- Boost-users - lists.boost.org

Re: [Boost-users] [boost] [Review] Boost.Atomic review starts today, October 17th 2011
by Tim Blechmann 17 Oct '11

17 Oct '11

> > The review of Helge Bahmann's Boost.Atomic library starts today, October > > 17th 2011. It will end on October 26th. > > > > I encourage everyone to participate the the discussion on the Boost > > mailing > > > list about this very important library. In addition the code can be > > reviewed > > > line-by-line in code collaborator (http://demo.smartbear.com/boost/) > > Is there some explanation on how this works? I have created an account on > that site for myself and see this Boost.Atomic review as being in progress, > but it also tells me I'm not a participant. not really .. you can simply browse the files, select files and add per-line comments. not sure, if you have get the `reviewer' role or if you can add yourself to the project (iac, i've just made your a `reviewer'). from my experience, code collaborator is better for code reviews, while the lists are better to discuss the architecture, documentation, concepts etc ... cheers, tim

1 0

[BGL] using subgraph and SCC in algorithm
by Avishay Ben Shabtai 17 Oct '11

17 Oct '11

hi i am trying to figure out how is the best way to use the capabilities of subgraph. i am using the following declarations typedef boost::adjacency_list<boost::listS ,boost::vecS ,boost::bidirectionalS // ,GVertexProp // ,CFGEdgeProp> // GBase; typedef GBase::vertex_descriptor G_v_id; class Graph : public GraphBase { public: remove(const std::string& vName){ G_v_id vToRemove = findVertexByName(vName); boost::remove_vertex(vToRemove,*this); } } now my algorithm uses SCCs and works and modify each component separately and the idea of subgraph charmed me so i performed the following change typedef boost::subgraph<boost::adjacency_list<boost::listS ,boost::vecS ,boost::bidirectionalS // ,GVertexProp // ,CFGEdgeProp> > // GBase; typedef GBase::vertex_descriptor G_v_id; and now it stops to compile on boost::remove_vertex. is there an elegant way to use subgraph instead of copying it to separate graph and working from there.

1 0

[Review] Boost.Atomic review starts today, October 17th 2011
by Tim Blechmann 16 Oct '11

16 Oct '11

Hi all, The review of Helge Bahmann's Boost.Atomic library starts today, October 17th 2011. It will end on October 26th. I encourage everyone to participate the the discussion on the Boost mailing list about this very important library. In addition the code can be reviewed line-by-line in code collaborator (http://demo.smartbear.com/boost/) About Boost.Atomic: Boost.Atomic is a library that provides atomic data types and operations on these data types, as well as memory ordering constraints required for coordinating multiple threads through atomic variables. It implements the interface as defined by the C++11 standard, but makes this feature available for platforms lacking system/compiler support for this particular C++11 feature. Tarball: http://www.chaoticmind.net/~hcb/projects/boost.atomic/boost.atomic.tar.gz Documentation: http://www.chaoticmind.net/~hcb/projects/boost.atomic/doc/index.html Please always state in your review, whether you think the library should be accepted as a boost library! Additionally please consider giving feedback on the following general topics: - What is your evaluation of the design? - What is your evaluation of the implementation? - What is your evaluation of the documentation? - What is your evaluation of the potential usefulness of the library? - Did you try to use the library? With what compiler? On which platform? Did you have any problems? - How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? - Are you knowledgeable about the problem domain? Cheers, Tim Review Manager

1 0

building bjam vc9 troubles...
by Dan Bloomquist 16 Oct '11

16 Oct '11

I'm trying to figure out what is wrong with my system, lots of google searching and everything seems to check. My |%VS90COMNTOOLS% is good and my VS works fine in every respect I use it. But when I try a 'build' in ../v2/engine even with 'msvc' or 'vc9' as an argument it just loops like: C:\cpp\boost_1_47\tools\build\v2\engine>build ### ### "Could not find a suitable toolset." ### ### You can specify the toolset as the argument, i.e.: ### .\build.bat msvc ### ### Toolsets supported by this script are: borland, como, gcc, gcc-nocygwin, ### intel-win32, metrowerks, mingw, msvc, vc7, vc8, vc9, vc10 ### ### ### "Could not find a suitable toolset." ### ### You can specify the toolset as the argument, i.e.: ### .\build.bat msvc ### Over and over until I control c it. No, I have yet to dig into the workings of build.bat for a better answer. I was hoping someone knew of the top of their head what may be my problem. I had last built 1_42 without a hitch back when it was released. But I've started using boost::geometry and need the lib for boost::filesystem. I tried to mix 1.42 filesystem and 1.47 geometry headers, but it bumps the filesystem lib needed to 1.47, I can't seem to cheat. ~~ cannot open file 'libboost_filesystem-vc90-mt-gd-1_47.lib' Thanks, Dan. |

2 2

Locale on Mac OS X
by Alberto Simões 16 Oct '11

16 Oct '11

Hello, friends I am kind of desesperate. I am trying to use a Natural Language Processing application written in C++ that uses libboost on my Mac OS X. The problem is that locales seem to be broken for C++ in darwin. I know that is is (as far as I can tell) libboost fault. If you run the c++ program in attach under Leopard, you get the behavior I paste below. My question is: anybody found a way to make locale working on darwin? Couldn't libboost try to implement an indirection level for mac os X using the C locale API? (not sure if you are interested, just an idea). Please let me know if there is a solution to this problem. Kindest regards, Alberto [ambs@rachmaninoff tmp]$ ./a.out Trying c locales... Succeeded in calling setlocale(LC_ALL, "C") Succeeded in calling setlocale(LC_ALL, "POSIX") Succeeded in calling setlocale(LC_ALL, "en_US") Succeeded in calling setlocale(LC_ALL, "en_US.UTF-8") Succeeded in calling setlocale(LC_ALL, "pl_PL") Succeeded in calling setlocale(LC_ALL, "pl_PL.ISO8859-2") Trying C++ locales... Succeeded ctor for std::locale with "C" Succeeded ctor for std::locale with "POSIX" Failed ctor for std::locale with "en_US": locale::facet::_S_create_c_locale name not valid Failed ctor for std::locale with "en_US.UTF-8": locale::facet::_S_create_c_locale name not valid Failed ctor for std::locale with "pl_PL": locale::facet::_S_create_c_locale name not valid Failed ctor for std::locale with "pl_PL.ISO8859-2": locale::facet::_S_create_c_locale name not valid -- Alberto Simões

2 2

[Objective-C++] Specializing boost::lexical_cast
by Kelvin Chung 16 Oct '11

16 Oct '11

I'm trying to specialize boost::lexical_cast to convert Objective-C objects into various standard C++ classes (for example, leveraging NSNumber::-intValue to create lexical_cast<int>(NSNumber*)). However, as Objective-C objects exist as pointers, I would imagine that they would fail the requirements of needing well-defined streaming operator<<() in order to use the stock boost::lexical_cast. Any words of advice on how to do this in a way that will do as little damage as possible?

2 2

[threadpool] Using with Boost.Bind
by Robert Jones 15 Oct '11

15 Oct '11

Hi All I'm making use of a threadpool implementation from sourceforge, which I know is not properly part of Boost, but I'm hoping someone has some experience which can help me. I'm trying to use one of the functions from pool_adaptors.hpp, which I've reproduced below. I think this contains an error in the first template as the parameter passed as a Pool is used as a pointer to Pool, and I'm guessing there should four functions in here, not three, as the set seems incomplete. On either count I might be missing something. Stripped down my code says approximately this... #include "threadpool.hpp" #include "boost/threadpool/pool_adaptors.hpp" using namespace boost::threadpool; void task( int ); vector<int> v; v += 1,2,3,4,5,6,7,8,9; pool my_pool; for_each( v.begin( ), v.end( ), bind( schedule, my_pool, _1 ) ); ...and of course the problem is that 'schedule' is both templated and overloaded, so the compiler complains it can't find/resolve 'schedule'. The usual solution is to express the exact type of 'schedule' required, but it's huge, and uses types internal the the threadpool library. Is there a better solution? Thanks - Rob. namespace boost { namespace threadpool { // TODO convenience scheduling function /*! Schedules a Runnable for asynchronous execution. A Runnable is an arbitrary class with a run() * member function. This a convenience shorthand for pool->schedule(bind(&Runnable::run, task_object)). * \param * \param obj The Runnable object. The member function run() will be exectued and should not throw execeptions. * \return true, if the task could be scheduled and false otherwise. */ template<typename Pool, typename Runnable> bool schedule(Pool& pool, shared_ptr<Runnable> const & obj) { return pool->schedule(bind(&Runnable::run, obj)); } /*! Schedules a task for asynchronous execution. The task will be executed once only. * \param task The task function object. */ template<typename Pool> typename enable_if < is_void< typename result_of< typename Pool::task_type() >::type >, bool >::type schedule(Pool& pool, typename Pool::task_type const & task) { return pool.schedule(task); } template<typename Pool> typename enable_if < is_void< typename result_of< typename Pool::task_type() >::type >, bool >::type schedule(shared_ptr<Pool> const pool, typename Pool::task_type const & task) { return pool->schedule(task); } } } // namespace boost::threadpool

2 2

Using boost::signal2 for generic event handling with overloaded methods
by Alessandro Giulianelli 15 Oct '11

15 Oct '11

Hi People, I'm studying boost to understand if I can remove from a project a homemade event-dispatcher with boost::signal2 but I found a lot of troubles and I cannot solve it. My actual code which I would migrate to boost is made by a "Sender" and "Multiple Receiver" The sender has the ability to register many "signals" (I use a map for do this" ) and each listener has a map of "slots" connected to the sender of the sender. The sender code is: struct Event {}; struct ID {}; struct OneSecondElapsedEvent : public Event { static ID id; }; struct UpdatePresentationTracksReqEvent : public Event { UpdatePresentationTracksReqEvent() {}; static ID id; }; typedef boost::signals2::signal<void (Event*)> signal_t; typedef signal_t::slot_type slot_t; class MainTask: public Elt::Thread , public Sender { boost::ptr_map<ID*, signal_t> signals_map; public: sig; MainTask(char* taskName); virtual ~MainTask(); void Main(); void registerListener(IListener* listener) { if(listener != NULL) { SlotMap::const_iterator it = listener->slots_map.begin(); for(;it != listener->slots_map.end();it++) { ID *id = it->first; const slot_t *handler = it->second; signals_map[id].connect(*handler); } } else printf("Listener NULLO!!!!\n"); } }; now my generic "listener" or Receiver should be able to register to sender and the code is: #include "SpActiveContext.h" #include <cassert> #include "ActiveObject.h" #include "GenericMessageHandler/ActiveMessage.h" #include "boost/bind.hpp" #include "boost/function.hpp" #include <stdio.h> #include "EventDefinitions.h" #include "MainTask.h" SpActiveContext::SpActiveContext(SpContext * component): _ActiveObject("SpActiveContext"), _Component(component){ assert(_Component); * //boost::function<void (OneSecondElapsedEvent*)> g = boost::bind(&SpActiveContext::Handle, this, _1);* * * * //const slot_t& t = static_cast<void (SpActiveContext::*)(OneSecondElapsedEvent* )>(&SpActiveContext::Handle);* * //registerEvent(OneSecondElapsedEvent::id, t);* * * * //typedef boost::signals2::signal<void (Event*)> signal_t;* * //typedef signal_t::slot_type slot_t;* * * * //boost::bind(&SpActiveContext::Handle(OneSecondElapsedEvent*), this, _1); * * * * //void (SpActiveContext::*Handle) (OneSecondElapsedEvent*) = &SpActiveContext::Handle;* * * * //boost::bind( static_cast<void (SpActiveContext::*)( int )>(&Foo::Bar) , foo, _1);* * * * * * //const slot_t1& t = boost::bind( static_cast<void (SpActiveContext::*)(OneSecondElapsedEvent*)>(&SpActiveContext::Handle) , this, _1);* * //registerEvent(OneSecondElapsedEvent::id, t );* * * * //const slot_t1& t = boost::bind( static_cast<void (SpActiveContext::*)( UpdatePresentationTracksReqEvent *)>(&SpActiveContext::Handle) , this, _1); ** //registerEvent(UpdatePresentationTracksReqEvent ::id, t ); * } void SpActiveContext::Handle(OneSecondElapsedEvent* msg) { printf("OneSecondElapsedEvent 2\n"); } void SpActiveContext::Handle(UpdatePresentationTracksReqEvent* msg) { printf("UpdatePresentationTracksReqEvent\n"); } In the constructor I tried to "connect" the function "Handle(*OneSecondElapsedEvent*)" to the slot but I cannot to this..the compiler write errors sayng that cannot connvert from "Event*" to "OneSecondElapsedEvent*".* * * *Is there any solution to make it working?* * * *Thanks in advance,* *AlexGiul* * *

2 1

serialization basic question
by Luis A. Montes 15 Oct '11

15 Oct '11

Hi, I'm trying to use the serialization library and was having a compile error using MSVC 2005. So I decided to try and compile "A Very Simple Case" from the tutorial, and that of course works. But then I noticed that the example serializes a const object, that sort of is rather useless, so I removed the const specifier, and I got the same error that I was getting with my program, namely: warning C4099: 'boost::serialization::static_warning_impl<false>::STATIC_WARNING' : type name first seen using 'struct' now seen using 'class' I went googling for an answer and I found that I should look at the rationale on the docs, but the rationale (http://www.boost.org/doc/libs/1_47_0/libs/serialization/doc/rationale.html) doesn't seem at all relevant. I finally sort of guessed that I wanted was probably to modify the tracking behavior on my program (looking at the thread http://lists.boost.org/boost-users/2010/05/59107.php ). But my objects are not on the stack, so I'm not really sure. My class structure is a bit more complicated than the simple example, but my question basically is, what changes would be appropriate on the simple example so that it would compile for non-const objects? Thanks

4 6

Boost Object Pool Ticket #3789
by Gokulakannan Somasundaram 15 Oct '11

15 Oct '11

Hi, I just want to know, when the patch will get applied for this bug. We want to use Boost object Pool, but we find this bug very disturbing. Thanks, Gokul.

1 1