June 2006 - Boost-users - lists.boost.org

still problems with has_xxx and shared_ptr...
by Ovanes Markarian 27 Jun '06

27 Jun '06

Hello all! Sorry to disturb again... I have the following code: namespace detail { BOOST_MPL_HAS_XXX_TRAIT_DEF(element_type) BOOST_MPL_HAS_XXX_TRAIT_DEF(value_type) BOOST_MPL_HAS_XXX_TRAIT_DEF(type) BOOST_MPL_HAS_XXX_TRAIT_DEF(type_t) } struct B1 {}; int main() { typedef boost::shared_ptr<B1> ptr_t; typedef mpl::if_c < detail::has_type<ptr_t>::value , mpl::int_<0> , mpl::if_c < detail::has_element_type<ptr_t>::value , mpl::int_<1> , mpl::if_c < detail::has_value_type<ptr_t>::value , mpl::int_<2> , mpl::if_c < detail::has_type_t<ptr_t>::value , mpl::int_<3> , mpl::int_<-1> > > > > type_id; type_id just_for_debug_an_instance; } MSVC 8.0 debugger shows for just_for_debug_an_instance the following type: boost::mpl::if_c<1,boost::mpl::int_<0>,boost::mpl::if_c<1,boost::mpl::int_<1>,boost::mpl::if_c<1,boost::mpl::int_<2>,boost::mpl::if_c<0,boost::mpl::int_<3>,boost::mpl::int_<-1> > > > > How is it possible, if I really can find in the shared_ptr template class the following typedefs: public: typedef T element_type; typedef T value_type; typedef T * pointer; typedef typename detail::shared_ptr_traits<T>::reference reference; For me it is really unclear why does: detail::has_type<ptr_t>::value has the value 1 in the first static if statement? I looked through the shared_ptr template class and found nowhere the 'type' typedef, function, or anything else, what could be evaluated by has_value. Any help would be really apreciated. Many Thanks, Ovanes Markarian

1 1

Re: [Boost-users] boost 64bit support on windows x64 on amd64 hardware
by Sohail Somani 27 Jun '06

27 Jun '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Chu-Ming Ng > Sent: Tuesday, June 27, 2006 1:42 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] boost 64bit support on windows x64 on > amd64 hardware > > > Hi, > > I am porting some code that uses boost from 32-bit to the > 64-bit platform and > would like to know if the boost library supports this platforms. I don't know about officially, but I haven't had any problems on 64 bit.

1 0

inspecting templates and boost::is_pointer
by Ovanes Markarian 27 Jun '06

27 Jun '06

Hello all, I am trying to specialize boost::is_pointer so, that it can determine if a type is a pointer wrapper (shared_ptr, auto_ptr, weak_ptr). Does anybody has ideas how this can be done? May be it would be possible to handle if e.g. one can determine if a class/struct has some typedef inside: template<...> class shared_ptr { typedef ... element_type; }; class some_other_class { typedef ... type; }; class some_other_class2 { //no typedef... }; Somehow I am not certain if I am on the right path. My problem is that I have a variant container type with different types. One could access a value using -> operator. The operator returns the pointer to the value. But if this value is a pointer (that can be determined by boost::is_pointer type trait) it should return the pointer and not pointer to pointer value. Now what happens if I store shared_ptr or any other type which overload -> operator? This will result in compiler error, because the compiler tries to access member of the pointer to pointer and not of the pointer instance. Is there may be a way to determine if the member access operator is overloaded? What if -> operator is defined in the base? And what if child class inherits base as protected or private? The other solution would be to force every user to write a specialized class is_pointer for every type stored in variant, but this would be not so nice. With Kind Regards, Ovanes Markarian

3 4

Re: [Boost-users] function traits
by Gottlob Frege 27 Jun '06

27 Jun '06

> > From: David Abrahams <dave(a)boost-consulting.com> "Ovanes Markarian" <om_boost(a)keywallet.com> writes: > > > Does anyone know if it possible to determine if a class > > overloads a certain function _where_ the return type is unknown. > > > > An example is the -> operator. It can return any type. My question > > is: is it possible to determine at compile time if some type > > overloads the -> operator, without knowing the return type when this > > operator is overloaded? > > That would only be possible if -> could be a free function, but IIRC > it is required to be a member function... so the answer is no. > > You can do it for any function that can be implemented as a non-member > (see e.g. boost/detail/is_incrementable.hpp). > > Also, if your end goal is still something like is_pointer_like<T>, instead of detecting operator->, you might be able to use the same techniques as is_incrementable to detect operator * (unary) - ie implement is_dereferenceable<T>. Would it be OK to assume is_pointer_like<T> '==' is_dereferenceable<T> ? Tony

1 0

[date_time] extract date from string using a list of patterns
by Jason Dolan 27 Jun '06

27 Jun '06

I'm looking to take a string and convert it to a date. The only problem is the string can be one of many patterns. i.e. ("%Y%m%d", "%Y-%m-%d", "%d/%m/%Y", etc...). It is also possible that the given string will fail all pattern matches, and thus return false. Can anyone give me an example of how to do this? I know you need to use the date_input_facet, but I am VERY new to boost and can't get it to work. Thanks! -Jason

3 5

[serializtion] Problem with multiple archives and Intel compiler
by Bill Lear 27 Jun '06

27 Jun '06

I am having great difficulty in architecting my boost serialization solution with the Intel compiler. The problem is this: I need to support text, xml, and binary archive formats. I allow the user to select which format to save in, by default using the portable text format. If the user wants greater speed, they can choose binary, or if they want to visually inspect the data, they can choose xml. On load, I dynamically detect which archive is used and call the appropriate load routine (text, xml, binary). However, if I include all three archives in the same compilation unit, the default (text) serialization is slowed down dramatically, with the runtimes increasing approximately 50% (from 200 to 300 seconds, and this on a fast 64-bit Opteron box). I tried breaking things apart, putting my archive includes and BOOST_CLASS_EXPORT calls into separate compilation units. I have the following code structure: class AbstractBase { virtual void method() = 0; }; BOOST_IS_ABSTRACT(AbstractBase) class Derived: public AbstractBase { virtual void method() {} }; class A { Def* def; template <class Archvie> void serialize(Archive& ar, const unsigned int version) { ar & def; } }; I break my serialization code into three different files, one for each type of archive, each of which includes a common file that has a templated method for loading a serialized file, along with BOOST_CLASS_EXPORT for the derived class. Here is the text version: #include <boost/archive/text_oarchive.hpp> #include "Serialization.hh" A* BoostTextLoader::load() { return Serialization<text_iarchive>::load(fileName); } #include <boost/serialization/export.hpp> BOOST_CLASS_EXPORT(Derived) Here is the binary: #include <boost/archive/binary_oarchive.hpp> #include "Serialization.hh" A* BoostTextLoader::load() { return Serialization<binary_iarchive>::load(fileName); } #include <boost/serialization/export.hpp> BOOST_CLASS_EXPORT(Derived) and a similar one for XML. The Serialization.hh is roughly this: template<typename Archive> struct FanSerialization { static A* load(const std::string& filename) { A* a = 0; stream<mapped_file_source> ifs(filename); Archive ia(ifs); ia >> BOOST_SERIALIZATION_NVP(a); return a; } } When I compile and try to link a test program, I get errors saying "multiple definition of `boost::archive::detail::guid_initializer<Derived>::instance'" and so on. So, I try removing the BOOST_CLASS export from the binary and XML code modules, leaving it in the text version. Compile and link are now good. The text archive performance is back down to 200 seconds. When I try to serialize binary (or xml) archives, I get. /boost-1_33_1/boost/archive/detail/oserializer.hpp:418: static void boost::archive::detail::save_pointer_type<Archive, TPtr>::polymorphic<T>::save(Archive &, const T &, const boost::archive::detail::basic_pointer_oserializer *) [with T = AbstractBase, Archive = boost::archive::xml_oarchive, TPtr = AbstractBase*]: Assertion `0L != bpos_ptr' failed. The code that is failing is from the assert below (oserializer.hpp): // sice true_type is valid, and this only gets made if the // pointer oserializer object has been created, this should never // fail bpos_ptr = archive_pointer_oserializer<Archive>::find(* true_type); assert(NULL != bpos_ptr); if(NULL == bpos_ptr) boost::throw_exception( archive_exception(archive_exception::unregistered_class) ); ar.save_pointer(vp, bpos_ptr); So, needless to say, I'm a bit confused by what is going on. Is there any way at all to fix this? I have run a profiler on the Intel code and it simply seems to be "doing more" than the non-Intel version (more comparison operators, chiefly). Could there be any way to manually instantiate things, or substitute the containers boost uses for look-ups at run-time with different, separate ones? Thanks. Bill

2 3

[BGL][compressed_sparse_row_graph] add_vertices
by Johan Oudinet 27 Jun '06

27 Jun '06

Hi, I try using the compressed_sparse row_graph graph type, and I want add lots vertices. When I read the documentation (http://www.boost-consulting.com/boost/libs/graph/doc/compressed_sparse_row.…) I saw the appropriate add_vertices function described as bellow: vertex_descriptor add_vertices(vertices_size_type count, compressed_sparse_row_graph& g) But, in compressed_sparse_row_graph.hpp:346: <--8----------------------------------------------------- template<BOOST_CSR_GRAPH_TEMPLATE_PARMS> inline Vertex add_vertices(Vertex count, BOOST_CSR_GRAPH_TYPE& g) { Vertex old_num_verts_plus_one = g.m_rowstart.size(); g.m_rowstart.resize(old_num_verts_plus_one + count, EdgeIndex(0)); return old_num_verts_plus_one - 1; } <--8----------------------------------------------------- I don't understand why `count' is a `Vertex' type! Why not `vertices_size_type' type ? Regards, -- Johan

2 1

[BGL] Betweenness Centrality Clustering
by Rui Carvalho 27 Jun '06

27 Jun '06

Hi, Is the BGL betweenness centrality clustering algorithm an implementation of: Community structure in social and biological networks M. Girvan and M. E. J. Newman Proc. Natl. Acad. Sci. USA 99, 7821-7826 (2002). http://aps.arxiv.org/PS_cache/cond-mat/pdf/0112/0112110.pdf thanks, Rui

2 1

1.33.0 documentation
by Joe McCay 27 Jun '06

27 Jun '06

I do I generate readable documentation for 1.33.0 out of a docbook file or a .fo file? jmccay

2 1

progress display for parallel environment
by Levent Yilmaz 27 Jun '06

27 Jun '06

Hi All, This little utility (progress_display in progress.hpp) is so useful that I couldn't help but send out a feature request (hoping that this list is the right place to do that). For programs using multiple threads or multiple distributed processes (as in MPI) it is useful to output progress indication only from some root node. However, these programs are usually SPMD (single program multiple data) and it is not convenient/possible to simply not create a progress_display object. One solution would be: // create a progress_display that does nothing for non root nodes int total_run = 100; boost::progress_display pg( i_am_root ? total_run : pg::DUMMY_DISPLAY ); // Then operator++() will output only from some root process. for (int i=0; i!=total_run; pg++); So my feature request is, can we have some facilities for creating trivial and does-nothing-useful type of progress_displays? best regards, - Levent.

2 1