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

[MPL] MPL use cases overview
by Olivier Tournaire 26 Mar '09

26 Mar '09

Hi all, I am currently reading "C++ template metaprogramming" and I am very interested in the MPL provided by boost. However, I am curious on how users integrate this library in their developments. Please feel free to post not only code snippets, but also description of your usages of the MPL ! Thanks, Olivier

4 4

[gil] Unexplainable compiler error
by Robert Dailey 26 Mar '09

26 Mar '09

Below is the compiler error that I'm getting in Visual Studio 2008: C:\IT\work\redsword\third_party\boost\1_38_0\boost/gil/image.hpp(119) : error C2039: '_view' : is not a member of 'boost::gil::image_view<Loc>' with [ Loc=boost::gil::rgba8c_loc_t ] ..\..\..\..\components\libs\renderer\source\Image.cpp(44) : see reference to function template instantiation 'boost::gil::image<Pixel,IsPlanar,Alloc> &boost::gil::image<Pixel,IsPlanar,Alloc>::operator =<boost::gil::image_view<Loc>>(const Img &)' being compiled with [ Pixel=boost::gil::rgba8_pixel_t, IsPlanar=false, Alloc=std::allocator<unsigned char>, Loc=boost::gil::rgba8c_loc_t, Img=boost::gil::image_view<boost::gil::rgba8c_loc_t> ] And the code that it is compiling: #include <boost/gil/image_view.hpp> #include <boost/gil/image.hpp> #include <boost/gil/typedefs.hpp> Image::Image( unsigned char const* buffer, unsigned int width, unsigned int height, unsigned int bitsPerPixel, int pitch ) { // Provide a pitch the exact width of the image if no // explicit pitch was provided. if( !pitch ) { pitch = width * bitsPerPixel; } using namespace boost::gil; m_imageData = interleaved_view( width, height, reinterpret_cast<rgba8_pixel_t const*>( buffer ), pitch ); } I'm still new to boost::gil and it's very hard to learn. Can anyone guide me in the right direction as to how to resolve this compiler error?

2 1

Re: [Boost-users] boost::array and lower_bound
by Hicham Mouline 26 Mar '09

26 Mar '09

I just didn't the see the std::lower_bound version with the comparator. That does what I want, Regards. > -----Original Message----- > From: Hicham Mouline [mailto:hicham@mouline.org] > Sent: 26 March 2009 9:30 > To: 'boost-users(a)lists.boost.org' > Subject: boost::array and lower_bound > > Hello, > I maintain a boost::array< pair< time, value > > B, > sorted by time, where I defined operator< for pair<time, value> > > I wish to get a given entry from the array for a given time y. > > If y matches any entry, value is returned, if y is between 2 entries, an > interpolated value is returned. > > std::lower_bound( &B[0], &B[supSize], ... ) //this doesn't work > > I suppose for this I need to define some other related iterator to pass to > lower_bound, > not boost::array< pair< time, value > >::const_iterator. > > I suppose this is what "Iterator adaptors" are about ? > I have seen the term "views" as well, is this related? > > Regards,

1 0

boost::array and lower_bound
by Hicham Mouline 26 Mar '09

26 Mar '09

Hello, I maintain a boost::array< pair< time, value > > B, sorted by time, where I defined operator< for pair<time, value> I wish to get a given entry from the array for a given time y. If y matches any entry, value is returned, if y is between 2 entries, an interpolated value is returned. std::lower_bound( &B[0], &B[supSize], ... ) //this doesn't work I suppose for this I need to define some other related iterator to pass to lower_bound, not boost::array< pair< time, value > >::const_iterator. I suppose this is what "Iterator adaptors" are about ? I have seen the term "views" as well, is this related? Regards,

1 0

Re: [Boost-users] [Graph] Adjacency list scalability with a Dijks tra algorithm
by Florian.PONROY＠fr.thalesgroup.com 26 Mar '09

26 Mar '09

Hi Juan, Thanks for your answer. I definitly agree with the map overhead, I will try to better understand how the property map works in order to implement your solution. Florian -----Message d'origine----- De : boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org]De la part de Juan Antonio Farré Basurte Envoyé : mercredi 25 mars 2009 19:27 À : boost-users(a)lists.boost.org Objet : Re: [Boost-users] [Graph] Adjacency list scalability with a Dijks tra algorithm Hi, It's as easy as instantiating the adjacency_list template with mapS for the vertex list selector. This is the second template argument. For example, typedef boost::adjacency_list<boost::vecS,boost::mapS> myGraph; Anyway, remember that a map is more expensive than a vector in terms of both space and time (access time to an element is logarithmic). Another option is that you stay with the default vector for the vertex list, and you use a property to store the vertex id. This way, your vertex 4242 would have descriptor 6 and you'd store 4242 as its id-property value. Cheers, Juan > Hi Andrew, > > Would you please tell me which part of the documentation I should rely on to use a std::map as you'd suggested? I should actually admit I'm not yet > familiar with the BGL :-) > > > -----Message d'origine----- > De : boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org]De la part de Andrew Sutton > Envoyé : mercredi 25 mars 2009 17:25 > À : boost-users(a)lists.boost.org > Objet : Re: [Boost-users] [Graph] Adjacency list scalability with a Dijkstra algorithm > > > > > Here is the problem: I would have imagined that, in my case, > boost::num_vertices(graph) would have returned 6, but it actually returns > the higher vertex number + 1, that is... 4243 ! So each of my vector has a > size of 4243 * sizeof(tVertex) bytes... Is there a way to handle that? I > probably misuse the API, but it would be better if the container would be > dimensioned to the actual number of vertices, since I am in a CPU and memory > constraint environment... > > > > You're confusing the notions of "index" and "descriptor". With the vertex set using vecS, the descriptor of a vertex is the same as its index -- > hence the confusion. Therefore, when you call add_edge(4242, 1, ...), the graph thinks you mean the vertex whose descriptor is 4242 and resizes > itself accordingly. > > You have to remember - whatever the examples may indicate - that you should never actually use a literal value and descriptors interchangeably. This > isn't documented very well, either. > > You could use a std::map to associate your own identifiers with vertex descriptors. > > > Andrew Sutton > andrew.n.sutton(a)gmail.com <mailto:andrew.n.sutton@gmail.com> > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

unicode, C++0x, UTF-16
by Brent Arias 26 Mar '09

26 Mar '09

I see that wstring_convert and codecvt_utf8_utf16 are in the current C++0x spec. Are there plans to backfill boost with implementations for these? I'm also curious why the critique of wstring_convert (http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2005/n1750.html) was not incorporated for C++0x. Thoughts? I've also spotted a comment in boost codecvt_utf8_facet saying: // note the following code will generate on some platforms where // wchar_t is defined as UCS2. The warnings are superfluous as // the specialization is never instantitiated with such compilers. Does that mean that boost.codecvt_utf8_facet is capable of acting as a UTF-8 / UCS2 converter on a Windows box, where wchar_t is 2 bytes in size? Brent Arias Software Architect Lumension Security(tm) Tel ~ (602) 692-8088 Fax ~ (480) 970-6323 brent.arias(a)lumension.com<mailto:brent.arias@lumension.com>

2 1

[fusion] complexity of filter_view
by er 26 Mar '09

26 Mar '09

Hi All, If say vector_type has N elements and predicate is true only for one of them. O(N)? typedef filter_view<vector_type const, predicate > view_t O(1)? view_t view(v);

2 1

Steven Watanabe template profiler
by Joel Falcou 26 Mar '09

26 Mar '09

A while back, I claimed having problems using Steven template profiler. I'm ashamed to say that the problem was that gcc is actually localized. So the warning I got in the filtered output was : "./test.hpp:17: attention : ‘int template_profiler::f(int)’ is deprecated (declared at ./template_profiler.hpp:50) " So I guess I have to edit the regex in postprocess ? I am not a regex guru so If Steven walk by, could you give me some directions to fix this on my end. -- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35

2 1

compile-time check that a member function exists
by John Pretz 25 Mar '09

25 Mar '09

Hi, I'm pretty new to boost. I want to write some code with a compile-time check whether a member function exists or not. The code below shows kinda what I need except I'd obviously like to avoid a specialized template for each class that I use. Can anybody point me in the right direction? Thanks, John **************************************** #include <iostream> using namespace std; struct WithMemberGo { void Go() const {cout<<__PRETTY_FUNCTION__<<" going...."<<endl;} }; struct WithoutGo { }; template <class T> void CallGoIfItExists(const T& t){ t.Go(); } // I want to avoid having to specialize this function for every class without the // right member function. template <> void CallGoIfItExists<WithoutGo>(const WithoutGo&){ cout<<"not going anywhere...."<<endl; } int main(){ WithMemberGo with; CallGoIfItExists(with); WithoutGo without; CallGoIfItExists(without); } -- John Pretz -- P-23, MS H803 Los Alamos National Lab Los Alamos, NM 87545 Phone: (505)665-5847 Fax: (505)665-4121

5 10

Re: [Boost-users] [Graph] Adjacency list scalability with a Dijks tra algorithm
by Florian.PONROY＠fr.thalesgroup.com 25 Mar '09

25 Mar '09

Hi Andrew, Would you please tell me which part of the documentation I should rely on to use a std::map as you'd suggested? I should actually admit I'm not yet familiar with the BGL :-) -----Message d'origine----- De : boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org]De la part de Andrew Sutton Envoyé : mercredi 25 mars 2009 17:25 À : boost-users(a)lists.boost.org Objet : Re: [Boost-users] [Graph] Adjacency list scalability with a Dijkstra algorithm Here is the problem: I would have imagined that, in my case, boost::num_vertices(graph) would have returned 6, but it actually returns the higher vertex number + 1, that is... 4243 ! So each of my vector has a size of 4243 * sizeof(tVertex) bytes... Is there a way to handle that? I probably misuse the API, but it would be better if the container would be dimensioned to the actual number of vertices, since I am in a CPU and memory constraint environment... You're confusing the notions of "index" and "descriptor". With the vertex set using vecS, the descriptor of a vertex is the same as its index -- hence the confusion. Therefore, when you call add_edge(4242, 1, ...), the graph thinks you mean the vertex whose descriptor is 4242 and resizes itself accordingly. You have to remember - whatever the examples may indicate - that you should never actually use a literal value and descriptors interchangeably. This isn't documented very well, either. You could use a std::map to associate your own identifiers with vertex descriptors. Andrew Sutton andrew.n.sutton(a)gmail.com <mailto:andrew.n.sutton@gmail.com>

2 1