December 2008 - Boost - lists.boost.org

New XML library
by Themis Vassiliadis 11 Dec '08

11 Dec '08

Hi, I have been working in a C++ library like Apache Digester (http://commons.apache.org/digester) I'm intending to convert it following boost policies described in Requirements and Guidelines. What are the chances of it become a Boost library ? -- Themis Vassiliadis

9 8

Re: [boost] [flyweight] help needed with Sun Studio (was:[flyweight][sun] internal compiler error processing a Boost.Flyweight test)
by JOAQUIN M. LOPEZ MUÑOZ 10 Dec '08

10 Dec '08

________________________________________ De: boost-bounces(a)lists.boost.org [boost-bounces(a)lists.boost.org] En nombre de Simon Atanasyan [atanasyan(a)gmail.com] Enviado el: miércoles, 10 de diciembre de 2008 21:35 Para: boost(a)lists.boost.org Asunto: Re: [boost] [flyweight] help needed with Sun Studio (was:[flyweight][sun] internal compiler error processing a Boost.Flyweight test) > > On Tue, Dec 9, 2008 at 12:34 PM, Joaquin M Lopez Munoz <joaquin(a)tid.es> wrote: > > I've just filed a bug report, still at the acceptance queue, internal review > > ID 1410119. Thank you very much for your help with this one. Anyway > > I'm not so interested right now in blaming the compiler as in finding a > > suitable workaround. > > Here is the patch for flyweight.hpp. I have checked them by: > CC: Sun C++ 5.9 SunOS_sparc Patch 124863-08 2008/10/16 Simon, you're the man! Thank you very much for this, I've just commited the patch to the trunk, let's see what the regressions tests say about its efectiveness for Sun C++ 5.7 and 5.8. Best regards, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

1 0

[flyweight][sun] internal compiler error processing a Boost.Flyweight test
by JOAQUIN M. LOPEZ MUÑOZ 10 Dec '08

10 Dec '08

I' ve posted this to the testing list to no avail, resending here: Hi, I've recently added Boost.Flyweight to the trunk so that the first testing results are coming out now. Sun Studio (all plaforms) has the following issue when (seemingly) linking test_init: >> Assertion: (../lnk/dbg_dbggen.cc, line 1806) while processing ../libs/flyweight/test/test_init.cpp at line 0. *http://tinyurl.com/588y3e* So I'm basically clueless. I'd be grateful if someone with access to the compiler could investigate this issue a little more. I can provide as much remote support as needed wrt to library aspects. Thank you in advance, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

4 6

[non Boost] Template troubles.
by Robert Jones 10 Dec '08

10 Dec '08

Ok, first off, this is not explicitly about Boost, which I know is against the Boost guidelines, but I hope you'll all forgive me. Indirectly it is through trying to use Boost that i've bumped into this. A little code snippet... #include <cstddef> template < typename T, std :: size_t sz > struct Block { typedef T type[ sz ][ sz ]; }; template < typename T > void assign( T & assignee, const T & value ); template < typename T, std :: size_t sz > void assign( T ( & assignee )[ sz ], const T ( & value )[ sz ] ); int main( ) { Block< int, 16 > :: type a; const Block< int, 16 > :: type b = { }; assign( a, b ); } This compiles under Gcc & Comeau, but MS VisStudio reckons the call to assign in main is ambiguous. Any thoughts? Also, if I put all the template declarations in an anonymous namespace, Comeau fails to compile, while Gcc is still ok! Have we encountered this before, and is there a known work around? Thanks. - Rob.

3 4

[bind/lambda] Matching signature constness?
by Robert Jones 10 Dec '08

10 Dec '08

Here's a bit of code (and my apologies for it's length, but I'm unsure which bit is the failure point) #include <cstddef> #include "boost/lambda/lambda.hpp" #include "boost/lambda/bind.hpp" #include "boost/function.hpp" #include "boost/type_traits.hpp" namespace { template < bool > struct cmp_detail { template < typename T > static bool cmp( const T & lhs, const T & rhs ) { return lhs != rhs; } }; template < > struct cmp_detail< true > { template < typename T, std :: size_t sz > static bool cmp( const T ( & lhs )[ sz ], const T ( & rhs )[ sz ] ) { using boost :: lambda :: _1; using boost :: lambda :: bind; using boost :: lambda :: var; boost :: function< bool( unsigned ) > cmp_element = bind( cmp_detail< boost :: is_array< T > :: value > :: cmp, var( lhs )[ _1 ], var( rhs )[ _1 ] ); for ( unsigned i = 0; i != sz; ++ i ) if ( cmp_element( i ) ) // if ( cmp_detail< boost :: is_array< T > :: value > :: cmp( lhs[ i ], rhs[ i ] ) ) return true; return false; } }; template < typename T > bool cmp( const T & lhs, const T & rhs ) { return cmp_detail< boost :: is_array< T > :: value > :: cmp( lhs, rhs ); } template < typename T, std :: size_t sz > struct Block { typedef T type[ sz ][ sz ]; }; } int main( ) { Block< int, 16 > :: type a; Block< int, 16 > :: type b; cmp( a, b ); } The problem is the bind call in cmp_detail<true>::cmp(...). Error is this block.cpp: In static member function `static bool <unnamed>::cmp_detail< true>::cmp(const T (&)[sz], const T (&)[sz]) [w ith T = int[16], unsigned int sz = 16u]': block.cpp:45: instantiated from `bool <unnamed>::cmp(const T&, const T&) [with T = int[16][16]]' block.cpp:61: instantiated from here block.cpp:32: error: no matching function for call to `bind(<unknown type>, const boost::lambda::lambda_functor<boost::l ambda::lambda_functor_base<boost::lambda::other_action<boost::lambda::subscript_action>, boost::tuples::tuple<boost::lam bda::lambda_functor<boost::lambda::identity<const int (&)[16][16]> >, boost::lambda::lambda_functor<boost::lambda::place holder<1> >, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boo st::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type> > >, const boost::l ambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::other_action<boost::lambda::subscript_action>, b oost::tuples::tuple<boost::lambda::lambda_functor<boost::lambda::identity<const int (&)[16][16]> >, boost::lambda::lambd a_functor<boost::lambda::placeholder<1> >, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples:: null_type> > >)' I think perhaps the var() calls are not preserving constness quite right? Any help would be welcome. Thanks. - Rob.

2 1

Compile-time autoincrementable values
by Agustín K-ballo Bergé 10 Dec '08

10 Dec '08

Hello everyone, I'm working on a small library to safely work with options made of flags combinations. I'm currently looking for a way to automatically generate contiguous indices, pretty much as an enum does. I was able to achieve this in a very impractical way, which I fear may be based on a compiler bug or an implementation-defined issue. I do not have access to the standard, and I will appreciate if someone with more experience than I could look at the code to confirm my fears. My doubts on standard conformance are mainly - compile time flags: the sizeof a function call expression changes as new overloads (better matches) are added. - sizeof as a default template non-type parameter: (this one really surprised me) when using sizeof as a default parameter it gets reevaluated each time the template is instantiated, so if the sizeof result changes a new type is created. >From there I just built a binary tree of flags, which I use as bits or binary digits. I have implemented it for values of 1, 2 and 3 bits only. The code can be found at the vault at: http://www.boostpro.com/vault/index.php?action=downloadfile&filename=autoin… Please excuse me for the quality of the code and the description of the method, I'm a college student quite far from being a C++ template expert. Agustín K-ballo Bergé.-

2 1

[iterators] zip_iterator doesn't work with GCC's TR1 tuple?
by Dean Michael Berris 10 Dec '08

10 Dec '08

Has anybody successfully used Boost.Iterator's zip_iterator with TR1 tuples from GCC's libstdc++? I keep getting errors like "error: no type named 'head_type' in 'class std::tr1::tuple<..." when using TR1 tuples with the zip_iterator implementation in Boost 1.35.0 -- am I missing something? I have code like this: using namespace std::tr1; vector<tuple<string, int, tuple<string, string, string> > > result; copy( make_zip_iterator( make_tuple( string_vector.begin(), int_vector.begin(), make_zip_iterator( make_tuple( string_vector1.begin(), string_vector2.begin(), string_vector3.begin() ) ) ) ), make_zip_iterator( make_tuple( string_vector.end(), int_vector.end(), make_zip_iterator( make_tuple( string_vector1.end(), string_vector2.end(), string_vector3.end() ) ) ) ), back_inserter(result) ); When I qualify the call to 'make_tuple' as 'boost::make_tuple' I don't see the error mentioned above, but I run into the conversion problem between a boost::tuple and an std::tr1::tuple. Any ideas how I can get around this? -- Dean Michael C. Berris Software Engineer, Friendster, Inc.

5 10

Regression tests report page stuck
by joaquin＠tid.es 10 Dec '08

10 Dec '08

Hello, The regression tests report page seems to be stuck, last update date is 7 Dec 2008 23:04:35: http://www.boost.org/development/tests/trunk/developer/summary.html Best, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

5 4

Re: [boost] [Boost-announce] [review][constrained_value] Review of Constrained Value Library begins today
by Daryle Walker 10 Dec '08

10 Dec '08

On Dec 1, 2008, at 7:59 AM, Jeff Garland wrote: > The review of the Robert Kawulak's Constrained Value library begins > today December 1, 2008, and will end on December 10th -- I will be > the review manager. Please post reviews to the developer list. > > Here's the library synopsis: > > The Boost Constrained Value library contains class templates useful > for creating constrained objects. A simple example is an object > representing an hour of a day, for which only integers from the > range [0, 23] are valid values. > > bounded_int<int, 0, 23>::type hour; > hour = 20; // OK > hour = 26; // exception! > > Behavior in case of assignment of an invalid value can be > customized. The library has a policy-based design to allow for > flexibility in defining constraints and behavior in case of > assignment of invalid values. Policies may be configured at compile- > time for maximum efficiency or may be changeable at runtime if such > dynamic functionality is needed. > > The library can be downloaded from the here: > http://rk.go.pl/f/constrained_value.zip > > The documentation is also available online here: > http://rk.go.pl/r/constrained_value > > --------------------------------------------------- > > Please state in your review, whether you think the library should be > accepted as a Boost library. Additionally, please consider the > following aspects in your review of the library: Yes, please accept it. > > - What is your evaluation of the design? It's OK. > - What is your evaluation of the implementation? Fine. I liked the attempts to minimize extra space per object. I didn't like that there were neither tests nor conventional examples. The file where the Quickbook docs took its code examples was the only *.cpp file provided. Fortunately, it's (intentionally) compilable. > - What is your evaluation of the documentation? It's extensive, both the usage (Quickbook) and code (Doxygen) guides. On the usage guide, I don't think the tutorial's "ignore" type does what you think it does. > - What is your evaluation of the potential usefulness of the library? I find restrained types to be useful. > - Did you try to use the library? With what compiler? Did you > have any > problems? Yes, with Mac OS X 10.4 (PowerPC) using XCode. No problems. > - How much effort did you put into your evaluation? A glance? A quick > reading? In-depth study? A day of reading and trying out the examples. > - Are you knowledgeable about the problem domain? Somewhat -- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com

2 3

Is there a special reason to use inline before template functions?
by Thomas Klimpel 10 Dec '08

10 Dec '08

Hi, in boost, many template functions (but not all) use inline. (Example from boost/accumulators/framework/accumulator_set.hpp that uses inline: template<typename Args> inline accumulator_visitor<Args> const make_accumulator_visitor(Args const &args) { return accumulator_visitor<Args>(args); } Example from boost/algorithm/minmax.hpp hat doesn't use inline: template <typename T> tuple< T const&, T const& > minmax(T const& a, T const& b) { return (b<a) ? make_tuple(cref(b),cref(a)) : make_tuple(cref(a),cref(b)); } ) Is there a special reason why inline is used for template functions? I would expect that current compilers know better when inlining a function provides a benefit than the programmer. So I guess there is a better reason for using the inline keyword before template function than just speculative performance optimization, but I can't figure it out. Can anybody help me? Most libs only prepend inline before template functions, but ublas prepends nearly every member function with the macro "BOOST_UBLAS_INLINE", which expands to "inline". Is there a reason why ublas uses this macro? Regards, Thomas

7 12