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

Spirit : Is possible to build rule step by step?
by Andrea Fontana 17 Jan '11

17 Jan '11

I'm trying to write a small example using spirit. I can write a rule in this way: my_rule = lit("example") | lit("example2"); Also in this way: std::string ex("example") std::string ex2("example2"); my_rule = lit(ex) | lit(ex2)) Is there a way to do this: std::string ex("example") std::string ex2("example2"); my_rule = lit(ex); my_rule = my_rule | lit(ex2)); I need to append strings from an array... Maybe I miss something from docs.

3 2

[Lambda] result_of support bug for placeholders?
by Michel MORIN 17 Jan '11

17 Jan '11

Hi, Let's we have using boost::lambda::_1; typedef boost::remove_reference<decltype(_1)>::type _1_type; Then, decltype(_1(1)) (and std::result_of<_1_type(int)>::type) is 'int const&', but boost::tr1_result_of<_1_type(int)>::type is 'int&". Is this a bug? Regards, Michel

1 0

Customize Message Queue
by derleader mail 16 Jan '11

16 Jan '11

Hi, I need advice how to customize Message Queue in boost / interprocess / ipc / message_queue . hpp > I need to insert additional argument in mq . send (& number , sizeof ( i ), 0 , ); and here mq . receive (& number , sizeof ( number ), recvd_size , priority, ); Its something like 2 messages in one queue. How I can do this? Regards Peter

2 1

Re: [Boost-users] functors created by boost::bind and move semantic
by Marat Abrarov 16 Jan '11

16 Jan '11

To: michi7x7 > > Marat Abrarov: > > Does anybody know why functional objects created by means of > > boost::bind have no move constructors. It would be rather useful > > especially when such functional objects consist of some smart pointers > > (like boost::shared_ptr). > > For example, move constructor of boost::shared_ptr is very light > > (comparing to its copy constructor) - it doesn't use any CAS at all. > > May be someone know the reason of move constructor absence or maybe > > someone know any plans of implementing it. > michi7x7: > I wondered about that too. The reason might be the huge ammounts of code > that need to be changed within boost::function. > However, I'm sure it will come soon. I mean not boost::function but functors created by boost::bind (boost::bind_t....) - they are rather simple - like tuples, so, I think, their move constructors would be simple too. But it would be great to have move constructors in boost::function too (which is more difficult because of current boost::function implementation). Interesting, what is about std::function (or std::tr1::function) in STL implementations with C++0x support, for example, in STL shipped with MS Visual C++ 2010. Marat Abrarov.

1 0

TTI library in the Boost sandbox
by Edward Diener 16 Jan '11

16 Jan '11

I have uploaded to the Boost sandbox 'tti' directory version 1.0 of the TTI library. The TTI library, which is an abbreviation for the Type Traits Introspection library, allows a programmer to introspect at compile time the inner elements of a C++ type. The introspection process depends on specifying the name of the inner element by different macros for different types of elements, and then using a generated metafunction to determine whether that element exists within the enclosing type. The inner elements which can be introspected are type, class template, member data, member function, static member data, and static member function. The TTI library is based on the type_traits_ext portion of the Concept Traits Library, with improvements and additions, and also lifts functionality, for the purposes of orthogonality, from Boost.MPL regarding introspection of types and templates. The purpose of the library is to provide a consistent set of interfaces for doing compile-time introspection which other template metaprogrammers can use in their code. There is a readme.txt in the top-level sandbox tti directory, for anyone browsing the sandbox for interesting libraries, which replicates what is written here. There is a build.txt file in the doc subdirectory for building the documentation and running the tests. The documentation is also included as part of the sandbox files for those who can not build the docs. There is a list of todo items in the sandbox on which I will be working for the next release, but the library is already feature rich. The library has been tested and works with gcc 4.30, 4.4.0, 4.5.0-1, 4.5.2, and VC++ 8.0, 9.0, and 10.0. It may also work with other compilers. There is a bug in versions of gcc prior to version 4 for which I will attempt to find a workaround if it is deemed important, which prevents it from working with the library. Questions, comments, suggestions, and bug reports are all welcome.

1 0

[Proto] Difficult Arity Check
by Hossein Haeri 16 Jan '11

16 Jan '11

Dear all, I have a syntax like ((_1 >> &f1 >> &f2) || (_2 >> &f3) || (_3 >> &f4 >> &f5 >> &f6)) >> g for which I already have a small grammar which along with my context do their job fine. The above expression is to act like g(f2(f1(x)), f3(y), f6(f5(f4(z)))) once invoked with _1 == x, _2 == y, and _3 == z. I am currently trying to add a safety layer on top of that which can emit a compile-error -- at the appropriate corner of my code -- that can prevent pass of functions with wrong arities. For example, g above needs to be a ternary. Although my context will notice it if g is not a ternary and produce a meaningful error message, that's too late. My poor user needs to be informed right at the instantiation time. So, I thought I'm going to use transforms to get my grammar manage that. To that end, I thought the _state of my grammar needs to be an MPL pair of an mpl::int_ and an mpl::bool_. Below is my algorithm with some plain English explanation. Please would someone help me for I found it utterly formulate that in Proto. 1) when<terminal<_>, mpl::pair<mpl::int_<1>, mpl::true_> > -- so far, so easy. 2) When logical_or<Arity, Arity>, return summation of the two arities, paired with the logical and of success value in the two branches: mpl::pair < mpl::plus < typename _state(EmtnArity(_left))::first, typename _state(EmtnArity(_right))::first >, mpl::and_ < typename _state(EmtnArity(_left))::second, typename _state(EmtnArity(_right))::second > > (This doesn't compile.) 3) When shift_right<Arity, Arity>, return mpl::int_<1> again, but, this time, paired with whether or not the right arity has been passed around logically anded with the success value so far: mpl::pair < mpl::int_<1>, mpl::and_ < typename _state(EmtnArity(_left))::second, is_callable_with_args(***) > > And, I'm totally stuck in the use of is_callable_with_args! Any helps please? Feel free to pass a completely new design too. TIA, --Hossein

2 3

Splitting Tiff RGBA into RGB and alpha data ...
by Corinna Kinchin 15 Jan '11

15 Jan '11

Hi there I have an uncompressed RGBA Tiff file (RGB + alpha channel) which I'm trying to split into two data files, one containing the RGB data, and the other containing the alpha (transparency) data. Here's a code snippet showing my (probably misguided) attempt at doing this: image_read_info< tiff_tag > info = read_image_info(in, tiff_tag()); if ( info._photometric_interpretation == 2 && info._samples_per_pixel == 4 ) { // RGBA rgba8_image_t rgba_img; // Read RGBA image ... read_image(in, rgba_img, tiff_tag()); // Set up planar image (to extract alpha channel) and rgb image // (to extract rgb data) rgba8_planar_image_t dst_planar_image(rgba_img.dimensions()); rgb8_image_t dst_rgb_image(rgba_img.dimensions() ); // Convert pixels to appropriate view copy_and_convert_pixels(view(rgba_img), view(dst_rgb_image)); copy_and_convert_pixels(view(rgba_img), view(dst_planar_image)); // Write out data ofstream out((ofilename+"rbg").c_str(), ios_base::binary), alphaout((ofilename +"alpha").c_str(), ios_base::binary); out.write(reinterpret_cast<const char*>(interleaved_view_get_raw_data(view(dst_rgb_image))), info._width * info._height * 3); alphaout.write(reinterpret_cast<const char*>(planar_view_get_raw_data(view(dst_planar_image), 3)), info._width * info._height); ... The alpha transparency data is ok as far as I can see, but the RGB data is mangled ... presumably I've misunderstood how the copy_and_convert_pixels() works. I'm new to the boost::gil library, and am finding the documentation a little hard to get to grips with, so any hints you could give would be greatly appreciated. Thanks a million in advance! All the best, ... Corinna

2 1

functors created by boost::bind and move semantic
by Marat Abrarov 15 Jan '11

15 Jan '11

Hi, Boost Users. Does anybody know why functional objects created by means of boost::bind have no move constructors. It would be rather useful especially when such functional objects consist of some smart pointers (like boost::shared_ptr). For example, move constructor of boost::shared_ptr is very light (comparing to its copy constructor) - it doesn't use any CAS at all. May be someone know the reason of move constructor absence or maybe someone know any plans of implementing it. Abrarov Marat.

2 1

Call C++ function from second Limux daemon
by derleader mail 15 Jan '11

15 Jan '11

Hi, I'm using boost for interprocess communication using message queue. There are 2 Linux daemons which must exchange data. In daemon A I want to call a C++ function from daemon B and pass some arguments for execution. The function in daemon B must execute the arguments and return the result to daemon A. Is it possible to implement this? Regards Peter

1 0

[Function Types] can_be_called
by Hossein Haeri 14 Jan '11

14 Jan '11

Dear all, May I please ask the people in charge of function_type to address: https://svn.boost.org/trac/boost/ticket/3783 I need it in my code and need to know when it's going to become available. TIA, --Hossein

2 1