August 2010 - Boost-users - lists.preview.boost.org

[1.44] Beta 1 available
by Beman Dawes 05 Aug '10

05 Aug '10

Boost release 1.43.0 beta 1 is now available from SourceForge See http://sourceforge.net/projects/boost/files/ There are two new libraries (Meta State Machine and Polygon) in this release, and lots of fixes and improvements to existing libraries. For details of what's in the release, see http://beta.boost.org/users/news/version_1_44_0. Note that the links to files on this web page are for the final release - use the SourceForge link above to get the beta files. Please download the beta, give it a try, and report any problems you encounter. Thanks, -- The Boost Release Team

3 3

Problem
by Luigi Seregni 05 Aug '10

05 Aug '10

Hello, I'm currently developing a windows ce application and I need to use some regular expression. My colleagues suggest me to use boost regex library, but I've experienced some troubles during installation procedures. I'm currently working on Visual Studio 9.0, so I've download the libraries' installer files from this site BoostPro computing. I've followed all the instructions stated in this document: http://www.boost.org/doc/libs/1_43_0/more/getting_started/windows.html#id33, but when I try to compile the example at point 6 (link your program to a boost library) I've obtained this error from compiler: 1>C:\Program Files\boost\boost_1_42\boost/regex/v4/regex_workaround.hpp(46) : fatal error C1083: Cannot open include file: 'locale': No such file or directory I've tried to find a solution but I still haven't found a solution! Can you please help me? Best regards, -- Luigi Seregni

1 0

ptr_vector sort usage
by archie14 05 Aug '10

05 Aug '10

I can't figure out how to use "sort" method in ptr_vector with the function object. Operator < should be defined, but for various reasons I cant define it in the class A, and I have to do the comparison between the A values using some other values. Example: class A { public: int i; }; class Compare { public: Compare (int i) : _i(i){} //this does not compile, but what's the proper way to define // function object for the sort method in ptr_vector? bool operator<(const A& v1, const A& v2) { return (v1.i > v2.i) &&(_i != 1); } private: int _i; }; ptr_vector<A> lst; Compare cmp(3); lst.sort(cmp);

4 7

boost MPI install error
by Jack Bryan 05 Aug '10

05 Aug '10

Dear All: I am trying to install boost on MPI clsuter with MPICH2. After ./bjam install, I compile a test program provided by boost.org . But I got errors: The program is :-----------------------------------------------------------------#include <iostream>#include <boost/mpi/environment.hpp>#include <boost/mpi/communicator.hpp> namespace mpi = boost::mpi; int main(int argc, char* argv[]){ mpi::environment env(argc, argv); mpi::communicator world; std::cout << "I am process " << world.rank() << " of " << world.size() << "." << std::endl; return 0;}-------------------------------------------------------------- Errors: ------------------------------------------------------------------mpic++ -o testp2p test_p2p.cpp -I/home/boost_2010_8_1/boost_1_43_0 -L/home/boost_2010_8_1/boost_1_43_0/stage/lib -lboost_mpi -lboost_serialization/usr/bin/ld: warning: liblam.so.0, needed by /home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so, not found (try using -rpath or -rpath-link)/usr/bin/ld: warning: liblamf77mpi.so.0, needed by /home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so, not found (try using -rpath or -rpath-link)/usr/bin/ld: warning: liblammpi++.so.0, needed by /home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so, not found (try using -rpath or -rpath-link)/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_short'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_unsigned'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `PMPI::Comm::mpi_comm_map'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `op_intercept'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_errors_return'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `PMPI::Intracomm::current_op'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_comm_world'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_unsigned_char'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_char'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `PMPI::Comm::mpi_err_map'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_group_empty'/home/boost_2010_8_1/boost_1_43_0/stage/lib/libboost_mpi.so: undefined reference to `lam_mpi_packed'collect2: ld returned 1 exit status ------------------------------------------------------------------- It seems that it needs LAM/MPI. But, the LAM/MPI has been installed: PATH=/home/openmpi_123/bin:/home/boost_2010_8_1/boost_1_43_0: ....... I used ./bjam --prefix=mypathand then ./bjam and then ./bjam installAny help is appreciated. Any help is appreciated. thanks Jack Aug. 3 2010

3 10

BGL: interrupting breadth_first_search() at a certain distance/predicate?
by Anders Wallin 05 Aug '10

05 Aug '10

Hi all, I have a graph where I start at a certain vertex and want to find the next undiscovered vertex at a maximum distance of 5 or so from the source vertex. Now I am running breadth_first_search() using a record_distances() visitor and I get the distances to all other vertices (say 10 000 of them) in my graph. This makes my algorithm slow, since I only need to know about vertices at a max distance of 5-6, not all of them. Is there a way of interrupting the algorithm when a certain predicate function supplied by me evaluates to true? (e.g. max distance reached, or a new interesting/valid vertex found, etc) thanks, AW

10 23

[distributions]: Inverse Gamma
by Thomas Mang 05 Aug '10

05 Aug '10

Hi, any plans of implementing the inverse gamma distribution as part of the distributions library ? What about multivariate (in particular the multivariate normal and t distributions and Wishart and Dirichlet ) ? thanks, Thomas

4 10

boost serialisation and enable_shared_from_this versus 1.43
by Avi Bahra 04 Aug '10

04 Aug '10

In 1.42, boost serialisation is broken when used with a class that is derived from enable_shared_from_this. This was supposed to fixed in the next release., but the online release notes for 1.43 make no mention of this. Does any one know if it made it into1.43 ? Best regards, Ta, Avi

2 1

Leandro Lima invited you to Dropbox
by Dropbox 04 Aug '10

04 Aug '10

Leandro Lima wants you to use Dropbox to sync and share files online and across computers. Get started here: http://www.dropbox.com/link/45.SygZEI0JRr.SygZEI0JRr/NjI3ODM0OTc5Nw - The Dropbox Team ____________________________________________________ To stop receiving invites from Dropbox, please go to http://www.dropbox.com/bl/207376696615/boost-users%40lists.boost.org

1 0

[GIL] Warnings when compiling items in channel_algorithms.hpp
by Nathan Crookston 04 Aug '10

04 Aug '10

All, While writing some image conversion code using scoped_channel_value I encountered several compiler warnings. The following minimal code demonstrates: Compiled with VC8, command line: "cl simple.cpp /Iboost /W3 /EHsc" ////// BEGIN CODE ///////////////////////////// #include "boost/gil/gil_all.hpp" namespace boost { namespace gil { const boost::uint16_t bits12_max_val = 0xFFF; struct bits12_min { static boost::uint16_t apply() { return 0; } }; struct bits12_max { static boost::uint16_t apply() { return bits12_max_val; } }; typedef boost::gil::scoped_channel_value<boost::uint16_t, bits12_min, bits12_max> bits12; namespace detail { template <> struct unsigned_integral_max_value<bits12> : public mpl::integral_c<uint32_t,bits12_max_val> {}; template <> struct unsigned_integral_num_bits<bits12> : public mpl::int_<12> {}; }//end detail GIL_DEFINE_BASE_TYPEDEFS(12, gray) const boost::uint16_t bits14_max_val = 0x3FFF; struct bits14_min { static boost::uint16_t apply() { return 0; } }; struct bits14_max { static boost::uint16_t apply() { return bits14_max_val; } }; typedef boost::gil::scoped_channel_value<boost::uint16_t, bits14_min, bits14_max> bits14; namespace detail { template <> struct unsigned_integral_max_value<bits14> : public mpl::integral_c<uint32_t,bits14_max_val> {}; template <> struct unsigned_integral_num_bits<bits14> : public mpl::int_<14> {}; }//end detail GIL_DEFINE_BASE_TYPEDEFS(14, gray) } } //end gil, boost using namespace boost::gil; int main() { gray14_image_t img14(2,2); color_converted_view<gray12_pixel_t>(view(img14))(1,0); gray32f_image_t img32f(2,2); color_converted_view<gray12_pixel_t>(view(img32f))(1,0); return 0; } ////// END CODE /////////////////////////////// It produces the following warnings, trimmed to only those I thought relevant: ///////////////// BEGIN Relevant Warnings ////////////////////////////////////// c:\cygwin\home\ncrookston\boost\boost\gil\channel_algorithm.hpp(257) : warning C 4244: 'argument' : conversion from 'const double' to 'boost::uint16_t', possible loss of data c:\cygwin\home\ncrookston\boost\boost\gil\channel_algorithm.hpp(252) : w hile compiling class template member function 'boost::gil::scoped_channel_value< BaseChannelValue,MinVal,MaxVal> boost::gil::detail::channel_converter_unsigned_i ntegral_nondivisible<SrcChannelV,DstChannelV,SrcLessThanDst,CannotFitInInteger>: :operator ()(SrcChannelV) const' with [ BaseChannelValue=boost::uint16_t, MinVal=boost::gil::bits12_min, MaxVal=boost::gil::bits12_max, SrcChannelV=boost::gil::scoped_channel_value<boost::uint16_t,boost:: gil::bits14_min,boost::gil::bits14_max>, DstChannelV=boost::gil::scoped_channel_value<boost::uint16_t,boost:: gil::bits12_min,boost::gil::bits12_max>, SrcLessThanDst=false, CannotFitInInteger=false ] <snip lots of instantiation backtrace> c:\cygwin\home\ncrookston\boost\boost\gil\channel_algorithm.hpp(268) : warning C 4244: 'argument' : conversion from 'float' to 'boost::uint16_t', possible loss o f data c:\cygwin\home\ncrookston\boost\boost\gil\channel_algorithm.hpp(268) : w hile compiling class template member function 'boost::gil::scoped_channel_value< BaseChannelValue,MinVal,MaxVal> boost::gil::channel_converter_unsigned<SrcChanne lV,DstChannelV>::operator ()(boost::gil::bits32f) const' with [ BaseChannelValue=boost::uint16_t, MinVal=boost::gil::bits12_min, MaxVal=boost::gil::bits12_max, SrcChannelV=boost::gil::scoped_channel_value<float,boost::gil::float _zero,boost::gil::float_one>, DstChannelV=boost::gil::scoped_channel_value<boost::uint16_t,boost:: gil::bits12_min,boost::gil::bits12_max> ] <snip lots more instantiation backtrace> /////////////////////// END Relevant Warnings ////////////////////////////////////////////////// Interestingly, g++ doesn't have an issue with the code. The following patch corrects the problem. I'd be happy to create a trac ticket, write some tests, etc. Please advise. ////////////// BEGIN Patch ////////////////////////////////////// Index: channel_algorithm.hpp =================================================================== --- channel_algorithm.hpp (revision 63088) +++ channel_algorithm.hpp (working copy) @@ -243,7 +243,18 @@ } }; +// Determines the base type of the channel. +template <typename Channel> +struct base_channel { typedef Channel type; }; +template <typename BaseChannelValue, typename MinVal, typename MaxVal> +struct base_channel<scoped_channel_value<BaseChannelValue,MinVal,MaxVal> > +: public base_channel<BaseChannelValue> {}; + +template <int NumBits> +struct base_channel<packed_channel_value<NumBits> > +{ typedef typename packed_channel_value<NumBits>::integer_t type; }; + // Both source and destination are unsigned integral channels, // the dst max value is less than (or equal to) the src max value, // and the src max value is not divisible by the dst max value @@ -253,8 +264,8 @@ typedef typename unsigned_integral_max_value<SrcChannelV>::value_type integer_t; static const double div = unsigned_integral_max_value<SrcChannelV>::value / double(unsigned_integral_max_value<DstChannelV>::value); - static const integer_t div2 = integer_t(div/2); - return DstChannelV((src + div2) / div); + static const integer_t div2 = static_cast<integer_t>(div/2); + return static_cast<typename base_channel<DstChannelV>::type>((src + div2) / div); } }; @@ -265,7 +276,8 @@ ///////////////////////////////////////////////////// template <typename DstChannelV> struct channel_converter_unsigned<bits32f,DstChannelV> : public std::unary_function<bits32f,DstChannelV> { - DstChannelV operator()(bits32f x) const { return DstChannelV(x*channel_traits<DstChannelV>::max_value()+0.5f); } + DstChannelV operator()(bits32f x) const { return + static_cast<typename detail::base_channel<DstChannelV>::type>(x*channel_traits<DstChannelV>::max_value()+0.5f); } }; template <typename SrcChannelV> struct channel_converter_unsigned<SrcChannelV,bits32f> : public std::unary_function<SrcChannelV,bits32f> { //////// END patch //////////////// Thanks, Nate

2 8

matrix function template for triangular_matrix and matrix datatypes
by Max S. Kaznady 04 Aug '10

04 Aug '10

Suppose we are working in namespace boost::numeric::ublas and I have 2 functions 1) template <class T> save_matrix(matrix<T> const &m) 2) template <class T> save_matrix(triangular_matrix<T,lower> const &m) and the code in both overloaded save_matrix functions is identical. How can I write a templated function which accepts any type of a matrix? I tried this, but it didn't work: template <class T> save_matrix(T const &m) Thanks, Max

2 3