advice is of help to you, Pete Bartlett

1 0

[graph] How to find the number of adjacent nodes
by Suresh Kumar 22 Dec '09

22 Dec '09

Hi Is there any easy method to find out the number of adjacent nodes in a graph? I am using adjacency_list with listS as parameters. Hence I i am unable to subtract the vertex iterator values returned by adjacent_vertices. Thanks suresh -- R. Suresh Kumar PhD Student, Vislab #216, Engineering II Bldg University of California, Riverside,

2 1

How to use boost::asio to implement a synchronous operation with timeouts?
by watsonsong 22 Dec '09

22 Dec '09

In the boost official documents, the example of timeouts in asio is implements by asynchronous operations and a deadline_timer. But what I want is a synchronous operation with a timeout, but I search nearly all of the reference manual and find nothing about it. In an other words, how can I used the asynchronous operations to simulate synchronous operations with timeouts. Maybe I could launch an asynchronous operation and call io_service::run_one immediately, but I don't know whether it is safe when I call io_service::run_one in a complete callback function which is call by io_service::run? Or there is any classic solution of this question. Thank you. -- View this message in context: http://old.nabble.com/How-to-use-boost%3A%3Aasio-to-implement-a-synchronous… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

storing data
by Sean Farrow 22 Dec '09

22 Dec '09

Hi: I am in need to store data. Basically I have an application, when I have 8 collumns that are named, however thee are 38 possible rows with differing nameson. I also need to store the screen co-ordinates for each column name on each row. So for example: Column 1, name Item, row 1, x 66, y 85, row 2 x 88 y 121. What is the best way of being able to store this data with boost and get access to the item name, and co-ordinates given a row and column number? Any help appreciated. Cheers Sean./

1 0

Echo server demo
by Michael Fowler 22 Dec '09

22 Dec '09

Hello. I tried using the code provided here: http://www.boost.org/doc/libs/1_41_0/doc/html/boost_asio/example/echo/block… to echo text back to a client but i haven't been successful. It seems to hang at size_t length = sock->read_some(boost::asio::buffer(data), error); what is wrong? Thanks

2 2

[Serialization] shared_ptr to const object
by elizabeta petreska 21 Dec '09

21 Dec '09

Hello I am using boost 1.38.0 How to serialize shared_ptr member which points to const object. The following compiles, but is not working i.e the restored boost_shared_ptr does not point to anything. I tried to use const_pointer_cast to cast the const away. I suspect that this is the problematic part. I think the question here is , how should I cast away the const correctly for boost::shared_ptr<const T>? #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/xml_oarchive.hpp> #include <boost/serialization/nvp.hpp> #include <boost/serialization/utility.hpp> #include <fstream> #include <boost/serialization/shared_ptr.hpp> class B { public: B() { } B(int a):ma(a) { } int ma; template<class Archive> void serialize(Archive & ar, const unsigned int version) { using boost::serialization::make_nvp; ar & make_nvp("ma",ma); } }; class BC { public: BC() { } BC(boost::shared_ptr<B>& t):mac(t) { } boost::shared_ptr<const B> mac; template<class Archive> void serialize(Archive & ar, const unsigned int version) { using boost::serialization::make_nvp; ar & make_nvp("mac", boost::const_pointer_cast<B> (mac)); } }; int main() { std::ofstream ofs(L"C:\\MyTest.xml"); std::ifstream ifs(L"C:\\MyTest.xml"); boost::shared_ptr<B> bb(new B(4)); BC aa(bb); { boost::archive::text_oarchive oa(ofs); oa & BOOST_SERIALIZATION_NVP(aa); } BC raa; { boost::archive::text_iarchive ia(ifs); ia & BOOST_SERIALIZATION_NVP(raa); } }

3 6

Bug using mpl with <ciso646> and MSVC?
by ebryan 21 Dec '09

21 Dec '09

Hello, I am using Spirit with some code that uses the keywords 'and', 'or', etc. MSVC (at least my version, 9.0) does not implement these keywords, but support can be added by including <ciso646>, which just has a bunch of macros, ie. "#define and &&". The problem I am running into is that when I try to compile something that uses Spirit and ciso646, I get this compiler error: boost\mpl\aux_\include_preprocessed.hpp(37) : fatal error C1083: Cannot open include file: 'boost/mpl/aux_/preprocessed/plain/&.hpp': No such file or directory The offending line in include_preprocessed.hpp is: # include BOOST_PP_STRINGIZE(boost/mpl/aux_/preprocessed/AUX778076_PREPROCESSED_HEADER) It looks like the macro 'bitand' is being expanded in here, so it tries to include '&.hpp'. Now, I think this problem has already been solved (mostly). See this archived discussion, where this problem was worked around for 'and' and 'or': http://lists.boost.org/Archives/boost/2004/05/65563.php I have checked the header files in boost/mpl and only see three more conflicts between header names and ciso646 macros: 'bitand', 'bitor', and 'not'. If someone with more expertise than myself could implement a similar workaround for these headers, I would be very grateful. Thanks, Erik -- View this message in context: http://old.nabble.com/Bug-using-mpl-with-%3Cciso646%3E-and-MSVC--tp26851153… Sent from the Boost - Users mailing list archive at Nabble.com.

1 1

Using Fusion adapted struct with mpl::is_sequence
by John Smith 21 Dec '09

21 Dec '09

Hi, I have adapted a struct to be a fusion sequence like so: struct s { uint8_t i; uint8_t j; uint8_t k; }; BOOST_FUSION_ADAPT_STRUCT( s, (uint8_t, i) (uint8_t, j) (uint8_t, k) ) which works fine. However, I pass the adapted struct to my MPL algorithm and it seems to always return false on mpl::is_sequence. If I typedef boost::fusion::fusion_sequence_tag tag in the struct, then it all works, also if I pass in a proper fusion sequence, then that works as well. I was wondering if there was a different way of doing things, I don't really want to require the tag to be typedef'ed in the structure. Thanks, Dave

4 5

Serialization of derived class in a plugin
by Jean-Charles Quillet 21 Dec '09

21 Dec '09

Hi, I've been using the boost serialization on many projects. But I can't manage to make it work on my current project. Basically, I've a main application and an abstract class located in a separate dll. The application loads dynamically plugins which defines classes derived from the abstract class. Then I'm trying to serialize a class created by a plugin using a pointer to the abstract class, I always get a "Unregistered class" exception. I've used the BOOST_CLASS_EXPORT macro in the plugins trying to make the boost serialization library aware of my classes without any success so far. Looking in the source directory and specially in the "test" directory, I've seen that what I'm trying to do is the same as the test described by the file test_dll_plugin.cpp, unfortunately, it seems the test is not compiled and I can't manage to make it link without error. So my main concern is: is this feature supposed to be working ? If not I'll have to review my design. If it does what is the magic to make it work (like how to compile test_dll_plugin.cpp or another simple working example) ? Thanks for any help, Jean-Charles

2 8

How can I create a ip::tcp::iostream from a SOCKET descriptor?
by Eric J. Holtman 21 Dec '09

21 Dec '09

I'm trying to convert my hand-rolled marshalling code into boost::serialization and boost::ip classes. My code gets handed a SOCKET handle (win XP). I have no control over this, the connection/server/client stuff is handled elsewhere, I just get the SOCKET after everything's already open. How can I create an iostream from that, so I can then use the Archive classes to send and receive objects over the socket? I understand the "harder" stuff (writing the overloads for my classes for input/output to a generic stream). I'm just stuck on finding that construction (or function) that will take an already opened SOCKET handle and wrap an iostream around it. any clues?

2 2

[serialization] "unregistered class" with 1.41, o.k. with 1.38
by Ruediger Berlich 21 Dec '09

21 Dec '09

Hi there, Consider a hierarchy of classes that is accessed through an abstract base class. The most derived class is a template which is "filled" with the basic types bool, char, boost::int32_t and double. I use typedefs of the kind "typedef GParameterT<char> GChar;" for ease of use. This is held in its own header file, in this case GChar.hpp . A corresponding GChar.cpp holds the export statement needed to register the class (or typedef in this case): #include <boost/serialization/export.hpp> BOOST_CLASS_EXPORT(Gem::GenEvA::GChar) There are also identical (except for base types) files for GDouble, GBoolean and GInt32 (holding the tell-tale typedefs and exports for the corresponding types). I'm using unit tests for my library, and there you will find code like: // Test (de-)serialization in different modes { // plain text format T gpt4(gpt0); BOOST_CHECK(gpt4.isEqualTo(gpt0)); BOOST_CHECK_NO_THROW(gpt4.fromString(gpt1.toString(TEXTSERIALIZATION), TEXTSERIALIZATION)); BOOST_CHECK(!gpt4.isEqualTo(gpt0)); BOOST_CHECK(gpt4.isSimilarTo(gpt1, exp(-10))); } // [ ... further serialization modes removed ... ] T can again be GChar, GBoolean, GDouble or GInt32 . to/fromString is a simple wrapper to the serialization calls (all happening through the base pointer). Now I have the problem that the toString/fromString throws an "unregistered class" with Boost 1.4.1, but **only** if GParameterT is fed a double or a boost::int32_t . (De-)Serialization works fine with char and bool as template parameter. With Boost 1.38, no errors occur at all, and also my production code works nicely with 1.41 (and it uses Boost.Serialization extensively). Any idea ? In any case: Thanks for a great library and have a nice christmas time and a happy new year! Best Regards, Ruediger

1 0

bimap find
by Bosko Markovic 21 Dec '09

21 Dec '09

Hi, I have one simple question for which I could not find answer in Bimap documentatio nor on the list. When searching bimap using find method what is the value of returned iterator if nothing matches search query? Example code: BimapDWORDVariant::right_map::const_iterator right_iter = bm.right.find(sSyncMsg); Thank you. BoskoM.

2 2

[boost::interprocess] Do I need to use lock before allocate shared memory?
by blp330 21 Dec '09

21 Dec '09

Hi, Do I need to use lock before allocate shared memory? I have two processes that will allocate string to the same shared memory, do I need to use lock? Thanks. -- View this message in context: http://old.nabble.com/-boost%3A%3Ainterprocess--Do-I-need-to-use-lock-befor… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

boost::iostreams
by Anand Patel 20 Dec '09

20 Dec '09

I have implemented iostream::filter and would like to set the stream(source/sink) error flags in the event of unexpected conditions. I did not find anyway to set the flags. Can someone please help. -Anand

1 0

asio: sys/epoll.h
by Pascal Kesseli 20 Dec '09

20 Dec '09

Hi everyone Trying to cross-compile my embedded code to a new platform, I hit the following error message: ... /home/beo/Software/C++/UC-8410/boost/include/boost-1_39/boost/asio/detail/epoll_reactor.hpp:27:23: warning: sys/epoll.h: No such file or directory ... My cross toolchain obviously does not feature such a file. Any way to circumvent or solve this issue? These are my compiler specs: g++ (GCC) 4.2.1 Copyright (C) 2007 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. It's the UC-8410 MOXA toolchain. Thanks in advance for any help and best regards Pascal _________________________________________________________________ Ski-Weltcup: Alle Rennen, alle Resultate und News auf MSN Sport http://sport.ch.msn.com/skialpin/

2 1

undefined reference to `boost::thread_resource_error::thread_resource_error()'
by Stefano 20 Dec '09

20 Dec '09

Hello, On ubuntu, I have a piece of software linking against the boost thread library. Linking with boost version 1.40 I don't have any problem, whereas switching to 1.41 I got the undefined reference to `boost::thread_resource_error::thread_resource_error()' In the sources I see that this constructor is now inlined and if I do nm -C -D libboost_thread.so | grep thread_resource_error from this (1.40) 0000ee80 T boost::thread_resource_error::thread_resource_error(int) 0000ee00 T boost::thread_resource_error::thread_resource_error() 0000ee40 T boost::thread_resource_error::thread_resource_error(int) 0000edc0 T boost::thread_resource_error::thread_resource_error() 0000f4c0 T boost::thread_resource_error::~thread_resource_error() 0000f500 T boost::thread_resource_error::~thread_resource_error() 0000f530 T boost::thread_resource_error::~thread_resource_error() 0000eec0 T boost::thread_resource_error::what() const 00013e00 V typeinfo for boost::thread_resource_error 0000ffc0 V typeinfo name for boost::thread_resource_error 00013d98 V vtable for boost::thread_resource_error I got this (1.41) 0000b6e0 W boost::thread_resource_error::~thread_resource_error() 0000b720 W boost::thread_resource_error::~thread_resource_error() 0000b480 W boost::thread_resource_error::what() const 0001282c V typeinfo for boost::thread_resource_error 0000ed00 V typeinfo name for boost::thread_resource_error 00012848 V vtable for boost::thread_resource_error The symbol is weak now. I should be not a problem, right? But I got the unresolved ref What is the best option to have my stuff compiled without the unresolved refs? Thanks Stefano

4 4

[iostreams] writing to filtering_ostream
by Igor R 20 Dec '09

20 Dec '09

Hello, I've got an application that originally used IOStreams of Boost 1.34. Now I recompiled it with 1.41, and everything stopped functioning. Meanwhile, I figured out that one of the failures occurs due to the following line: iostreams::copy(istream(&someStreambuf), someFilteringStream); The intention was just to copy a chunk of binary data from an input stream to a filtering stream, but it turns out that iostreams::copy() now *closes* the sink. I tried to re-write it like this: someFilteringStream->write(sourceData, sourceSize); or this: iostreams::write(someFilteringStream, sourceData, sourceSize); But neither of the above executes the filter. So what's the right way to do this in 1.41? Thanks.

1 1

[boost::variant] problem with swap
by Ondrej Kuncar 20 Dec '09

20 Dec '09

Hello! I have problems to compile this code with Visual Studio 9.0 SP1 and Boost 1.41: ------------------------------------------------------------------- #include <boost/variant.hpp> #include <deque> typedef boost::variant<std::deque<int>::iterator> SemanticValue; int main() { SemanticValue v1; SemanticValue v2; v1.swap(v2); } ------------------------------------------------------------------- I got this error: c:\program files\boost\boost_1_41_0\boost\variant\detail\move.hpp(155) : error C2668: 'std::swap' : ambiguous call to overloaded function c:\program files\boost\boost_1_41_0\boost\variant\detail\move.hpp(141): could be 'void boost::detail::variant::detail::move_swap::swap<T>(T &,T &)' with [ T=T0 ] c:\program files\microsoft visual studio 9.0\vc\include\utility(16): or 'void std::swap<T>(_Ty &,_Ty &)' [found using argument-dependent lookup] with [ T=T0, _Ty=T0 ] while trying to match the argument list '(T0, T0)' c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(698) : see reference to function template instantiation 'void boost::detail::variant::move_swap<T>(T &,T &)' being compiled with [ T=T0 ] c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(825) : see reference to function template instantiation 'void boost::detail::variant::swap_with<Variant>::operator ()<T>(T &) const' being compiled with [ Variant=boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>, T=T0 ] c:\program files\boost\boost_1_41_0\boost\variant\detail\visitation_impl.hpp(141) : see reference to function template instantiation 'void boost::detail::variant::invoke_visitor<Visitor>::internal_visit<T>(T &,int)' being compiled with [ Visitor=boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>>, T=T0 ] c:\program files\boost\boost_1_41_0\boost\variant\detail\visitation_impl.hpp(169) : see reference to function template instantiation 'void boost::detail::variant::visitation_impl_invoke_impl<Visitor,VoidPtrCV,T>(int,Visitor &,VoidPtrCV,T *,boost::mpl::false_)' being compiled with [ Visitor=boost::detail::variant::invoke_visitor<boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>>>, VoidPtrCV=void *, T=T0 ] c:\program files\boost\boost_1_41_0\boost\variant\detail\visitation_impl.hpp(252) : see reference to function template instantiation 'void boost::detail::variant::visitation_impl_invoke<Visitor,VoidPtrCV,T0,NoBackupFlag>(int,Visitor &,VoidPtrCV,T *,NoBackupFlag,int)' being compiled with [ Visitor=boost::detail::variant::invoke_visitor<boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>>>, VoidPtrCV=void *, NoBackupFlag=boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>::has_fallback_type_, T=T0 ] c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(1769) : see reference to function template instantiation 'void boost::detail::variant::visitation_impl<first_which,first_step,Visitor,VoidPtrCV,boost::variant<T0_>::has_fallback_type_>(const int,const int,Visitor &,VoidPtrCV,boost::mpl::false_,NoBackupFlag,Which *,step0 *)' being compiled with [ Visitor=boost::detail::variant::invoke_visitor<boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>>>, VoidPtrCV=void *, T0_=std::_Deque_iterator<int,std::allocator<int>,true>, NoBackupFlag=boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>::has_fallback_type_, Which=first_which, step0=first_step ] c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(1780) : see reference to function template instantiation 'void boost::variant<T0_>::internal_apply_visitor_impl<Visitor,void*>(int,int,Visitor &,VoidPtrCV)' being compiled with [ T0_=std::_Deque_iterator<int,std::allocator<int>,true>, Visitor=boost::detail::variant::invoke_visitor<boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>>>, VoidPtrCV=void * ] c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(1803) : see reference to function template instantiation 'void boost::variant<T0_>::internal_apply_visitor<boost::detail::variant::invoke_visitor<Visitor>>(boost::detail::variant::invoke_visitor<Visitor> &)' being compiled with [ T0_=std::_Deque_iterator<int,std::allocator<int>,true>, Visitor=boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>> ] c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(1637) : see reference to function template instantiation 'void boost::variant<T0_>::apply_visitor<boost::detail::variant::swap_with<Variant>>(Visitor &)' being compiled with [ T0_=std::_Deque_iterator<int,std::allocator<int>,true>, Variant=boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>, Visitor=boost::detail::variant::swap_with<boost::variant<std::_Deque_iterator<int,std::allocator<int>,true>>> ] c:\program files\boost\boost_1_41_0\boost\variant\variant.hpp(1631) : while compiling class template member function 'void boost::variant<T0_>::swap(boost::variant<T0_> &)' with [ T0_=std::_Deque_iterator<int,std::allocator<int>,true> ] d:\dokumenty\visual studio 2008\projects\pokus4\pokus4\pokus4.cpp(20) : see reference to class template instantiation 'boost::variant<T0_>' being compiled with [ T0_=std::_Deque_iterator<int,std::allocator<int>,true> ] Do you have any ideas? Thanks. OK

2 2

VxWorks support in boost 1.41
by Geoff Shapiro 19 Dec '09

19 Dec '09

Having recently upgraded to just released 1.41 version of boost, I noticed the new support for the VxWorks platform. I am now in the process of porting over a body of code using boost to a VxWorks/PPC platform and obviously want to take advantage of the builtin support, but I've run into a problem and would like advice on the "boost" way of rectifying the problem. I am using VxWorks 6.7 to be running on an MPC8548 PPC. When I try to compile my project files in the standard manner for incorporating boost with tr1 support (.../boost/boost/tr1/tr1 and .../boost first in the header search paths, #include <memory> as first thing in files), I am getting a compile warning. The complaint is for already defined preprocessor macros -- in boost/config/vxworks.hpp at lines 23 and 24 there are definitions: #define _POSIX_TIMERS 1 #define _POSIX_THREADS 1 A little further down, this header file includes boilerplate code in boost/config/posix_features.hpp and that header file then includes unistd.h, a system file. For VxWorks 6.7, unistd.h also has defines for the above preprocessor macros, but sets their values to 200112L. While the warning may be able to ignored and is probably benign, IF unistd.h is always included after vxworks.hpp (I am not sure this will be the case for all of our files), I would rather see this fixed. What would be the best way to fix this so as to remain conformant with the "boost" philosophy for platform independence?

2 2

metafunction to 'rangify'
by Hicham Mouline 19 Dec '09

19 Dec '09

Hi, struct inputs { double input1; int input2; bool input3; }; Given type 'inputs', how to I write a metafunction that can return to me this type: rangify<inputs>::type identical to struct inputs_rangified { union { double fixed; boost::tuple<double, double, double> range; } input1; // range is min, max, increment union { int fixed; boost::tuple <int, int, int> range; } input2; union { bool fixed; std::pair<bool, bool> range; } input3; std::bitset<3> is_range; }; 1. Is it possible to get the number of members of a struct at compile-time? 2. What would be the easiest way to fill inputs_rangified from an input stream? Typically, my application will have hundreds different 'inputs' types and I would like to automate this process. Regards,

3 3

[mpl][type_traits] Getting the minimum and maximum value for a buit-in type at compile time
by Edward Diener 19 Dec '09

19 Dec '09

I would have thought that there must be a way to get the minimum and maximum value for a C++ built-in type but my search through mpl and type_traits yields nothing and numeric_limits requires a run-time function call. Surely this information is available to the compiler so it must be available to the TMP programmer somewhere. Where is it ?

3 4

[mpl] Strange integral constant overflow warning in mpl::less_equal
by Edward Diener 19 Dec '09

19 Dec '09

This simplified version of some TMP code I am writing using mpl and integer_traits is giving me a strange warning. I'm using VC9 and Boost 1.40: ------------------------------------------------------------------------------------- struct ResultTypeIntegerValues { typedef boost::mpl::long_<0L> lzero; typedef boost::mpl::long_<boost::integer_traits<long>::const_min> lmin; typedef boost::mpl::less_equal<lmin,lzero>::type type; }; ------------------------------------------------------------------------------------- with warning message: ------------------------------------------------------------------------------------- c:\utilities\boost\boost_1_40_0\boost\mpl\aux_\integral_wrapper.hpp(73) : warning C4307: '-' : integral constant overflow 1> c:\utilities\boost\boost_1_40_0\boost\mpl\aux_\preprocessed\plain\less_equal.hpp(60) : see reference to class template instantiation 'boost::mpl::long_<N>' being compiled 1> with 1> [ 1> N=-2147483648 1> ] 1> c:\utilities\boost\boost_1_40_0\boost\mpl\aux_\preprocessed\plain\less_equal.hpp(70) : see reference to class template instantiation 'boost::mpl::less_equal_tag<T>' being compiled 1> with 1> [ 1> T=nmspace::ResultTypeIntegerValues::lmin 1> ] 1> c:\programming\programs\xxx\yyy.h(64) : see reference to class template instantiation 'boost::mpl::less_equal<N1,N2>' being compiled 1> with 1> [ 1> N1=nmspace::ResultTypeIntegerValues::lmin, 1> N2=nmspace::ResultTypeIntegerValues::lzero 1> ] ------------------------------------------------------------------------------------- As far as I can see I am doing a perfectly valid comparison. Is this a minor bug in mpl or am I doing something wrong or should this warning just be expected ? Warning C4307 gives: ------------------------------------------------------------------------------------- Error Message 'operator' : integral constant overflow The operator is used in an expression that results in an integer constant overflowing the space allocated for it. You may need to use a larger type for the constant. A signed int holds a smaller value than an unsigned int because the signed int uses one bit to represent the sign. The following sample generates C4307: // C4307.cpp // compile with: /W2 int i = 2000000000 + 2000000000; // C4307 int j = (unsigned)2000000000 + 2000000000; // OK int main() { } -------------------------------------------------------------------------------------

2 2

Re: [Boost-users] [Boost.Test] valgrind complains about invalid reads
by Scott Banachowski 18 Dec '09

18 Dec '09

--- Steven Watanabe wrote: > joel falcou wrote: > > I'm jumping in the wagon but this very piece of code > leads to a > > "incorrect free" on Snow Leopard using g++ 4.2.1 > Apple. > > Maybe it is related to the virtual matter > > We've already discussed this many times. The code above > is wrong for reasons unrelated to the use of a virtual > destructor, > although adding a virtual destructor masks the problem. > This has been fixed in the trunk. Thanks Steven, with some searching I think I found the patch you are referring to: https://svn.boost.org/trac/boost/changeset/56467 Seeing this patch, I now see how the virtual function is a red herring (although I don't know why it happens to work, in this case). Nevertheless, I still like the idea of making it virtual, just to avoid the necessity of switching on type and casting to delete, but that's my opinion. Scott

1 0

sizeof of boost::int8_t
by eial＠cs.bgu.ac.il 18 Dec '09

18 Dec '09

running sizeof on a boost::int8_t returns 1, but what is the type? int/char? thanks.

8 7

[containers] defining the insertion point of non-unique keys
by Olaf Krzikalla 18 Dec '09

18 Dec '09

Hi @boost, I'm looking for the solution to the following problem: there is a collection of objects and I want to find a particular object by two different keys. The first key is unique, so there is no problem. The second key however is non-unique and if I search for an object using this key I want to find the element that was inserted last in that sequence. Neither boost::multindex nor boost:intrusive seems to have solutions right at hand. If using boost::multindex I could try to insert a new element in the non-unique sequence using the lower_bound iterator of that sequence as a hint. However it is not documented that in that case the element is always inserted before the hint. Same is true for boost:intrusive::multi_set/multi_map. And I'm not able to swap the position of two adjacent elements in a multi_set or multi_map, if they have equal keys. To illustrate the problem, here is some code: struct A { int a; // non-unique key void* b; // unique key // stuff }; using namespace boost::multi_index; typedef multi_index_container<A, indexed_by< ordered_non_unique<member<A,int,&A::a> >, ordered_unique<member<A,void*,&A::b> > > > tCollection; void foo(tCollection& collection, int key) { typedef tCollection::nth_index<0>::type tItByInt; tItByInt i = collection.get<0>().lower_bound(key); // use the current lower bound as a hint: tItByInt insertPos = collection.get<0>.insert(i, getObject(key)); // the big question: does the assertion holds? assert(collection.get<0>().lower_bound(key) == insertPos); } Best regards Olaf Krzikalla

4 9

Re: [Boost-users] Boost Graph and GraphViz - Names and Subgraphs
by Graham Hemingway 18 Dec '09

18 Dec '09

I am building graphs with boost::graph and want to generate GraphViz output. Most everything works as expected, but I am having trouble getting two things to work correctly: 1) My subgraphs are not visualized in the GraphViz output 2) I am setting the graph_name property of both the root graph and all subgraphs, but the names never appear in the GraphViz output. Any help would be appreciated. Cheers, Graham

1 0

[phoenix] Staus of phoenix
by Edward Diener 18 Dec '09

18 Dec '09

Is phoenix an official Boost library ? I do not see it listed among the Boost libraries for 1.40 or 1.41 although I do see it in the pdf documentation for Boost libraries put together by John Maddock.

6 11

Template type questions
by Glenn Macgregor 18 Dec '09

18 Dec '09

Hi All, I am using the boost::variant with the binary visitor algorithm, apply_visitor. struct UnknownValue {}; typedef boost::variant<UnknownValue, long, unsigned long, double, string, char> ValueType; I am using apply_visitor for binary visitation as below: class AddVisitor : public boost::static_visitor<> { public: template <typename T> void operator()(const T &lhs, const T &rhs) { T val = lhs + rhs; } template <typename T, typename R> void operator()(const T &lhs, const R &rhs) { // do nothing...different types } template <typename T> void operator()(const T &lhs, const UnknownValue &rhs) { // do nothing...we have an unknown value } template <typename T> void operator()(const UnknownValue &lhs, const T &rhs) { // do nothing...we have an unknown value } void operator()(const UnknownValue &lhs, const UnknownValue &rhs) { // do nothing...we have an unknown value } }; The above is not complete because I can do addition if the types are long and unsigned long or long and double or long and char...etc. What is the preferred method for doing this? Do I need to create an overridden operator() for every argument combination that is valid or can I get the argument types in the function? Thanks Glenn

2 1

[spirit] help parsing data file into a struct
by Kenny Riddile 18 Dec '09

18 Dec '09

This is my first time using Spirit and I haven't messed with writing EBNF grammars since college, so I'm sorry if this is a simple question. Lets say I have a structure that looks like this: struct Foo { std::string someString; std::string someOtherString; std::vector<std::string> stringList; }; BOOST_FUSION_ADAPT_STRUCT( Foo, (std::string, someString) (std::string, someOtherString) (std::vector<std::string>, stringList) ) and I have a data file that looks like this: SOMESTRING hello SOMEOTHERSTRING goodbye STRINGLIST { string1 string2 string3 } I have written the following grammar to parse such a file into the struct above: template< typename Iterator > struct FooDefinition : spirit::qi::grammar< Iterator, Foo() > { FooDefinition() : FooDefinition::base_type( start ) { using namespace spirit::qi; start = someString >> +space >> someOtherString >> +space >> stringList; someString = lit("SOMESTRING") >> +space >> value; someOtherString = lit("SOMEOTHERSTRING") >> +space >> value; stringList = lit("STRINGLIST") >> *space >> lit('{') >> *space >> (value % +space) >> *space >> lit('}'); value = +char_("a-zA-Z_0-9"); space = lit(' ') | lit('\n') | lit('\t'); } spirit::qi::rule< Iterator, Foo() > start; spirit::qi::rule< Iterator, std::string() > someString; spirit::qi::rule< Iterator, std::string() > someOtherString; spirit::qi::rule< Iterator, std::vector<std::string>() > stringList; spirit::qi::rule< Iterator, std::string() > value; spirit::qi::rule< Iterator > space; }; However, I would like the order of the three statements in the data file (SOMESTRING, SOMEOTHERSTRING, and STRINGLIST) to be irrelevant. Right now, they must be in the order shown above for parsing to succeed. How would I go about accomplishing this with Spirit?

2 2

inheritance, input streams, parsing and arithmetic range of values
by Hicham Mouline 18 Dec '09

18 Dec '09

Hello I would like your opinions about design/implementation choices for an application I am writing. I have a list of models. Each model is a kind of black box that takes inputs and produces outputs. All the inputs are either of boolean or arithmetic type. (bool, int, double) A subset of inputs is valid for all models. Then different models have different number of other inputs. For a fixed set of inputs, the model produces 1 fixed set of outputs. It is desired to vary the values of some/all (at least 1) of the inputs and get different outputs sets, and then plot that. The inputs are specified in files (considered as std::istream in the code), the syntax is trivial: input1 = value1 ... inputn = valuen When it is desired to vary the input values, the syntax changes a little: input1 = value1 input2 = value21..value22 ... inputn = valuen in this example only input2 changes, and the idea is to plot one of the outputs over the differnt values of input2. Each derived model has a concrete struct input_i_s { ... } that holds its inputs. class model { public: virtual fixed_inputs get_fixed_inputs( std::istream& ) const =0; /// throws if syntax error, or not all inputs are specified, or some values are not valid virtual variable_inputs get_variable_inputs( std::istream& ) const =0; /// the same, allows the '..' notation though for ranges, at least of the inputs need to be a range }; // fixed_inputs is a pointer to a base class of all inputs for all models 1 implementation pseud-code looks like: fixed_inputs model_i::get_fixed_inputs( std::istream& input) const { size_t inputs = 7; std::string buffer; while (inputs>0 && input.good()) { input >> buffer; if ( buffer.substr(0, sizeof("input1")) == "input1=") { std::istringstream is(buffer.substr(0, sizeof("input1"))); is >> .... ; /// pseudo-code : stores the value into a field in fixed_input1s // if wrong value, throws an exception } else if ... ... } return fixed_inputs( new input_i_s() ); // pseudo-code, say for model number i } I would like then to write the implementation for the syntax which accepts ranges as well. (the get_variable_inputs function) I hope to be able to factorize code between the 2 implementations. 1. boost::program_options can read options from input streams. Given that I am interested in ranges as well, is it adviced to go down that road? 2. Is there a library that helps to generate the expected format of the input stream from the struct of inputs for each model, and reads the stream automatically. Can it do range values as well? 3. is it possible then to factorize code between the get_fixed_inputs and get_variable_inputs? Regards,

1 0

(no subject)
by Michael Fowler 18 Dec '09

18 Dec '09

Hello. I tried using the code provided here: http://www.boost.org/doc/libs/1_41_0/doc/html/boost_asio/example/echo/block… to echo text back to a client but i haven't been successful. It seems to hang at size_t length = sock->read_some(boost::asio::buffer(data), error); what is wrong? Thanks

1 0

Function template specialization
by Surya Kiran Gullapalli 18 Dec '09

18 Dec '09

Hello all, The following question may not be related to boost, but if a solution can be found with boost or otherwise would be great. I've a base class with a function template. I derive from base class and try to have a specialization for the function template in derived class I did something like this. ////////////////////////// Header file begin ////////////////////////// class Base { .. template <typename T> fun (T arg) { ... } }; class Derived : public Base { ... } ; template <> Derived::fun(int arg); ///////////////////////// Header file end //////////////////////////// and in .cpp file I've provided implementation for the template specialization. This works fine with MSVC 8.0 and g++ complains about lack of function declaration fun in Derived class. I do not know which compiler is behaving correctly. Any help in this is greatly appreciated. Thanks in advance, Surya

2 1

boost lambda and bind
by Conoscenza Silente 18 Dec '09

18 Dec '09

Hi All I am trying to use boost lambda for looping over a std::map to get each element and pass those to a class function using lambda functions. I don't know why but the binding inside my std::for_each is not compiling and I have 2 errors pointing to the line where boost::assign is used. 'invalid initialization of reference of type int&' 'invalid initialization of reference of type std::basic_string<>&' Do you know how to fix this and why it is happening Thank you very much Regards CS //Boost //#include<boost/assign/std/map.hpp> #include<boost/assign/list_of.hpp> #include<boost/lambda/lambda.hpp> #include<boost/lambda/core.hpp> #include<boost/lambda/bind.hpp> #include<boost/lambda/construct.hpp> class CInnerClassHandler{ public: void DoSomething( int aIntValue, const std::string& aStringValue ){ std::cout << "INT:" << aIntValue << "STRING:" << aStringValue << std::endl; // here we just trace } }; void map_lambda_test(){ typedef std::pair< int, std::string > this_pair; std::map< int, std::string > dictionary = boost::assign::map_list_of(3,"A")(4,"B")(5,"C"); // Errors points to // this line of code CInnerClassHandler obj; std::for_each( dictionary.begin(), dictionary.end(), boost::lambda::bind( &CInnerClassHandler::DoSomething, &obj, boost::lambda::bind(&this_pair::first, boost::lambda::_1), boost::lambda::bind(&this_pair::second,boost::lambda::_1)) ); }

2 2

Build boost for i686 (instead of i386)
by Christoph Duelli 18 Dec '09

18 Dec '09

What do I have to pass to bjam to build Boost with i686 instead of the default i386 as cpu? (I am using Linux, with gcc 4.3.2) So far I am calling ./bjam architecture=x86 address-model=32 Best regards Christoph

2 1

Re: [Boost-users] [Boost.Test] valgrind complains about invalid reads
by Scott Banachowski 18 Dec '09

18 Dec '09

Thanks for the reply Gennadiy, see inline for comments: --- On Wed, 12/16/09, Gennadiy Rozental <rogeeff(a)gmail.com> wrote: > From: Gennadiy Rozental <rogeeff(a)gmail.com> > Subject: Re: [Boost-users] [Boost.Test] valgrind complains about invalid reads > To: boost-users(a)lists.boost.org > Date: Wednesday, December 16, 2009, 9:32 PM > Scott Banachowski <sbanacho > <at> yahoo.com> writes: > > > > > > > I agree with Peter here, the destructor for test_unit > should be virtual. > > 'Should' is a strong statement. Do you have anything to > based it on? > Ok, should is too strong. Change that to "it would be nice to", because the possibility that the base class destructor may be called (even if not by your code, possibly by others). > > I find that it is being deleted when > > test_unit* is a stored in reference counted pointer, > > Where? My mistake, it's held in a map. See the next comment: > > > and at program exit it is being deleted by calling > ~test_unit(). > > Where do I call destructor explicitly? You don't. It's being called implicitly at program exit, by valgrind. I think what is happening is that valgrind replaces the delete operator with its own, which is "operator delete(void*)" The typeinfo is likely lost by this replacement of delete. Just because you don't call the destructor, it doesn't mean others won't. If a base class is derived from, best practice is to declare its destructor virtual. It is relatively common to replace allocators (e.g. tcmalloc). > > > If you use a derived class (test_suite), this is > leading to an error. > > Where? As already mentioned, reported by valgrind. > > > Also, in the code I'm looking at (1.40.0), the > test_suite class has a > > virtual destructor, but its base class test_unit > does not. > > Not making the base class virtual seems to be > defeating the purpose. > > What purpose? The only reason that test_suite destructor is > virtual is because > this class in fact is being polymorphically. It has > subclass master_test_suite. > By defeating the purpose, I mean it forces code like the following: if( ut_detail::test_id_2_unit_type( tu.second->p_id ) == tut_suite ) delete static_cast<test_suite const*>(tu.second); else delete static_cast<test_case const*>(tu.second); A virtual base case simplifies this code (especially if in the future other classes are added, or in the case of replacement allocator it does the right thing). thanks, Scott > Gennadiy > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

4 3

[serialization] object tracking for shared strings
by Nils Krumnack 17 Dec '09

17 Dec '09

Hi all, I hope one of you will be able to help me with this: I am working with a data structure that contains a large number of (mostly identical) strings. To save memory I put together a shared memory string implementation, which ensures that each string is kept in memory only once (implemented through a global map with all strings). However, once I write them to disk, I again have n copies of each string. I would really prefer, if I could avoid the space overhead. At first glance, object tracking seems to be the way to go. If my strings were tracked, there would automatically be only one copy of each string. However I can't quite figure out how. I don't really want to change the global tracking behavior for std::string, since that might end up affecting code all over the place. Serializing pointers would be an option, but then I would have to deal with the issue of when to deallocate that object. I suppose I could serialize shared pointers, but then I would have to change my implementation to use shared pointers internally. Alternatively I could also wrap a struct around the strings internally and turn on tracking for that struct, but again I would have to complicate my implementation for the sake of serialization. What I was hoping to find was some way of telling the archive at the point where I am serializing the object, that it should track this one object. I am thinking of something like ar & BOOST_SERIALIZATION_TRACK_THIS (mystring); I was looking for something like that in the documentation, but wasn't very successful. Any ideas? Thanks in advance for any help, Nils

3 7

are boost::interprocess containers thread-safe?
by blp330 17 Dec '09

17 Dec '09

Hi, I must implement a multiple producer-consumer IPC application in win32 (but will port to linux in near future). I am wondering whether the boost::interprocess container like list, vector or map is thread-safe (actually, interprocess-safe?) or I need to use lock in my producer-consumer application? I mean, can Producer just simply put shared memory string to container and take it from Consumer without any lock? Or is there any multiple producer-consumer example that uses boost::interprocess? Thanks a lot. -- View this message in context: http://old.nabble.com/are-boost%3A%3Ainterprocess-containers-thread-safe--t… Sent from the Boost - Users mailing list archive at Nabble.com.

2 3

[proto] extends example not compiling
by alfC 17 Dec '09

17 Dec '09

Is this example, contained in the boost.Proto manual supposed to compile? #include<complex> #include <boost/proto/core.hpp> #include <boost/proto/context.hpp> #include <boost/proto/debug.hpp> #include<boost/proto/extends.hpp> template<class T, class Dummy = boost::proto::is_proto_expr> struct my_complex { BOOST_PROTO_EXTENDS( typename boost::proto::terminal<std::complex<T> >::type , my_complex<T> , boost::proto::default_domain ) }; int main() { my_complex<int> c0, c1; c0 + c1; // OK, operator+ found now! } I get this error (with gcc 4.3 and gcc 4.4): proto_complex.cpp:10: error: expected nested-name-specifier before ‘typename’ proto_complex.cpp:10: error: two or more data types in declaration of ‘proto_base_expr’ proto_complex.cpp:10: error: ‘proto_base_expr’ has not been declared proto_complex.cpp:10: error: ‘proto_base_expr’ has not been declared The strangest thing is that when I replace the macro by its explicit code the code seems to compile. Thank you, Alfredo

2 5