August 2011 - Boost-users - lists.preview.boost.org

[Container] Forward declaration of movable types
by Szymon Gatner 17 Aug '11

17 Aug '11

Hi, I am evaluating container library by Ion Gaztanaga. I know it is not officially a part of boost yet but I didn't know where else to ask this question. My problem: If there is a movable class 'MyMovable' (I use boost.move) it can be stored without unnecessary copies in move-aware vector that is declared like so: vector<MyMovable> movables; which is very cool until this vector is suppose be a member of other class like: //widget.hpp class Widget { vector<MyMovable> movables_; }; which is not cool at all because now widget.hpp must include my_movable.hpp as opposed to just forward declaring it if it stored say vector<shared_ptr<MyMovable>> which in turn makes movability of MyMovable rather useless. I guess my question is: how can one take advantage of move semantics in context of member containers without adding unnecessary class dependencies? Regards, Simon

2 3

About Bug #4978 (deadlock on interrupt)
by Fabien Chêne 17 Aug '11

17 Aug '11

Hello, The bug #4978 seems to be fixed, according to its status. However, is it really fixed in the release 1.47, as indicated by a comment at the end ? I cannot see anything about that in the release note for 1.47. I think that it would be worth mentioning it. Regards, -- Fabien

1 0

Re: [Boost-users] [containers review] A reminder that the containers review ends tomorrow
by John Maddock 17 Aug '11

17 Aug '11

The containers library review is now officially over. However, there are still a few drips a drabs of reviews coming in.... are there any more still pending? Thanks, John.

1 0

[MPI][Boost] Some Questions About Delta Stepping Dijkstra in Boost
by Chen Kai 17 Aug '11

17 Aug '11

Environment: Boost1.47 + Microsoft HPC 2008 SDK I find Delta Stepping Dijkstra and Crauser Dijkstra in boost, and both of them are slower than sequential dijkstra in boost. (About 10 times slower. I used 10 graphs, the biggest graph contains 10 million nodes and 10 million edges) Q1: I wonder why the parallel dijkstra is slower than sequential version? Q2: Can I adjust the number of threads that Delta Stepping used? (For exemple, I want it use 1, 2, 3, 4 threads respectively. And Can I make it sequential version?) Thx --- Chen Kai School of Computer Science, Fudan University blog: http://www.cnblogs.com/remlostime/

1 0

problems with installing boost python
by Wilson Tan 17 Aug '11

17 Aug '11

Hello everyone, I had been trying to build BOOST for SST, but so far I had been unsuccessful with the Python(and IOSTREAM) components. I followed Section 3.1( http://www.boost.org/doc/libs/1_47_0/libs/python/doc/building.html) but even that is not fully successful: wilson@vorpalsword:~/SST/boost_1_47_0$ ./bjam --build-dir=~/SST/boost_1_47_0/libs/python/example/quickstart toolset=gcc --verbose-test test notice: could not find main target test notice: assuming it is a name of file to create. Building the Boost C++ Libraries. Performing configuration checks - has_icu builds : no - ../config//has_gcc_visibility builds : yes - ../config//has_long_double_support builds : yes don't know how to make <e>test ...found 1 target... ...can't find 1 target... wilson@vorpalsword:~/SST/boost_1_47_0$ The first lines of my Python related errors while installing are the following ...skipped <p/usr/local/lib>libboost_iostreams.so for lack of <p/usr/local/lib>libboost_iostreams.so.1.47.0... gcc.compile.c++ bin.v2/libs/python/build/gcc-4.4.3/release/threading-multi/numeric.o In file included from ./boost/python/detail/prefix.hpp:13, from ./boost/python/numeric.hpp:8, from libs/python/src/numeric.cpp:6: ./boost/python/detail/wrap_python.hpp:75:24: error: patchlevel.h: No such file or directory ./boost/python/detail/wrap_python.hpp:78:2: error: #error Python 2.2 or higher is required for this version of Boost.Python. ./boost/python/detail/wrap_python.hpp:142:21: error: Python.h: No such file or directory ./boost/python/detail/wrap_python.hpp:203:4: error: #error "HAVE_LONG_LONG defined but not PY_LONG_LONG or LONG_LONG" In file included from ./boost/python/object/pointer_holder.hpp:14, from ./boost/python/to_python_indirect.hpp:10, from ./boost/python/converter/arg_to_python.hpp:10, Result of whereis: wilson@vorpalsword:~/SST/boost_1_47_0$ whereis python python: /usr/bin/python2.6 /usr/bin/python /etc/python2.6 /etc/python /usr/lib/python3.1 /usr/lib/python2.6 /usr/local/lib/python2.6 /usr/include/python2.6 /usr/share/python /usr/share/man/man1/python.1.gz For good measure, I also specified the interpreter location in project-config.jam. # Python configuration using python : 2.6 : /usr/bin/python ; using mpi ; Could somebody give me pointers on how to fix this and get Python running on Boost? I'd really appreciate it.

2 2

Re: [Boost-users] Custom services using boost::asio
by Marat Abrarov 16 Aug '11

16 Aug '11

Hi, Allan. > The Io part is easy to fit into asio, but I'm having problems seeing what > the best way is to implement the core application event processing. > > Using glib::source I could register an interface where two function is > called: > > bool prepare (int &timeout); > bool dispatch (sigc::slot_base *slot); > > Prepare could set a time out identifying how much time should elapse before > dispatch is ready to process an event. > > When the time is passed, the glib main loop will call the dispatch function, > where I then can process the internal events. > > How can this be implemented in boost::asio io_service? May be boost::asio::deadline_timer can help you? http://www.boost.org/doc/libs/1_47_0/doc/html/boost_asio/overview/timers.ht… http://www.boost.org/doc/libs/1_47_0/doc/html/boost_asio/tutorial/tuttimer2… http://www.boost.org/doc/libs/1_47_0/doc/html/boost_asio/reference/deadline… Regards, Marat Abrarov.

1 0

Building boost from straight Windows Command Prompt
by Michael Powell 16 Aug '11

16 Aug '11

So... If I want to build boost from a straight up Windows Command Prompt in the Visual Studio environment, the key is that we run the VCVARSALL.BAT file prior to running the bootstrap, etc, etc? For instance, if we wanted to route a conditional build of boost through our Continuous Integration process using NAnt, for instance... Thank you... Regards, Michael

2 3

How to iterate nested 'structs' in property tree
by Bill Moo 16 Aug '11

16 Aug '11

Hello List, I have an .info file with the following structure and as I am unable to use BOOST_FOREACH to iterate these I was wondering how it was done? Items { Item { ID 999 Description "Text to describe the item" Stuff1 1 Stuff2 0 Stuff3 7 ItemSettings { Setting1 9 Setting2 12 Setting3 0.1 } SubItems { SubItem { ID 0 Status 0 } SubItem { ID 1 Status 0 } } } Item { ; Structure as above } ; And so on... } -- Bill

1 1

[qi and lambda] Get wrong reaction of the qi::rule if I use lambda for the action
by boost_mailinglist＠gmx.de 16 Aug '11

16 Aug '11

Hello, if I execute the following example, I get the following output "The bad one The good one a" but I only expected "The good one a". Why does it happens? Boost 1.47.0 was used. #include <boost/config/warning_disable.hpp> #include <boost/spirit/include/qi.hpp> #include <boost/lambda/lambda.hpp> #include <iostream> int main() { using boost::spirit::qi::char_; using boost::spirit::qi::parse; char const *first = "{a}", *last = first + std::strlen(first); parse(first, last, '{' >> char_("a")[std::cout << "The good one " << boost::lambda::_1 << '\n'] | char_("b")[std::cout << "The bad one " << boost::lambda::_1 << '\n'] >> '}'); return 0; }

2 3

boost multithread use all cores in parallel ?
by Jack Bryan 16 Aug '11

16 Aug '11

Hi, How to make sure that a multithreaded C++ program is run by ALL cores on a multicore server to do computing in parallel physically (not logically). I have a multicore server, which has 24 CPUs, each of which has 6 cores. It is Intel Xeon X5650 2.67GHz cpu cores : 6 (support 6 threads) cpu MHz : 1596.000 totally, I have 24 * 6 = 144 cores. I designed a multithreaded C++ program with boost/thread. How to make sure that my program is run by all 144 cores ? Any help is really appreciated. thanks

10 12

iterator_adaptor : const and regular form
by John M. Dlugosz 15 Aug '11

15 Aug '11

I wrote an iterator by deriving from the boost::iterator_adaptor template. Now I need the const_iterator for my collection, too. The instructions for iterator_adaptor are not clear, but the implication is that I should transfer what I read from iterator_facade. So, is using a const type for the Value (the 3rd template argument) all it takes to write a proper const_iterator? (assuming the functions I supply are proper, too) So, I can just write my iterator as a template of T, passing T through as Value to the iterator_adaptor base class, and then instantiate _that_ with my value type and with const value type to make iterator and const_iterator respectively, right? —John

1 0

[Range] Why adaptor stores iterator_range instead of its underlying range?
by TONGARI 15 Aug '11

15 Aug '11

Hi, I've had this question for a long time, but I still haven't seen others asked the same question. I wonder what's the rationale for range adaptor holding an iterator_range, instead of storing the underlying range (by reference when the range is a container) and calculating begin() and end() on the fly. For C++11 code, it's not unusual to store the temporary expression aside and use it more than once. E.g. vector<int> v = {1, 2, 3}; auto view = v | filtered(...); do_something(view); v.push_back(4); // X do_something(view); Above code is broken since 'v''s iterator is invalidated after X. Could anybody shed some light on this matter?

1 0

[Boost.Range][Boost.Iterators] - iterator as pointer
by Christoph Heindl 15 Aug '11

15 Aug '11

Hi folks, I wonder if it is possible to determine if an iterator or a boost.range can be handled like a pointer to contiguous memory using an iterator/range trait? I need to enable functions if the given range is convertible to a c-array like pointer range. Best regards, Christoph

3 6

[geometry] weird define in adapted fusion sequence
by alfC 15 Aug '11

15 Aug '11

Hi, Trying (again) to use Boost.Geometry with coordinates with Boost.Units (BTW, what is the status of the compatibility between geometry and units?) I found the following macro definition in boost/ geometry/geometries/adapted/boost_fusion.hpp // Convenience registration macro to bind a Fusion sequence to a CS #define BOOST_GEOMETRY_REGISTER_BOOST_FUSION_CS(CoordinateSystem) \ namespace boost { namespace geometry { namespace traits { \ template <typename Sequence> \ struct coordinate_system \ < \ Sequence, \ typename boost::enable_if \ < \ fusion_adapt_detail::is_fusion_sequence<Sequence> \ >::type \ > \ { typedef cs::cartesian type; }; \ }}} Unless I am missing something, the definition of the macro does not depend on the parameter "CoordinateSystem" !. To make the situation even more confusing the manual's example (http://www.boost.org/doc/ libs/1_47_0/libs/geometry/doc/html/geometry/reference/adapted/ boost_fusion.html) uses the struct as the parameter and not the coordinate system ! Unless I am wrong the last lines of the definition of BOOST_GEOMETRY_REGISTER_BOOST_FUSION_CS should be: { typedef CoordinateSystem type; }; \ }}} and the manual should be updated accordingly. E.g. the example should read BOOST_FUSION_ADAPT_STRUCT(sample_point, (double, x) (double, y) (double, z)) BOOST_GEOMETRY_REGISTER_BOOST_FUSION_CS(cs::cartesian) Am I missing something? Thanks, Alfredo

3 2

Problem building boost\optional\optional.hpp in VS2010
by Pally Sandher 15 Aug '11

15 Aug '11

Hi, We've been using boost v1.36 in Visual Studio 2005 for a few years now. I'm attempting to upgrade our projects & solutions to build in Visual Studio 2010 & thus I've upgraded to boost 1.47.0. However I'm getting an error when trying to compile a project which uses some boost functions Full text of the error message is as follows: 1> ASIMNET.CPP 1>c:\code\inc\boost\optional\optional.hpp(346): error C2061: syntax error : identifier 'm_storage' 1> c:\code\inc\boost\optional\optional.hpp(345) : while compiling class template member function 'void boost::optional_detail::optional_base<T>::construct(const std::locale &)' 1> with 1> [ 1> T=boost::io::detail::locale_t 1> ] 1> c:\code\inc\boost\optional\optional.hpp(500) : see reference to class template instantiation 'boost::optional_detail::optional_base<T>' being compiled 1> with 1> [ 1> T=boost::io::detail::locale_t 1> ] 1> c:\code\inc\boost\format\internals.hpp(56) : see reference to class template instantiation 'boost::optional<T>' being compiled 1> with 1> [ 1> T=boost::io::detail::locale_t 1> ] 1> c:\code\inc\boost\format\internals.hpp(57) : see reference to class template instantiation 'boost::io::detail::stream_format_state<Ch,Tr>' being compiled 1> 1>Build FAILED. 1> The only references to boost in ASIMNET.CPP are as follows: #include <boost/math/special_functions/round.hpp> translation[X] = boost::math::round( (translation[X]/TILEWIDTH) ) * TILEWIDTH; translation[Y] = boost::math::round( (translation[Y]/TILEWIDTH) ) * TILEWIDTH; Any help will be most appreciated. Thanks, Palbinder Sandher Software Deployment Engineer T: +44 (0) 141 945 8500 F: +44 (0) 141 945 8501 http://www.iesve.com **Design, Simulate + Innovate with the <Virtual Environment>** Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456 Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP Email Disclaimer

2 2

boost.clang -- linq like operations in C++
by Julien Nitard 15 Aug '11

15 Aug '11

Hi All, Some (rather long) time ago there were some talks and some code written about LINQ-like feature for C++, notably one named boost.qlang. Were there any followups ? Is there any equivalent in other libraries ? The original thread: http://boost.2283326.n4.nabble.com/C-LINQ-is-cool-What-about-boost-qlang-td… Thanks for any help, Julien

4 3

boost timer
by Kraus Philipp 15 Aug '11

15 Aug '11

Hello, I would like to calculate the runtime of my algorithms. I have written a MPI based algorithms and must determine the duration of the algorithm on each CPU. I would like to show the really duration, which is based on Amdahls law, so I need a "exactly" timer. Can I use the Boost Timer / Boost MPI Timer to calculate the duration between start and end of my algorithm? On large datasets I think the timer is a good choice. Is the resolution of the timer object just enough for small amounts of data? I would like to do this: timer t; run_algorithm; std::cout << t.elapsed << std::endl; Is a timer object a good choice to determine runtime structure or do you a better structure? Thanks Phil

6 8

Compiler warnings when compiling property_tree v1.47.0
by Bill Moo 15 Aug '11

15 Aug '11

Hello Users, I am using the Intel compiler v12.0.1029.2010 on Windows 7 x64 and when compiling the property_tree (v1.47.0) I get lots of warnings reported: infoStruct.cpp ..\..\Includes\boost/mpl/aux_/has_begin.hpp(20): warning #2586: 'test' : decorated name length exceeded, name was truncated BOOST_MPL_HAS_XXX_TRAIT_NAMED_DEF(has_begin, begin, true) This is one of many similar warnings generated and I was wondering if anyone could shed some light on how this can be solved. The code compiles of course and the library builds but I'd rather not have them at all. I would appreciate your comments on this thanks. -- Bill

2 1

[serialization] bad cast for shared library compiled by g++ with visibility=hidden
by Pavol Juhas 15 Aug '11

15 Aug '11

Hello, I am working on a C++ shared library (DLL) that should have a hierarchy of serializable classes using fixed archive format. The serialize template methods of the classes are all instantiated in the DLL, the idea is the library clients should be able to load/save derived objects using shared_ptr<Base>, without an explicit knowledge of the Derived classes. Everything seems to work fine if I build the shared library with g++ using the default visibility option (i.e., every function in the DLL is exported). However, if I compile the code with "-fvisibility=hidden" the test program crashes with bad_cast exception, whenever it tries to serialize a derived object stored as a shared_ptr<Base>. Strangely, if I build the library with NDEBUG defined, the code works fine. The difference seems to be in serialization/smart_cast.hpp which uses different cast code with the NDEBUG macro defined. Any idea why is this happening? Is serialization supported for the shared libraries build with g++ -fvisibility=hidden option? I am using g++ 4.4.3 and the Boost libraries 1.40. Attached is a minimum example where the testlib.hpp and testlib.cpp are the shared library and testprg.cpp is the test program. The code can be built with the attached Makefile as follows: (a) make clean runtest - default g++ visibility, test program works (b) make clean runtest VISIBILITY=1 - builds the library with -fvisibility=hidden, test program fails with the bad_cast exception (c) make clean runtest VISIBILITY=1 NDEBUG=1 - builds the library with -fvisibility=hidden and NDEBUG defined. test program works. Thank you for any pointers, Pavol

1 0

Complier errors using property_tree 5 minute tutorial
by Bill Moo 14 Aug '11

14 Aug '11

Hello Users, I am using boost 1.47.0's property_tree on Windows 7 x64 using the Intel Compiler v 12.0.1029.2010 and having followed the 5 Minute Tutorial to the letter I get errors at complie time. This is my header file : #ifndef INFOSTRUCT_HPP #define INFOSTRUCT_HPP #include <string> struct INFO { unsigned m_Type ; std::string m_Desc ; std::string m_User ; std::string m_Pass ; std::string m_Server ; void load(const std::string & fileName) ; void save(const std::string & fileName) ; } ; #endif // INFOSTRUCT_HPP and this is the .cpp file: #include <boost/property_tree/ptree.hpp> #include <boost/property_tree/xml_parser.hpp> #include "infoStruct.hpp" void INFO::load(const std::string &fileName) { boost::property_tree::ptree pTree ; boost::property_tree::xml_parser::read_xml(fileName, pTree) ; m_Type = pTree.get("Database.Driver", 0) ; m_Pass = pTree.get < std::string>("Database.Pass") ; } void INFO::save(const std::string &fileName) { return ; } But when I compile this I get lots and lots of compiler errors this most significant of those being: ..\..\Includes\boost/property_tree/detail/ptree_implementation.hpp(52): error #501: argument of type "const std::string std::_Pair_base<const std::string, boost::property_tree::basic_ptree<std::string, std::string, std::less<std::string>>>::*" is incompatible with template parameter of type "const std::string std::pair<const std::string, boost::property_tree::basic_ptree<std::string, std::string, std::less<std::string>>>::*" &value_type::first>, ^ detected during: instantiation of class "boost::property_tree::basic_ptree<Key, Data, KeyCompare>::subs [with Key=std::string, Data=std::string, KeyCompare=std::less<std::string>]" at line 182 instantiation of "boost::property_tree::basic_ptree<Key, Data, KeyCompare>::basic_ptree() [with Key=std::string, Data=std::string, KeyCompare=std::less<std::string>]" at line 8 of "..\DataFactory\infoStruct.cpp" With the line in question being : boost::property_tree::ptree pTree ; I honestly cannot see where or what I am doing wrong, so I am hoping someone here can help me. -- Bill

2 4

[mpl][fusion] Issues in the mpl/fusion interface
by Gabriel Redner 14 Aug '11

14 Aug '11

Hi all, I am writing some code which uses MPL and fusion together, and in general I am happy with their interoperability. However, I have stumbled on a few areas in which functionality I expected was missing, or where expressions I expected to work did not. All of these issues have workarounds, but I'd appreciate some input on whether the problem is with the libraries, or with my expectations. 1. mpl::pair and fusion::pair Fusion sequences are conforming MPL sequences, and MPL sequences can be transmuted into fusion sequences. However, the same is not true of their pair classes. I feel that fusion::pair ought to conform to some "MPL pair" concept such that: - I can access its types by the same 'first' and 'second' members as in mpl::pair - mpl::first and mpl::second metafunctions should work against it For instance, I'd like to be able to extract a sequence of key types from a map: // OK! typedef mpl::map<mpl::pair<int, char>, mpl::pair<bool, short> > m1; typedef mpl::transform< m1, mpl::first<mpl::placeholders::_1>, mpl::back_inserter<mpl::vector<> > >::type keys1; // Not OK! typedef fusion::map<fusion::pair<int, char>, fusion::pair<bool, short> > m2; typedef mpl::transform< m2, mpl::first<mpl::placeholders::_1>, mpl::back_inserter<mpl::vector<> > >::type keys2; // OK, but requires different code for mpl and fusion maps typedef mpl::transform< m2, fusion::result_of::first<mpl::placeholders::_1>, mpl::back_inserter<mpl::vector<> > >::type keys3; Also, I feel there should be a canned way to translate mpl::pair into fusion::pair. fusion::result_of::as_pair? 2. mpl::map and fusion::map This code works nicely: typedef mpl::vector<int, char, bool> v1; typedef fusion::result_of::as_vector<v1>::type v2; v2 v(1, 'a', false); However, this does not: typedef mpl::map<mpl::pair<int, char>, mpl::pair<bool, short> > m1; typedef fusion::result_of::as_map<m1>::type m2; m2 m('a', 3); The type 'm2' ends up as fusion::map<mpl::pair<int, char>, mpl::pair<bool, short> >, which clearly is no good. I feel that as_map should automatically translate mpl::pair to fusion::pair so that the result is a usable container. Please let me know if I am barking up the wrong tree, or whether these seem like useful features to anyone else. If there is consensus that they are useful, I would be interested in developing patches for them. Thanks, -Gabe

2 2

Re: [Boost-users] adding elements to a std::container in a
by Martin Elzen 14 Aug '11

14 Aug '11

> Subject: [Boost-users] adding elements to a std::container in a single > expression > Message-ID: <j2787j$bj9$1(a)dough.gmane.org> > Content-Type: text/plain; format=flowed; charset="iso-8859-1"; > reply-type=original > > until now I've usually written my own template to do this: > > > template<typename T> > struct push_back_container:T > { push_back_container &_push_back(const typename T::value_type &_r) > { push_back(_r); > return *this; > } > }; > > one can use this like this: > > push_back_container<std::vector<something> > >()._push_back(something1)._push_back(_something2) > > Is there some more elegant way to do this? Been there, and had the same issues. I've always wondered why std::vector gives you so many ways to assign values to it, but so few to append items, as personally I want to append items much more often than totally change a vector's contents. For what you want to do there is also a standard assign-with-back-inserter whatchamacallit : include vector include algorithm // assign include iterator // back_inserter std::copy( source.begin(), source.end(), std::back_inserter( targetSequenceWithpush_back_memberFn ) );

1 0

Chrono - setting system time
by asif saeed 14 Aug '11

14 Aug '11

Hi, Can Chrono be used to set system time? I have a piece of Windows code using SYSTEMTIME struct for setting all the fields and then calling SetLocalTime(). Is there any facility in boost::chrono to do the same portably? Best regards, Asif

2 1

Issues with brent_find_minima
by Frank Astier 14 Aug '11

14 Aug '11

I must be missing something: the following program doesn't seem to find the right minimum (boost 1.47.0). double f(double x) { return x * cos(x); } int main(int argc, char** argv) { cout << brent_find_minima(f, 3.0, 4.0, 1e-8) << endl; return 0; } I get: 4 -2.61457 But I think the minimum is -3.28837 at x = 3.42562. Did I miss something in calling brent_find_minima?? Also, this function should be called "brent_find_minimum" since it finds a single minimum, IMHO. The singular is minimum in Latin, minima is the plural. Or can the function be called repetitively and each time return a different minimum?? Frank fastier(a)yahoo.com

3 2

[serialization] unregistered class exception when using BOOST_CLASS_EXPORT_GUID
by Ron Hashimshony 14 Aug '11

14 Aug '11

Ron Hashimshony wrote: > Hi all boost users, > We have been using boost.serialization for years. We want to move > from 1.33 to 1.47 and encountered a problem. > Many of our classes are abstract, and the derived are defined with > BOOST_CLASS_EXPORT_GUID. > At one point in time, a new class (B) was added that derives from a > class that was defined with the BOOST_CLASS_EXPORT_GUID macro (A). > It all worked well in 1.33. In 1.47, when we get to de-serialize > class A, we get an exception of unregistered class. > We traced it down, and found out that if we remove the derived class > (B), it > works fine, but this is not an option, and then the derived class (B) > cannot be de-serialized from the file... > If there any way to bypass this problem? was there a change made that > broke this > functionality? > Many Thanks! > Ron > I believe this is a problem introduced in version 140 which is under investigation. Which type of archive are you using? Robert Ramey Hi Robert, We are using XML archives only at the moment. we want to use the portable binaries as well. Any idea when this problem will be solved? Thanks.

1 0