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

[fusion] proposal BOOST_FUSION_ADAPT_FUNCTION, BOOST_FUSION_DEFINE_FUNCTION, etc
by alfC 15 Oct '10

15 Oct '10

Hi, I am using Boost.Fusion lately. Sometimes one has a function with several parameters, int f(double d , int i , std::string s){ ...bla bla ... } the logic of the program says that sometimes that same funcion will be called by generic code and in such case it is better if "f" where defined as taking a fusion::vector as argument. int f(fusion::vector<double, int, std::string>) // another function, could be an overload even since there is only one why to define this adaptor function, e.g. int f(fusion::vector<double, int, std::string> x){ return f(at_c<0>(x), at_c<1>(x), at_c<2>(x));\ } and it is pretty mechanical. I was wondering if it would be a good idea to have, with consistency with the BOOST_FUSION_ADAPT_STRUCT, something called BOOST_FUSION_ADAPT_FUNCTION BOOST_FUSION_ADAPT_FUNCTION( int, f, double, int, std::string ) (or something like that) that defines the second / fusion-compatible version of f. It could be even a BOOST_FUSION_DEFINE_FUNCTION( int, f, (double, d), (int, i), (std::string, s), ( ... code using variables d, i, s or at_key<d_>, at_key<i_>, at_key<s_> ... ) ) that defines simultaneously the raw-C function, f(double, int, string) *and* the fusion friendly version f(vector<...>). I could probably program these macro for specific cases but not a general one since I don't know enough macro syntax to make it work. Do you think it is a good idea or it is already doable with existing Fusion features? Thank you, Alfredo

2 6

[iostreams] changing char_type of bzip2_compressor
by Matthias Vallentin 15 Oct '10

15 Oct '10

I am having trouble with typedef'ing a custom bzip2_compressor that works with unsigned char instead of plain char. This is the pipeline: typedef boost::uint8_t uint8 typedef boost::iostreams::filtering_stream< boost::iostreams::input, uint8> filtering_istream; typedef boost::iostreams::filtering_stream< boost::iostreams::output, uint8> filtering_ostream; typedef boost::iostreams::basic_bzip2_compressor< std::allocator<uint8>> bzip2_compressor; // FIXME: The bzip2_compressor as defined above does not fit in // the pipeline, what is the correct way to typedef it? std::vector<uint8> raw; auto range = boost::make_iterator_range(start, end); // args of type uint8*. filtering_istream in(range); filtering_ostream out; -> out.push(bzip2_compressor()); out.push(boost::iostreams::back_inserter(raw)); boost::iostreams::copy(in, out); The line with the arrow causes the following errors below, presumable due to an incorrect typedef of the basic_bzip2_compressor. I could not find the correct way to typedef it by crawling through the bzip2.hpp header file of the iostreams library. Any ideas? I am using g++ 4.6 with -std=gnu++0x for C++0x support. Matthias Error subset: [..] ../boost/iostreams/chain.hpp:258: error: no matching function for call to 'std::list<boost::iostreams::detail::linked_streambuf<unsigned char, std::char_traits<unsigned char> >*, std::allocator<boost::iostreams::detail::linked_streambuf<unsigned char, std::char_traits<unsigned char> >*> >::push_back(std::auto_ptr<boost::iostreams::stream_buffer<boost::iostreams::basic_bzip2_compressor<std::allocator<unsigned char> >, std::char_traits<unsigned char>, std::allocator<unsigned char>, boost::iostreams::output> >::element_type*)' [..] /opt/local/include/gcc46/c++/bits/streambuf.tcc:89: error: invalid conversion from 'std::basic_streambuf<char, std::char_traits<unsigned char> >::char_type*' to '__gnu_cxx::char_traits<unsigned char>::char_type*' [-fpermissive] /opt/local/include/gcc46/c++/bits/char_traits.h:188: error: initializing argument 1 of 'static __gnu_cxx::char_traits<_CharT>::char_type* __gnu_cxx::char_traits<_CharT>::copy(__gnu_cxx::char_traits<_CharT>::char_type*, const char_type*, std::size_t) [with _CharT = unsigned char, __gnu_cxx::char_traits<_CharT>::char_type = unsigned char, std::size_t = long unsigned int]' [-fpermissive] /opt/local/include/gcc46/c++/bits/streambuf.tcc:89: error: invalid conversion from 'const char_type*' to 'const char_type*' [-fpermissive] /opt/local/include/gcc46/c++/bits/char_traits.h:188: error: initializing argument 2 of 'static __gnu_cxx::char_traits<_CharT>::char_type* __gnu_cxx::char_traits<_CharT>::copy(__gnu_cxx::char_traits<_CharT>::char_type*, const char_type*, std::size_t) [with _CharT = unsigned char, __gnu_cxx::char_traits<_CharT>::char_type = unsigned char, std::size_t = long unsigned int]' [-fpermissive] /opt/local/include/gcc46/c++/bits/streambuf.tcc: In member function 'std::streamsize std::basic_streambuf<_CharT, _Traits>::xsgetn(std::basic_streambuf<_CharT, _Traits>::char_type*, std::streamsize) [with _CharT = char, _Traits = std::char_traits<unsigned char>, std::streamsize = long int, std::basic_streambuf<_CharT, _Traits>::char_type = char]': [..] .._client.cc:170: instantiated from here /opt/local/include/gcc46/c++/bits/streambuf.tcc:55: error: invalid conversion from 'std::basic_streambuf<char, std::char_traits<unsigned char> >::char_type*' to '__gnu_cxx::char_traits<unsigned char>::char_type*' [-fpermissive] /opt/local/include/gcc46/c++/bits/char_traits.h:188: error: initializing argument 1 of 'static __gnu_cxx::char_traits<_CharT>::char_type* __gnu_cxx::char_traits<_CharT>::copy(__gnu_cxx::char_traits<_CharT>::char_type*, const char_type*, std::size_t) [with _CharT = unsigned char, __gnu_cxx::char_traits<_CharT>::char_type = unsigned char, std::size_t = long unsigned int]' [-fpermissive] /opt/local/include/gcc46/c++/bits/streambuf.tcc:55: error: invalid conversion from 'std::basic_streambuf<char, std::char_traits<unsigned char> >::char_type*' to 'const char_type*' [-fpermissive] /opt/local/include/gcc46/c++/bits/char_traits.h:188: error: initializing argument 2 of 'static __gnu_cxx::char_traits<_CharT>::char_type* __gnu_cxx::char_traits<_CharT>::copy(__gnu_cxx::char_traits<_CharT>::char_type*, const char_type*, std::size_t) [with _CharT = unsigned char, __gnu_cxx::char_traits<_CharT>::char_type = unsigned char, std::size_t = long unsigned int]' [-fpermissive]

1 0

math: combinations of i amongst n elements
by Hicham Mouline 15 Oct '10

15 Oct '10

Hi, Is there an available function to generate all the i-tuples among a sequence of size n (runtime) and 1<= i <=n For e.g. in a vector of 3 ints 1 2 and 3 we want to get all of the singletons: 1 2 3 all the pairs 1 2 1 3 2 3 and all the triplets 1 2 3 If not and I were to write one (n known at runtime only), what signature should I use? template <typename iterator> .... combinations(iterator begin, iterator end) { } what return type would hold all of the singletons, pairs, triplets, ... n-tuples Perhaps a better signature is template <typename iterator> .... combinations(iterator begin, iterator end, int i) { } but then the return type is still not easy to define. I'm thinking a 2D multi_array where not all the space is used (dim1 size is 2^n and dim2 size is n), and pass it by ref template <typename iterator> void combinations(multi_array& out, iterator begin, iterator end) { } regards,

3 2

Determine 32 vs 64 architecture
by Ryan McConnehey 15 Oct '10

15 Oct '10

Is there a library that lets me determine on what architecture my code is compiled? I needed to redefine a couple of typedef for fundamental types. Ryan

3 2

build failed for mpi on homogeneous platform
by Alain Miniussi 14 Oct '10

14 Oct '10

Hi, I am trying to build 1.44 on linux systems and the MPI component fails when I uncomment the #define BOOST_MPI_HOMOGENEOUS line in boost/mpi/config.hpp It is failing with both intel 11 and g++ compilers, here is the first error message for gcc 4.4.3: gcc.compile.c++ bin.v2/libs/mpi/build/gcc-4.4.3/release/threading-multi/broadcast.o In file included from ./boost/mpi/packed_iarchive.hpp:26, from ./boost/mpi/communicator.hpp:27, from ./boost/mpi/collectives.hpp:21, from ./boost/mpi/collectives_fwd.hpp:19, from ./boost/mpi/collectives/broadcast.hpp:11, from libs/mpi/src/broadcast.cpp:9: ./boost/mpi/detail/binary_buffer_iprimitive.hpp: In member function ‘void boost::mpi::binary_buffer_iprimitive::load(T&) [with T = boost::archive::version_type]’: ./boost/archive/detail/iserializer.hpp:107: instantiated from ‘static void boost::archive::load_access::load_primitive(Archive&, T&) [with Archive = boost::mpi::packed_iarchive, T = boost::archive::version_type]’ ./boost/archive/detail/iserializer.hpp:356: instantiated from ‘static void boost::archive::detail::load_non_pointer_type<Archive>::load_primitive::invoke(Archive&, T&) [with T = boost::archive::version_type, Archive = boost::mpi::packed_iarchive]’ ./boost/archive/detail/iserializer.hpp:433: instantiated from ‘static void boost::archive::detail::load_non_pointer_type<Archive>::invoke(Archive&, T&) [with T = boost::archive::version_type, Archive = boost::mpi::packed_iarchive]’ ./boost/archive/detail/iserializer.hpp:586: instantiated from ‘void boost::archive::load(Archive&, T&) [with Archive = boost::mpi::packed_iarchive, T = boost::archive::version_type]’ ./boost/archive/detail/common_iarchive.hpp:66: instantiated from ‘void boost::archive::detail::common_iarchive<Archive>::load_override(T&, int) [with T = boost::archive::version_type, Archive = boost::mpi::packed_iarchive]’ ./boost/mpi/packed_iarchive.hpp:101: instantiated from ‘void boost::mpi::packed_iarchive::load_override(T&, int, mpl_::false_) [with T = boost::archive::version_type]’ ./boost/mpi/packed_iarchive.hpp:118: instantiated from ‘void boost::mpi::packed_iarchive::load_override(T&, int) [with T = boost::archive::version_type]’ ./boost/archive/detail/interface_iarchive.hpp:60: instantiated from ‘Archive& boost::archive::detail::interface_iarchive<Archive>::operator>>(T&) [with T = boost::archive::version_type, Archive = boost::mpi::packed_iarchive]’ ./boost/archive/detail/common_iarchive.hpp:45: instantiated from ‘void boost::archive::detail::common_iarchive<Archive>::vload(boost::archive::version_type&) [with Archive = boost::mpi::packed_iarchive]’ libs/mpi/src/broadcast.cpp:151: instantiated from here ./boost/mpi/detail/binary_buffer_iprimitive.hpp:87: error: no matching function for call to ‘assertion_failed(mpl_::failed************ boost::serialization::is_bitwise_serializable<boost::archive::version_type>::************)’ ./boost/mpi/detail/binary_buffer_iprimitive.hpp: In member function ‘void boost::mpi::binary_buffer_iprimitive::load(T&) [with T = boost::archive::object_id_type]’: Regards, -- Alain Miniussi ---- mail: alain.miniussi(a)oca.eu SIT https://sitweb.oca.eu tel: +33 4 92 00 30 09 CRIMSON https://crimson.oca.eu web: http://www.oca.eu/alainm

2 1

[Pool] Problems using Boost.Pool with the Eigen matrix library
by lfrfly＠icqmail.com 14 Oct '10

14 Oct '10

Since Eigen matrices and vectors use SSE and therefore require a particular alignment, Eigen requires that when these types (specifically, the compile-time sized versions) are within a struct or class, that struct be given the EIGEN_MAKE_ALIGNED_OPERATOR_NEW macro. But here's the problem. What if that struct is allocated by a boost::pool_allocator<T>? Is there any way to enforce alignment then? I'm trying it and it just isn't working. I keep getting breakpoints hit inside of Eigen calls, presumably due to unaligned accesses. Has anyone investigated this, and are there any reasonable options? Perhaps a way to force the pool's start to be aligned, and a way to make sure the sizeof() my type as a whole is aligned? (But what about allocator conversion, eg, std::allocate_shared<T> may take an allocator<T> but it converts it internally to an allocator<somewrapper<T>>?)

2 1

bzip2_compressor help
by Bryan S 14 Oct '10

14 Oct '10

I know that I'm doing something stupid but it isn't obvious to me what. I'm trying to write a compressed file using: namespace BI = boost::iostreams; BI::filtering_ostream out; std::fstream mfile("test.gz", std::ios::binary|std::ios::out); out.push(BI::bzip2_compressor()); out.push(mfile); out << "This is a gz file" << endl; but when I do so, the file is always empty. Commenting out the push of the bzip2_compressor works like I'd expect. Any help for a newbie? Thanks B

2 1

Re: [Boost-users] [Boost.Spirit] Synthesised attributes and sub-rules
by Hartmut Kaiser 14 Oct '10

14 Oct '10

> On 14 October 2010 22:36, Hartmut Kaiser <hartmut.kaiser(a)gmail.com> wrote: > > That should be easy, see below. > > > > start = a_rule | b_rule | c_rule; > > > > qi::rule<Iterator, std::vector<int>()> a_rule; > > qi::rule<Iterator, std::vector<int>()> b_rule; > > qi::rule<Iterator, std::vector<int>()> c_rule; > > > I think I tried something like that, but it didn't work. Unless I'm > mistaken, each of those rules will create their own std::vector<int> > instance, so the values will not end up inserted into the vector specified > in the grammar parameter. No, it should work. The vector attribute is propagated to the alternatives, where it is filled. > I was able to eventually achieve my goal by specifying each of the rules > like so: > > qi::rule<Iterator, std::vector<int> &> a_rule; > > and in the main grammar I passed the _val into them. > > start = a_rule(_val) | b_rule(_val) | c_rule(_val) > > Then I referenced the arguments with _r1 in the sub-rules. > > Does that look correct? Is there a better way? That should work as well. This might be a tick better in terms of performance, even. But I'm guessing here, I have not done any measurements. Regards Hartmut --------------- http://boost-spirit.com

1 0

where is lexical_cast.hpp?
by alapex0310＠gmail.com 14 Oct '10

14 Oct '10

Hi, I want to use boost::lexical_cast but cannot find lexical_cast.hpp in the boost tree. I am using VS2008. Has it been moved somewhere? I am trying to use boost 1.44. Thanks a lot in advance, -Al

3 3

[fusion] memory layout of fusion::vector vs boost::tuple/boost::array
by alfC 14 Oct '10

14 Oct '10

Hi, I am trying to convert a boost::array into a fusion::vector. reinterpreting the memory from boost::array. It works for tuples but not fusion::vector, why is that. Is there a way to make it work? ( I was hoping that the memory layout is the same to make conversion from one to the other very easy.) #include <boost/fusion/container.hpp> #include <boost/fusion/include/container.hpp> #include<boost/array.hpp> #include<iostream> #include<boost/tuple/tuple.hpp> using std::clog; using std::endl; int main(){ boost::array<double, 3> ba={{1.0, 2.1, 3.2}}; clog << "ba: " << ba[0] << ", " << ba[1] << ", " << ba[2] << endl; using namespace boost::fusion; vector<double, double, double> bfv( reinterpret_cast<vector<double, double, double>&>(ba) ); //doesn't work as expected clog << "bfv: " << at_c<0>(bfv) << ", "<< at_c<1>(bfv) << ", " << at_c<2>(bfv) << endl; using boost::tuple; tuple<double, double, double> btt(reinterpret_cast<tuple<double, double, double>&>(ba) ); // works! clog << "btt: " << btt.get<0>() << ", "<< (btt).get<1>() << ", " << (btt).get<2>() << endl; return 0; } outputs: ba: 1, 2.1, 3.2 //original data bfv: -9.25597e+61, -2.35344e-185, 4.87507e-270 //not expected btt: 1, 2.1, 3.2 //works! Let me give some context, the reason I am doing this is that I need to interface fusion::vector of boost::quantities (all have the sizeof double) and adimensionalize the list of quantities into an array of doubles and vise-versa In reality, the final goal is to get this to work: boost::array<double, 3> ba={{1.0, 2.1, 3.2}}; vector<quantity<si::length>, quantity<si::time> , quantity<si::mass> > bfv( reinterpret_cast<vector<quantity<si::length>, quantity<si::time> , quantity<si::mass> >&>(ba) ); (or other generic option to go from array<double, N> to vector<quantity<T1>, quantity<T2>, ... quantity<TN> >) but since I can't make it work with simple 'doubles' I am asking why this could be done. Thank you again, Alfredo

9 20