- Boost-users - lists.boost.org

Generically connecting a signal to a member function
by Chris Stankevitz 18 Oct '12

18 Oct '12

Hello, I wrote a function called "Connect" which generically connects a signal to a member function. The pseudo-code for this function is: void Connect(Signal, MemberFunction, pClass) { Signal.connect(Signal::slot_type(MemberFunction, pClass)); } Problem: I had to create "overloads" for different number of signal arguments. This was somewhat expected and we see it in boost with things like boost::bind. What I did not expect was that each of my overloads had to have a different name. This is because all overloads have the same function signature: one signal, one member function, one class pointer. Question: Is there a way to write a generic "Connect" that handles signals of different number of arguments? If overloading is involved, can it be done like bind such that all overloads have the same name? Below is my solution that exhibits the "problem" I described above. Thank you for your help, Chris === #include <boost/bind.hpp> #include <boost/signals2.hpp> #include <iostream> //------------------------------------- // Connect0 // // Generically connect a member // function to a signal producing zero // arguments. //------------------------------------- template < class TASignal, class TAMemberFunction, class TAClass> void Connect0( TASignal& Signal, TAMemberFunction MemberFunction, TAClass* pClass) { typedef typename TASignal::slot_type TDSlot; TDSlot Slot(MemberFunction, pClass); Signal.connect(Slot); } //------------------------------------- // Connect0 // // Generically connect a member // function to a signal producing one // arguments. //------------------------------------- template < class TASignal, class TAMemberFunction, class TAClass> void Connect1( TASignal& Signal, TAMemberFunction MemberFunction, TAClass* pClass) { typedef typename TASignal::slot_type TDSlot; TDSlot Slot(MemberFunction, pClass, _1); Signal.connect(Slot); } //------------------------------------- // Connect2 // // Generically connect a member // function to a signal producing two // arguments. //------------------------------------- template < class TASignal, class TAMemberFunction, class TAClass> void Connect2( TASignal& Signal, TAMemberFunction MemberFunction, TAClass* pClass) { typedef typename TASignal::slot_type TDSlot; TDSlot Slot(MemberFunction, pClass, _1, _2); Signal.connect(Slot); } //------------------------------------- //------------------------------------- struct TCClass { void f0() {} void f1(int) {} void f2(int, int) {} }; //------------------------------------- //------------------------------------- int main() { boost::signals2::signal<void()> Signal0; boost::signals2::signal<void(int)> Signal1; boost::signals2::signal<void(int, int)> Signal2; TCClass Class; Connect0(Signal0, &TCClass::f0, &Class); Connect1(Signal1, &TCClass::f1, &Class); Connect2(Signal2, &TCClass::f2, &Class); Signal0(); Signal1(1); Signal2(2, 2); return 0; }

1 0

Intrsuive multisets and stable sorting
by Dan Posluns 18 Oct '12

18 Oct '12

Hi there, I'm attempting to use intrusive associative containers, but it's essential to the application that as I add "equivalent" objects to the multiset that their order is maintained, ie. that the sort is stable. I'm looking at using avl_multiset for its fast retrieve times. I ran an experiment several times over where I added 1000 entries of random numbers from 0-9 that also contained ordering information to a multiset, and that ordering information wound up being preserved, even when I removed some items and added new ones to the initial tree. So that looks promising. If someone were able to confirm the results of my experiment it would be better, though. Best would be documentation that lists which multiset implementations are stable and which aren't, but I can't find such a thing in the docs. Thanks, Dan.

3 5

[fusion] making an inner struct boost fusion compatible?
by Oswin Krause 17 Oct '12

17 Oct '12

Hi list, I have the following problem: template<class T, class U, class V> struct MyType{ struct nested{ typename MetaFunc<T>::type a; typename MetaFunc<U>::type b; typename MetaFunc<V>::type c; }; }; now this inner struct must be a boost fusion compatible random access sequence. The examples cover only structs on namespace lvel, how can i handle this way? Later these classes will be generated by a macro automatically so even dirty hacks are allowed :). But i don't want to clatter the namespace by pulling nested outside of the class. Is there an easy way or do i have to write the necessary mtafunctions to make this possible? Greetings, Oswin

3 8

undefined reference to boost::hash_range
by Michael Dickey 17 Oct '12

17 Oct '12

I ran into an odd error with 1.51.0 when compiling the following code using gcc 4.4.6 on rhel6/64 with -fkeep-inline-functions (using this flag for debug builds): #include <boost/filesystem/path.hpp> int main(void) { boost::filesystem::path a_path; return 0; } The linker complains about an undefined reference to boost::hash_range<> referenced from boost/filesystem/path.hpp. Looking at path.hpp I can see it includes hash_fwd.hpp not hash.hpp. This is despite the fact that it's actually using various hash methods inside the header file (including hash_range). When I change it to use hash.hpp instead, the linker error goes away. Should path.hpp really use hash.hpp instead, or is this some weird compiler issue, or something else entirely? I figured someone here would know =) Thanks, -Mike

1 0

[Phoenix] Restricting actors
by paul Fultz 17 Oct '12

17 Oct '12

Hello, I was wondering if there was a way to restrict actors(rather than extend them). That is, I would like to create an actor that only supports say the `+` operator. Right now, if I inherit from `actor` it will bring in all the C++ operators. Is there a different base class I can inherit from? Or is it even possible? Thanks, Paul

1 0

IOStream - expose MemoryInputStream buffer and it's size
by Gheorghe Marinca 16 Oct '12

16 Oct '12

Hy, I want to implement a boost iostream MemoryInputStream like this: class MemoryInputDevice : public boost::iostreams::source { public: MemoryInputDevice(char* buffer, size_t size): buffer(buffer), size(size), pos(0) { } std::streamsize read(char* s, std::streamsize n); private: char* buffer; std::streamsize size; std::streamsize pos; }; typedef boost::iostreams::stream<MemoryInputDevice> MemoryInputStream; I want at the same time to expose to clients beside the istream interface a way for the client to get the pointer to the internal buffer and it's length (getting the pointer to the internal buffer could be questionable but in order not to copy buffer around there are cases where a parser of mine needs all the content of the internal buffer and I don't want to copy that around before passing it to the parser) Is therea way for MemoryInputStream user to get that info ? I was thinking at the stream->rdbuffer() method but it seems it does not give me what I want. Thanks in advance for your help -Ghita

2 1

How to do line clipping using boost polygon?
by jian hu 16 Oct '12

16 Oct '12

Hi, I'm trying to clip line segments using boost polygon. Boost polygon does not provide public interface but somehow I find the function validate_scan(), which can do the line clipping. My input are two segments < (10, 10), (10, 2), 0 > < ( 6, 7), (11, 2), 1 > The numbers in the brackets are coordinates of two end points of the segments and the number after it is the segment id. So it looks like this. The result I expect should be four segments, which are the result of intersection <(10,10), (10,3), 0> <(10,3), (10,2), 0> <(6,7 ), (10,3), 1> <(10,3), (11, 2), 1> However, the <(10,3), (11,2), 1> segment is replace by two segments, which are <(10,3), (10,2), 1> and <(10,2), (11,2), 1>. Can someone tell me what happened? Replacing the segment with the other two are not right behavior. Thank you very much. -- Jian Hu

1 0

[boost.numeric] Poor Performance of numeric_cast
by Tang Jiang Jun 16 Oct '12

16 Oct '12

Hi, I have run a performance testing for numeric_cast recently, and found that the result was really unexpected bad, although the document mentioned that it will be no overhead if overflows don't happen. Could somebody please help me to verify this testing? If this is true, I doubt whether I should use numeric_cast in the production code. Here is my testing code and result. *#include <boost/numeric/conversion/cast.hpp> #include <boost/format.hpp> #include <boost/cstdint.hpp> #include <boost/chrono.hpp> #include <iostream> using namespace std; using namespace boost; using namespace boost::numeric; using namespace boost::chrono; int main() { const static int32_t COUNT = 1000000; high_resolution_clock::time_point start; start = high_resolution_clock::now(); for( int32_t n = 0; n < COUNT; ++n ) { int32_t i32 = 123; int16_t i16 = i32; } cout << format("Native Integer Cast: %1%\n") % ( ( high_resolution_clock::now() - start ) / COUNT ); start = high_resolution_clock::now(); for( int32_t n = 0; n < COUNT; ++n ) { try { int32_t i32 = 100; int16_t i16 = numeric_cast< int16_t >( i32 ); } catch( const bad_numeric_cast& e ) { cout << e.what() << endl; } } cout << format("Boost Integer Cast: %1%\n") % ( ( high_resolution_clock::now() - start ) / COUNT ); start = high_resolution_clock::now(); for( int32_t n = 0; n < COUNT; ++n ) { float f = 100.0f; int32_t i = static_cast< int32_t >( f ); } cout << format("Native Floating-Integer Cast: %1%\n") % ( ( high_resolution_clock::now() - start ) / COUNT ); start = high_resolution_clock::now(); for( int32_t n = 0; n < COUNT; ++n ) { try { float f = 123.0f; int32_t i = numeric_cast< int32_t >( f ); } catch( const bad_numeric_cast& e ) { cout << e.what() << endl; } } cout << format("Boost Floating-Integer Cast: %1%\n") % ( ( high_resolution_clock::now() - start ) / COUNT ); start = high_resolution_clock::now(); for( int32_t n = 0; n < COUNT; ++n ) { int32_t i = 132; float f = static_cast< float >( i ); } cout << format("Native Integer-Floating Cast: %1%\n") % ( ( high_resolution_clock::now() - start ) / COUNT ); start = high_resolution_clock::now(); for( int32_t n = 0; n < COUNT; ++n ) { try { int32_t i = 128; float f = numeric_cast< float >( i ); } catch( const bad_numeric_cast& e ) { cout << e.what() << endl; } } cout << format("Boost Integer-Floating Cast: %1%\n") % ( ( high_resolution_clock::now() - start ) / COUNT ); return 0; };* Result: *Native Integer Cast: 3 nanoseconds Boost Integer Cast: 311 nanoseconds Native Floating-Integer Cast: 4 nanoseconds Boost Floating-Integer Cast: 430 nanoseconds Native Integer-Floating Cast: 2 nanoseconds Boost Integer-Floating Cast: 106 nanoseconds*

4 12

[asio] Custom asynchronous service
by Acer Yang 16 Oct '12

16 Oct '12

Hi all, I would like to adapt my class to asio for offering async_do_something(param, callback) interfaces. After reading examples presented in asio/examples/services/ and some asio source code. I found the example does not do invoke callback stuff (e.g. asio/examples/service/logger_service.cpp). In case of asio source code, there are classes such as detail::wait_handler<> that wrap user callbacks for invoked later. Should I use these wrappers in my service implementation if I hope to comfort the strand design of asio? Do I have other choices? All suggestions are welcome. Best wishes, Acer.

2 3

boost test, valgrind errors in release mode
by Avi Bahra 15 Oct '12

15 Oct '12

I have several libs that I am testing with boost unit test Valgrind shows mismatched free/delete Using opensuse113, gcc 4.5, boost 1.47 I am bit perplexed because: a/ The valgrind errors only show up with release mode. b/ The valgrind errors suggest problems with unit test, but I have dozen or so separate unit tests, however, errors only shows up on one of the tests. has anyone come across these errors, and suggestion of how to overcome. + valgrind --leak-check=full --show-reachable=yes ACore/bin/gcc-4.5/release/coretest ==25529== Memcheck, a memory error detector ==25529== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. ==25529== Using Valgrind-3.5.0 and LibVEX; rerun with -h for copyright info ==25529== Command: ACore/bin/gcc-4.5/release/coretest ==25529== ==25529== Mismatched free() / delete / delete [] ==25529== at 0x4C25F7B: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==25529== by 0x4D25CA: _ZNSt6vectorISt4pairIN5boost9unit_test13basic_cstringIKcEES5_ESaIS6_EE13_M_insert_auxEN9__gnu_cxx17__normal_iteratorIPS6_S8_EERKS6_.clone.57 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4D3B09: _Z41__static_initialization_and_destruction_0ii.clone.0 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x50EEF5: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4265AA: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x511D22F: ??? (in /usr/lib64/libstdc++.so.6.0.14) ==25529== by 0x50EE74: __libc_csu_init (elf-init.c:120) ==25529== by 0x55C1B0F: (below main) (in /lib64/libc-2.11.2.so) ==25529== Address 0x5904c60 is 0 bytes inside a block of size 32 alloc'd ==25529== at 0x4C267D0: operator new(unsigned long) (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==25529== by 0x4D2561: _ZNSt6vectorISt4pairIN5boost9unit_test13basic_cstringIKcEES5_ESaIS6_EE13_M_insert_auxEN9__gnu_cxx17__normal_iteratorIPS6_S8_EERKS6_.clone.57 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4D2F62: _Z41__static_initialization_and_destruction_0ii.clone.0 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x50EEF5: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4265AA: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x511D22F: ??? (in /usr/lib64/libstdc++.so.6.0.14) ==25529== by 0x50EE74: __libc_csu_init (elf-init.c:120) ==25529== by 0x55C1B0F: (below main) (in /lib64/libc-2.11.2.so) ==25529== ==25529== Mismatched free() / delete / delete [] ==25529== at 0x4C25F7B: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==25529== by 0x4D25CA: _ZNSt6vectorISt4pairIN5boost9unit_test13basic_cstringIKcEES5_ESaIS6_EE13_M_insert_auxEN9__gnu_cxx17__normal_iteratorIPS6_S8_EERKS6_.clone.57 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4D3B29: _Z41__static_initialization_and_destruction_0ii.clone.0 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x50EEF5: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4265AA: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x511D22F: ??? (in /usr/lib64/libstdc++.so.6.0.14) ==25529== by 0x50EE74: __libc_csu_init (elf-init.c:120) ==25529== by 0x55C1B0F: (below main) (in /lib64/libc-2.11.2.so) ==25529== Address 0x5904cc0 is 0 bytes inside a block of size 64 alloc'd ==25529== at 0x4C267D0: operator new(unsigned long) (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==25529== by 0x4D2561: _ZNSt6vectorISt4pairIN5boost9unit_test13basic_cstringIKcEES5_ESaIS6_EE13_M_insert_auxEN9__gnu_cxx17__normal_iteratorIPS6_S8_EERKS6_.clone.57 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4D3B09: _Z41__static_initialization_and_destruction_0ii.clone.0 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x50EEF5: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4265AA: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x511D22F: ??? (in /usr/lib64/libstdc++.so.6.0.14) ==25529== by 0x50EE74: __libc_csu_init (elf-init.c:120) ==25529== by 0x55C1B0F: (below main) (in /lib64/libc-2.11.2.so) ==25529== ==25529== Mismatched free() / delete / delete [] ==25529== at 0x4C25F7B: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==25529== by 0x4D25CA: _ZNSt6vectorISt4pairIN5boost9unit_test13basic_cstringIKcEES5_ESaIS6_EE13_M_insert_auxEN9__gnu_cxx17__normal_iteratorIPS6_S8_EERKS6_.clone.57 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4D3B69: _Z41__static_initialization_and_destruction_0ii.clone.0 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x50EEF5: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4265AA: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x511D22F: ??? (in /usr/lib64/libstdc++.so.6.0.14) ==25529== by 0x50EE74: __libc_csu_init (elf-init.c:120) ==25529== by 0x55C1B0F: (below main) (in /lib64/libc-2.11.2.so) ==25529== Address 0x5904d40 is 0 bytes inside a block of size 128 alloc'd ==25529== at 0x4C267D0: operator new(unsigned long) (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==25529== by 0x4D2561: _ZNSt6vectorISt4pairIN5boost9unit_test13basic_cstringIKcEES5_ESaIS6_EE13_M_insert_auxEN9__gnu_cxx17__normal_iteratorIPS6_S8_EERKS6_.clone.57 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4D3B29: _Z41__static_initialization_and_destruction_0ii.clone.0 (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x50EEF5: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x4265AA: ??? (in /var/tmp/ma0/clientRoot/workspace/MyProject/ACore/bin/gcc-4.5/release/coretest) ==25529== by 0x511D22F: ??? (in /usr/lib64/libstdc++.so.6.0.14) ==25529== by 0x50EE74: __libc_csu_init (elf-init.c:120) ==25529== by 0x55C1B0F: (below main) (in /lib64/libc-2.11.2.so) ==25529== Best regards, Ta, Avi

1 0