August 2009 - Boost-users - lists.preview.boost.org

memory overhead of boost::function<>
by Zachary Turner 12 Aug '09

12 Aug '09

Is there an easy way to calculate the minimum and maximum memory overhead required by a single instance of boost::function? It's not easy to trace through the code and find all instances of allocations / deallocations and member variables. sizeof(boost::function<T>) appears to be 32 on my machine for all values of T but I'm not sure if there are certain types of functions which might be less than this or more than this. Is there any way to reduce the memory overhead of boost::function? 32 is quite a bit on an x86 architecture, considering a normal function pointer on the same machine takes only 4 bytes. I was thinking of using boost::function<> as the basis for a generic inter-object messaging system where there may be thousands of objects in the systems any of which might want to communicate via event sinks / sources implemented via boost::function<> and callback registration. But if the memory overhead is 32-bytes per object at a minimum, and possibly more with more complicated function objects or on 64-bit architectures, then it seems like it could quickly grow out of hand if there were many listeners.

2 1

[Test] Making use of boost::exception_diagnostic_information and friends
by Alexander Lamaison 12 Aug '09

12 Aug '09

Are there any plans to let the Boost.Test runner make use of the extra information that the Boost.Exception library is able to 'inject' into exceptions? Currently, when an exception is caught by the test runner and it is derived from std::exception, the report includes the what() string but is not able to give any infomation about the location the exception was thrown from: unknown location(0): fatal error in "cfstr_shellidlist_multiple_items": std::runtime_error: All items in a DataObject must have the same parent. When an exception is thrown using the BOOST_THROW_EXCEPTION() macro, extra information including the location it was thrown from are added to the exception: http://www.boost.org/doc/libs/1_39_0/libs/exception/doc/BOOST_THROW_EXCEPTI…. This allows callers to pass an exception to diagnostic_infotmation() and get a string detailing the exact nature of the exception thrown. It should be possible to let the test runner extract this information in the same way that diagnostic_infomation does and use it to improve the test runner output. Many thanks. Alex

4 17

[mpi] what is the equivalent of MPI_COMM_SELF?
by alfC 12 Aug '09

12 Aug '09

Is there something equivalent to the constant MPI_COMM_SELF in Boost.MPI? It is easy to refer the MPI_COMM_WORLD since this is the default constructor of boost::mpi::communicator, but it seems that the only way to have the equivalent of MPI_COMM_SELF is to do boost::mpi::communicator self(MPI_COMM_SELF, boost::mpi::communicator::comm_attach); does it make sense to define a global object with that name to refer to the communicator in current process? (this can be handy to switch to serial bahavior and parallel executrion when debugging the code, I guess). Thanks, Alfredo

3 3

[BGL] Can't correctly copy data
by Caligula 12 Aug '09

12 Aug '09

Hi, I've been working on this bug for a few weeks now. Its sort of a show stopper for me. I'm not able to correctly copy a simple string datum from one element to another such element. Just go to http://pastebin.com/m6f57be95 And see pastebin references. Either I'm doing something wrong or BGL is seriously flawed. Thanks Caligula

3 6

[Boost.Array] Checked iterator in debug mode
by dariomt＠gmail.com 12 Aug '09

12 Aug '09

Hi list, I can see that boost::array::operator[] has bounds checking in debug mode. Is there a way to get checked iterators also? Like safe STL style. If not, would that be a useful addition to boost::array? And if I was to add checked iterators to boost::array, do you have any recommendations? Any boost::detail::checked_iterator I can reuse? Thanks!

1 0

[serialization] boost::system::error_category
by Rutger ter Borg 12 Aug '09

12 Aug '09

Dear list, I'm trying to write serialization code for Boost.System's error_codes, however I'm having trouble to serialize/deserialize an error_category, which are polymorphic singleton objects. I guess when saving, type information needs to be stored, like with polymorphic pointers? When loading, a global function (get_generic_category(), get_system_category(), etc.) could be used to obtain a reference to the singleton object. Is there any way to use the exported type information in such way that a reference to the correct singleton is obtained? Any suggestions on how to approach this? Thanks in advance, Kind regards, Rutger ter Borg

1 1

mpl::map ---> fusion::map conversion problems
by Zachary Turner 11 Aug '09

11 Aug '09

Hello, I'm trying to convert an MPL map to a fusion map using boost::fusion::result_of::as_map. I have the following code, with some of the leading-in code omitted for brevity although I don't think it's relevant: //listener_info_types := mpl::list< pair<handler<E>, token<H> > > typedef typename boost::mpl::transform< handler_types, token_types, parameterize_pair<boost::mpl::_1, boost::mpl::_2> >::type listener_info_types; //listener_info_lookup = mpl::map<EventType, pair<handler<E>, token<H> > typedef typename boost::mpl::transform< SupportedEvents, listener_info_types, boost::mpl::inserter< boost::mpl::map<>, boost::mpl::identity<boost::mpl::_1> > >::type listener_info_lookup; typedef typename boost::fusion::result_of::as_map<listener_info_lookup>::type fusion_listener_map; Deleting the last line and everything works. I've successfully used a similar technique to convert mpl vectors to fusion vectors in the past, is it not possible for maps? Or more likely I'm just doing something wrong? I can post more code if necessary including full type definitions leading up to all these previous typedefs, but basically I have a class parameterized initially with an mpl::list of some number of types. I then do a sequence of mpl::transform calls, none of which with use a custom inserter. The end result of all of that is listener_info_types above. So listener_info_types should be an mpl::list. The next call should then create an mpl::map, (SupportedEvents is the initial type used to parameterize the class, so an mpl::list as previously mentioned). Any suggestions? Thanks Zach

2 2

casting with boost
by zeev mintz 11 Aug '09

11 Aug '09

hi, i'm trying to do some simple casts using boost, to and from webToolkit widgets, i.e int i = 42; WString str; str = boost::lexical_cast<WString>(i); i get a compilaton error that goes something like this: lexical_cast.hpp(768): no operator ">>" matches these oprands... could it be a matter of updating my boost version? ( i'm using boost version 1.35) is there another way to make conversions without boost? thanks, Shalev.

2 1

Synchronous problem with asio
by Dan Benčík 11 Aug '09

11 Aug '09

Dear everybody, I have a problem using asio. My client/server application requires only synchronous communication. So, using the examples for synchro from the boost homepage, I have set up two procedures to send and receive data. Their code is as follows: void vReceive(tcp::socket & socket, std::string & szDest){ char szTmp_Buf [BUF_LEN + 1]; szDest = ""; std::cout << "Entering vReceive . . ." << std::endl; for (;;){ char szBuf [BUF_LEN]; boost::system::error_code error; uInt uiBytes_Recv = socket.read_some(boost::asio::buffer(szBuf), error); std::cout << " Read " << uiBytes_Recv << " bytes" << std::endl; if (error == boost::asio::error::eof) break; // Connection closed cleanly by peer. else if (error) throw boost::system::system_error(error); // Some other error. memcpy((void*) szTmp_Buf, (void*) szBuf, uiBytes_Recv ); szTmp_Buf[ uiBytes_Recv ] = '\0'; szDest += szTmp_Buf; }; std::cout << "Received" << szDest << std::endl; std::cout << "Leaving vReceive . . ." << std::endl << std::endl; }; void vSend(tcp::socket & socket, std::string & szSrc){ std::cout << "Entering vSend . . . " << std::endl; std::cout << "Sending " << szSrc << std::endl; boost::system::error_code ignored_error; boost::asio::write(socket, boost::asio::buffer(szSrc), boost::asio::transfer_all(), ignored_error); std::cout << "Leaving vSend . . . " << std::endl << std::endl; }; These procedures are just wrappers for the lines of code extracted from the boost examples. In my test applications, the client calls vReceive(socket, szDate); vSend(socket, std::string("Chop Suey!") ); vReceive(socket, szDate); vSend(socket, std::string("Halo")); and the server calls vSend(socket, message); vReceive(socket, szReceived); vSend(socket, message); vReceive(socket, szReceived); just to test the functionality. The problem is that both applications freeze after the first information exchange, as I have depicted on the following picture ( http://eubie.sweb.cz/asio_problem.JPG ). It seems that vReceive() procedure on the client side doesn´t finish while vSend() finishes on the server side. So, does anybody have any idea, what might be wrong? Just in case someone wanted to replicate the problem, I uploaded the complete sources to http://eubie.sweb.cz/asio_problem.rar file. Thank you all in advance, Daniel.

3 3

[Boost-users] http://www.boost.org/doc/libs/1_39_0/libs/regex/doc/vc71-performance.html
by Jerry Jeremiah 11 Aug '09

11 Aug '09

I have a couple questions. http://www.boost.org/doc/libs/1_39_0/libs/regex/doc/vc71-performance.htm l references PCRE 3.9 but http://www.pcre.org/ says that the current version is PCRE 7.9. I would really like to see current values for performance comparisons. Where would I get the software for this comparison? Is the source part of the Boost library? Thanks, Jerry

2 1