- Boost - lists.boost.org

Bits and Ints: Intention Request
by Murilo Adriano Vasconcelos 15 Jul '10

15 Jul '10

Hi, I am working on GSoC Bits and Ints project wich is a library with functions for integral data types. The motivation for this library is to provide integer algorithms and utilities optimized by doing binary manipulations so that it uses less CPU cycles, memory access and conditional branches. This library contains some classical integer algorithms such as population count, reversing bits, sign extension and bit interleaving. Would be great to Boost to incorporate this library because its functions can be used on a widely types of applications that needs to perform fast operations on integral types. Since one of slowest operations perfomed by the CPU is conditional braching (it involves pipelining, recaching, etc) this library avoid it. For the runtime versions, conditional branch is not used (if, switch, for, while, etc). Other motivation to adopt this library is the integration with Boost.MPL. Almost every function implemented have it MPL and non-MPL static version. For example: // Runtime int pop_count(uintmax_t value); // non-MPL metafunction template <uintmax_t Value> struct static_pop_count {}; // MPL template <typename IC> struct pop_count : integral_c<int, static_pop_count<IC::value>::value> {}; The contents of this library is inside of boost/integer folder and can be viewed in [0]. The static and MPL compatible metafunctions are declared on headers with static_ prefix. The partial documentation can be viewed in [1]. What already is implemented (underlined are out of proposal) - Sign extension from a length of m bits to a length of n - Bit reversal - Bit-interleaving and uninterleaving - Programming utilities for bit and bitmask manipulation - Sign function - Transfer of sign - Clear Least-Significant Nonzero Bit - Round Up or Down to Next Power of 2 - Count Trailing Zeros - Population Count - *Integer Log2* - *Count Leading Zeros* - Compile time MPL compatible abs - Compile time MPL compatible lcm - Compile time MPL compatible gcd - Safe average functions - Swap without a temporary - Templated class is_integral_constant<> My plan is to implement these functions: - Run-time GCD - Run-time LCM - Functions for getting carry bits from addition and the high-half of an integer multiply - Rounding right-shifts - Saturating arithmetic operations - Multi-word shifts - Two-word divide by one-word value - Bounds Checking - Power of 2 Crossing - Find First String of 1-Bits of a Given Length - Incrementing a Reversed Integer - Decoding a “Zero Means 2n” Field - Integer Log Base 10 I want to know wich functions you think that it is interesting to implement in this library. Also, comments and suggestions are appreciated. [0] - http://svn.boost.org/svn/boost/sandbox/SOC/2010/bits_and_ints/boost/integer/ [1] - http://svn.boost.org/svn/boost/sandbox/SOC/2010/bits_and_ints/libs/integer/… References: http://aggregate.org/MAGIC/ http://graphics.stanford.edu/~seander/bithacks.html http://bits.stephan-brumme.com/ http://www.boost.org/doc/libs/1_35_0/libs/integer/index.html Hacker's Delight / Henry S. Warren, Jr Regards, -- Murilo Adriano Vasconcelos http://murilo.wordpress.com

3 5

Is svn.boost.org down?
by Marshall Clow 14 Jul '10

14 Jul '10

I can't get to Trac. I can't check out from SVN. [ 0810 PDT ] -- Marshall

2 1

[uuid] - test fail for gcc-mingw-4.3.3
by Andy Tompkins 14 Jul '10

14 Jul '10

Hi, I'm hoping someone can help me figure out why the following test fails. See http://tinyurl.com/2w796gq It appears that it compiles but there is such little information I don't know what it wrong. Thanks Andy.

2 2

[Docs] Where's the documentation on compare_pointees.hpp?
by Mostafa 14 Jul '10

14 Jul '10

Hi all, Is there any mention in the docs on compare_pointees.hpp? (I couldn't find any.) I came across it was via some convoluted Google searches, after exhausting the boost docs route. -Mostafa

2 2

Suggested change to type_traits/function_traits
by Joseph Garvin 14 Jul '10

14 Jul '10

I only just now subscribed to this list so forgive me if this has been suggested and rejected before. Currently function_traits<F> exposes the argument types for F as individual typedefs, arg1_type, arg2_type, etc. I think this could be made more powerful with two changes: 1. First, change it to be 0 indexed, arg0_type should be first. This is more consistent with the rest of boost and C++ in general. 2. Have an inner arg struct, templated over an integer indicating which argument you want. So instead of: typedef some_type arg0_type; typedef some_other_type arg1_type; //... You would have: template<int i> struct arg template<> struct arg<0> { typedef some_type type; } template<> struct arg<1> { typedef some_other_type type; } // etc. I think this would make function_traits more usable for metaprogramming. As is, to access the different argN_types you have to resort to using boost preprocessor. For backwards compatibility the old interface could co-exist with the new one.

4 3

Boost format hex issue..
by Rob Kramer 14 Jul '10

14 Jul '10

Hi, I'm having an issue with boost::format, in that I can't get it to format a value as hexadecimal if that value is of float type. For example, the little test program below prints: "As int: 0018, 0x0012, as float: 0018, 0x0018" This is as intended, I suppose, since floats can't printed in hex form anyhow. But is there anyway I can get the 'x' format type attribute to interpret the value as integer? Simply casting it to int is not an option, because the way my float value is interpreted should depend only on the format string. I added a hack to my application that fixes it for now, but that doesn't seem very correct: boost::format (..) fmt; double v = ..; if (fmod (v, 1.0) == 0.0) fmt % (int) v; else fmt % v; Cheers! Rob #include <iostream> #include <iomanip> #include <boost/format.hpp> using namespace std; int main() { cout << "As int: " << boost::format ("%1$04d, 0x%1$04x") % 18 << ", as float: " << boost::format ("%1$04d, 0x%1$04x") % (double) 18 << endl; }

1 0

member function traits (was: Concept Traits Library)
by Barend Gehrels 14 Jul '10

14 Jul '10

Hi, Related to this message: John Maddock wrote: >> Even if the library doesn't provides an approach that works to define >> all >> the concepts, it has a lot of useful things. I also suggested to >> Terje at >> least to add the mpl and type traits extensions of the library. > > Agreed - in fact I don't see why we can't just add the new traits > direct to the type_traits library? > > Certainly if someone would like to package these up as direct > additions to type_traits (so I don't have to do much!) then I'd > certainly look favorably on adding them. > and especially related to the function_traits of the type_traits library. Is there any interest in class member function introspection? According to the current documentation, "function_traits is intended to introspect only C++ functions of the form R (), R( A1 ), R ( A1, ... etc. ) and not function pointers or class member functions.". We have, in our work on Boost.Geometry, the need to have the function_traits also working on class member functions. We first had defined a specific meta-function, only for the purpose of concept-checking. Bruno suggested that using function_traits we can avoid that. So we need it to work on member functions. I did some experiments and it is possible to let function_traits checking member functions as well, and it is easy to incorporate into function_traits. It can then be called, for a member function called "apply" of a (template)class called Strategy, like this: typedef typename boost::function_traits<typename BOOST_TYPEOF(&Strategy::apply)>::arg1_type ptype1; for the first argument (in our case a point type, which is then used in the concept-checking using BCCL, to check if it fulfills the point concept and to feed into this same apply method). Adding member function support only needs two relatively simple additions: 1) a specialization for member functions, like this (here only for a const function with 2 parameters) template <typename Class, typename R, typename T1, typename T2> struct function_traits_helper<R (Class::*) (T1, T2) const> { BOOST_STATIC_CONSTANT(unsigned, arity = 2); typedef R result_type; typedef T1 arg1_type; typedef T2 arg2_type; }; 2) the boost::add_pointer call in function_traits should be avoided for member functions, this can be solved by using a meta-function there: template<typename Function> struct function_traits : public detail::function_traits_helper < typename mpl::if_ < boost::is_member_function_pointer<Function>, Function, typename boost::add_pointer<Function>::type >::type > {}; Experiments are here: http://codepad.org/Qw4Uf1aG So I have a prototype, it seems to work fine and can easily be extended and incorporated. Is there more interest in it, and do the type_traits maintainers find this a good idea? Of is it already somewhere else within Boost (or planned)? Regards, Barend

3 7

Required disk space for test runner
by Michael Caisse 13 Jul '10

13 Jul '10

I just noticed my boost test runners are failing because 25GB of space is no longer enough. I could randomly start making the VM drives larger but since that is such a pain in the MS guest OS I am hoping somebody can provide guidance on how much space is currently required. Thanks - michael -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com

2 1

[1.44] Asking for permission to merge Spirit patches
by Hartmut Kaiser 13 Jul '10

13 Jul '10

Hey, I would like to ask for permission to merge a couple of outstanding Spirit patches to the release branch. These improve Spirit's compatibility with gcc 3.4.x, vc7.1 and Intel/Darwin. I had to wait for the tests to cycle before I could merge them. We have people waiting for those patches to be released. https://svn.boost.org/trac/boost/changeset?reponame=&new=63942@trunk%2Fboost &old=63938@trunk%2Fboost https://svn.boost.org/trac/boost/changeset?reponame=&new=63924@trunk%2Fboost &old=63923@trunk%2Fboost https://svn.boost.org/trac/boost/changeset?reponame=&new=63908@trunk%2Fboost &old=63902@trunk%2Fboost This one adds a new (minor) feature, which I missed to merge during the release preparations. It is well tested, shows no regressions on any platforms, and was supposed to be released (the related docs have been merged already): https://svn.boost.org/trac/boost/changeset?reponame=&new=63891@trunk%2Fboost &old=63838@trunk%2Fboost Regards Hartmut --------------- Meet me at BoostCon www.boostcon.com

2 2

[Filesystem] Add features to list all mount paths, determine mount path of a path
by Daniel Trebbien 13 Jul '10

13 Jul '10

John Reynolds came up with a great idea for a possible extension to Boost Filesystem, in response to a question on the Boost Users' List about listing all available drives on Windows. The basic idea is to have the ability to list mounted filesystems. I then suggested that `basic_path` could be extended with a new member (e.g.: `mount_path`) that will determine the mount path of the given path, and add a function to iterate over all mount paths. Before I open a feature request, I wanted to ask some questions: 1. Has this idea been discussed before? 2. Is `mount_path` a good name for this? 3. What would be a good name for the function to iterate over all mount paths? 4. Is the concept of mount paths cross-platform?

6 7