- Boost - lists.preview.boost.org

[hs] Quickbook highlighting doesn't work in KDE 4.3
by Andrey Semashev 06 Sep '09

06 Sep '09

I'm not sure it's the right place for this, so sorry if it's off-topic. I'm trying to use Boost.HS* with Kate from KDE 4.3 on Ubuntu 9.04 but it doesn't work. There appears no "boost::hs" group in the Toost->Highlighting menu, neither does it color the quickbook files. Have anyone run into such kind of problems? PS: I successfully used Boost.HS with previous KDE versions (4.1 for sure, not sure about 4.2). This tool really helps to document the library, many thanks to the author. * http://www.drivehq.com/web/matias.capeletto/boost_hs/doc/html/index.html

2 2

Failures for GCC 4.4.1 with gnu++0x
by Beman Dawes 06 Sep '09

06 Sep '09

9 19

[Boost.Function] How about parallel functions?
by Kasra 06 Sep '09

06 Sep '09

Hi, I required to call few functions is separate threads and use the results. However, the code for this looks really ugly, you have to create threads, then join the threads when ever you require the result of the functions. So I have created a boost.parallel_function library that does this seamlessly just like invoking a normal function and it only joins the thread when the value of the function result type is required. Is this library of any interest to any one? Here is the a hello world example: #include <cstdlib> #include <limits> #include <boost/parallel_function.hpp> using namespace boost; int F0() { volatile int x = 0; for(int i=0; i<std::numeric_limits<int>::max()/1024; i++) x ^= std::rand(); return (int)x; } int F2(int y) { volatile int x = 0; for(int i=0; i<std::numeric_limits<int>::max()/512; i++) x ^= std::rand(); return (int)x ^ y; } int main() { typedef boost::parallel_function< void ( ) > f0_type; typedef boost::parallel_function< void (int) > f1_type; f0_type f0( &::F0 ); f1_type f1( &::F1 ); f0_type::result_type r0 = f0(); f1_type::result_type r1 = f1(32); // use r0 and r1 as if they are int r0 += 34; int value = r1 *= r0 + 452; std::cout << "Hello world " << value << std::endl; return 0; } With Best Regards Mr. Kasra Nassiri Department of Mathematics & Computer Science Imperial College London

3 2

Bug in date_time::subsecond_duration
by Jon Spencer 06 Sep '09

06 Sep '09

There is a needless loss of range in the constructor of date_time::subsecond_duration. The argument is multiplied by the resolution of the base_duration, which could potentially overflow, and then divided by frac_of_second. The patch below reduces the amount of overflow and is identical if one resolution is an integer multiple of the other. The same technique is used elsewhere in the file. Also see this posting on the subject (http://lists.boost.org/boost-users/2008/10/41763.php) --- boost/date_time/time_duration.hpp +++ boost/date_time/time_duration.hpp @@ -267,7 +267,7 @@ public: typedef typename base_duration::traits_type traits_type; explicit subsecond_duration(boost::int64_t ss) : - base_duration(0,0,0,ss*traits_type::res_adjust()/frac_of_second) + base_duration(0,0,0,traits_type::res_adjust()<frac_of_second?ss/(frac_of_second/traits_type::res_adjust()):ss*(traits_type::res_adjust()/frac_of_second)) {} };

1 0

Weak pointer to object not managed by a shared_ptr
by Mateusz Loskot 05 Sep '09

05 Sep '09

Hi, One of the smart pointer techniques [1] explains use of weak_ptr pointer to object not managed by a shared_ptr. [1] http://www.boost.org/doc/libs/1_40_0/libs/smart_ptr/sp_techniques.html I'm wondering what would be practical use case of that? Is it a part of any of known idioms or patterns? Also, is it valid to assume that this technique can extend lifetime of X so of the internal integral X::i_ ? Here is a simple test case based on the technique explained in the smartptr docs [1] with my questions marked using XXX tag: /// begin of program /////////////////////////////////////////////// #include <exception> #include <iostream> #include <boost/shared_ptr.hpp> #include <boost/weak_ptr.hpp> using namespace std; using namespace boost; struct null_deleter { void operator()(void const *) const { std::cout << "deleter\n"; } }; class X { public: explicit X(int i): this_(this, null_deleter()), i_(i) {} X(X const & rhs): this_(this, null_deleter()), i_(rhs.i_) {} X & operator=(X const & rhs) { i_ = rhs.i_; return *this; } weak_ptr<X> get_weak_ptr() const { return this_; } int get_i() const { return i_; } private: shared_ptr<X> this_; int i_; }; int main() { try { shared_ptr<X> spx; { X x1(7); // XXX: lifetime is extended here? spx = shared_ptr<X>(x1.get_weak_ptr()); } // XXX: No exception. x1 is still alive, so the pointer valid? cout << spx->get_i() << std::endl; } catch (bad_weak_ptr const& e) { cerr << e.what() << endl; } catch (exception const& e) { cerr << e.what() << endl; } catch (...) { cerr << "unknown error\n"; } return 0; } /// end of program /////////////////////////////////////////////// No exception is thrown and I can It seem that the lifetime of x1, automatic object, has been extended beyond its "natural" scope. Am I correct? However, if I replace retrival of shared_ptr with weak_ptr then of course bad_weak_ptr exception is thrown: shared_ptr<X> spx; weak_ptr<X> wpx; { X x1(7); wpx = x1.get_weak_ptr(); } spx = shared_ptr<X>(wpx); cout << spx->get_i() << std::endl; I'm having problems with understanding of what causes the extention of lifetime of X (so of internal X::i_ object). Is this mechanism similar to binding r-value to "const reference"? I would appreciate if anyone could help me to understand it. References to the C++ standard would be great too. Best regards, -- Mateusz Loskot, http://mateusz.loskot.net Charter Member of OSGeo, http://osgeo.org

7 9

[ptr_container] Questionable strong guarantee of ptr_map::insert()
by Kazutoshi Satoda 05 Sep '09

05 Sep '09

Hi, With current implementation of ptr_map::insert(), if an exception is thrown, the objected pointed by the passed pointer is deleted. Is this really providing strong guarantee as documented? I understand strong guarantee as described here. http://www.boost.org/community/exception_safety.html > The strong guarantee: that the operation has either completed > successfully or thrown an exception, leaving the program state exactly > as it was before the operation started. I think ptr_map::insert() doesn't provide this guarantee because deleting the object is a change of the program state. As more concrete problem, see the attached code. I couldn't find any way to provide strong guarantee for delayed_inserter::insert_ready() using ptr_map. I want ready_value unchanged if an exception is thrown during the insertion. Looking for a solution on a similar situation, I found constructors of shared_ptr. Its documentation clearly says that "If an exception is thrown, delete p is called". And for auto_ptr version, it provides strong guarantee by taking the auto_ptr as non-const reference. I think ptr_map::insert() should follow this manner; document the deletion on exception for raw pointer version, and provide strong guarantee with insert(const key_type& key, std::auto_ptr<U>& x) . The the auto_ptr& version can be a better replacement of the raw pointer version. The required temporary of auto_ptr, instead of key_type, is more expressive (about the purpose of the temporary), and matches the Best Practices shown in the documentation of shared_ptr. http://www.boost.org/libs/smart_ptr/shared_ptr.htm#BestPractices But, introducing the auto_ptr& version seems to cause some change of behavior in client codes using the current pass-by-value version. Is it valid to request (create a ticket) for strong guaranteed (taking auto_ptr&) version while having this compatibility concern? Regards. -- k_satoda #include <memory> #include "boost/ptr_container/ptr_map.hpp" namespace ptr_map_insert_strong_guarantee_demo { class value { public: explicit value(int source); // ... }; class delayed_inserter { public: // ... // This is called from context that knows key and source of value, // but doesn't know the time of insertion. void ready_new(int key, int source) { // OK. // This implementation offers strong guarantee for this call. std::auto_ptr<value> new_value(new value(source)); ready_key = key; ready_value = new_value; } // This is called from context which doesn't know key and value, // but knows the time of insertion. void insert_ready() { if (ready_value.get() != 0) { // I want to offer strong guarantee for this call, // but boost::ptr_map deosn't provide the way to do that; // *** ready_value is deleted when an exception is thrown. *** map.insert(ready_key, ready_value); } } // ... private: int ready_key; std::auto_ptr<value> ready_value; boost::ptr_map<int, value> map; // ... }; }

8 17

[lockfree] how many items are queued?
by Oliver Kowalke 05 Sep '09

05 Sep '09

lockfree.fifo provides the info if the queue is empty or not. is it possible to provide a function std::size_t size() const - returning how many items are queued at a specific timepoint? regards, Oliver -- Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3 - sicherer, schneller und einfacher! http://portal.gmx.net/de/go/chbrowser

5 9

PropertyTree automated tests
by Sebastian Redl 04 Sep '09

04 Sep '09

Hi, There are no test results at all for the proptree library on trunk. There isn't even a row in the table for it. Why is that so? Sebastian

2 2

wait-free fast-pathed, bounded-time, 100% starvation-free rw-mutex...
by Chris M. Thomasson 04 Sep '09

04 Sep '09

I was wondering if the following invention of mine might be of use to Boost: http://software.intel.com/en-us/forums/intel-threading-building-blocks/topi… (please __read_all__; thank you) Here are some performance numbers wrt my algorithm versus NPTL on Linux (fedora 10), and OpenSolaris: http://groups.google.com/group/comp.programming.threads/msg/3e7531657a11baf0 My invention has some very interesting properties that simply do not exist for 99% of all the other rw-mutex algorithms I have seen to date. It's 100% starvation free and has bounded wait time for readers and writers. Heck, even the Microsoft rw-mutex algorithm in Vista does not come close to that: http://msdn.microsoft.com/en-us/library/aa904937(VS.85).aspx They clearly state that their algorithm is not fair. The algorithm I have observed in OpenSolaris does not meet this goal as well. Another nice property is that the algorithm has wait-free fast-paths. If readers hit a fast-path, then they are going to get wait-free access to the protected data-structure. The same goes for writers. The underlying operation is a single atomic fetch-and-add operation: http://software.intel.com/en-us/forums/intel-threading-building-blocks/topi… In fact, the algorithm only needs fetch-and-add; no stupid CAS-loops are required. Please take not that there are absolutely no loops in the algorithm; this is one reason why it's bounded time and starvation free. Also, it has conditional reader-to-writer upgrades and non-conditional writer-to-reader downgrade capabilities. This is something that does not exist in the Microsoft implementation, or even the OpenSolaris, or basically any Linux implementation I have observed. Same goes for OSX. Therefore, I think this particular algorithm might be of use to Boost. Any thoughts? Thank you all for you're valuable time; I really do appreciate it! :^)

4 14

1.40 regex
by Jonathan Brannan 04 Sep '09

04 Sep '09

I have code that comes out clean in 1.39 (as unit-test and under valgrind), but under 1.40 it's not clean in valgrind and it crashes when run normally (malloc(): memory corruption). I'm trying to write a reproducer that is smaller than what I have, but incase something pops out at someone (or can point me where my code might be wrong), here is the valgrind output: Running 16 test cases... ==25074== Invalid write of size 8 ==25074== at 0x4E7D4A8: boost::shared_ptr<boost::re_detail::named_subexpressions_base<char> >::shared_ptr() (shared_ptr.hpp:182) ==25074== by 0x4E813C1: boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char cons t*, std::string> > > >::match_results(std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > const&) (match_results.hpp:77) ==25074== by 0x5811371: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator <char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::construct_init(boost::basic_regex<char, boost::regex_traits<char, boost::cpp_rege x_traits<char> > > const&, boost::regex_constants::_match_flags) (perl_matcher_common.hpp:73) ==25074== by 0x4E856D2: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator <char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::perl_matcher(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::_ _normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_i terator<char const*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, _ _gnu_cxx::__normal_iterator<char const*, std::string>) (perl_matcher.hpp:366) ==25074== by 0x4E8583A: bool boost::regex_search<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterat or<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char c onst*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, __gnu_cxx::__no rmal_iterator<char const*, std::string>) (regex_search.hpp:55) ==25074== by 0x4E85BEF: boost::regex_token_iterator_implementation<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex _traits<char> > >::init(__gnu_cxx::__normal_iterator<char const*, std::string>) (regex_token_iterator.hpp:109) ==25074== by 0x4E8AC07: boost::regex_token_iterator<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::regex_token_iterator(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::basic_regex<char, boost::regex _traits<char, boost::cpp_regex_traits<char> > > const&, int, boost::regex_constants::_match_flags) (regex_token_iterator.hpp:200) ==25074== by 0x45D99C: application_ut_suite::test_logon::test_method() (application_ut.cpp:139) ==25074== Address 0x6a64588 is 0 bytes after a block of size 56 alloc'd ==25074== at 0x4C2506C: operator new(unsigned long) (in /usr/lib64/valgrind/amd64-linux/vgpreload_memcheck.so) ==25074== by 0x5811362: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator <char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::construct_init(boost::basic_regex<char, boost::regex_traits<char, boost::cpp_rege x_traits<char> > > const&, boost::regex_constants::_match_flags) (perl_matcher_common.hpp:73) ==25074== by 0x4E856D2: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator <char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::perl_matcher(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::_ _normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_i terator<char const*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, _ _gnu_cxx::__normal_iterator<char const*, std::string>) (perl_matcher.hpp:366) ==25074== by 0x4E8583A: bool boost::regex_search<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterat or<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char c onst*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, __gnu_cxx::__no rmal_iterator<char const*, std::string>) (regex_search.hpp:55) ==25074== by 0x4E85BEF: boost::regex_token_iterator_implementation<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex _traits<char> > >::init(__gnu_cxx::__normal_iterator<char const*, std::string>) (regex_token_iterator.hpp:109) ==25074== by 0x4E8AC07: boost::regex_token_iterator<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::regex_token_iterator(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::basic_regex<char, boost::regex _traits<char, boost::cpp_regex_traits<char> > > const&, int, boost::regex_constants::_match_flags) (regex_token_iterator.hpp:200) ==25074== by 0x442E4F: application_ut_suite::(anonymous namespace)::Test_Setup::Test_Setup(std::istream&, unsigned long) (application_ut.cpp:96) ==25074== by 0x45D99C: application_ut_suite::test_logon::test_method() (application_ut.cpp:139) ==25074== by 0x45F620: _ZN20application_ut_suiteL18test_logon_invokerEv (application_ut.cpp:133) ==25074== ==25074== Invalid write of size 8 ==25074== at 0x46CE8A: boost::detail::shared_count::shared_count() (shared_count.hpp:71) ==25074==:regex_traits<char, boost::cpp_regex_traits<char> > > const&, int, boost::regex_constants::_match_flags) (regex_token_iterator.hpp:200) ==25074== by 0x442E4F: application_ut_suite::(anonymous namespace)::Test_Setup::Test_Setup(std::istream&, unsigned long) (application_ut.cpp:96) ==25074== Address 0x6a64590 is 8 bytes after a block of size 56 alloc'd ==25074== at 0x4C2506C: operator new(unsigned long) (in /usr/lib64/valgrind/amd64-linux/vgpreload_memcheck.so) ==25074== by 0x5811362: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::construct_init(boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags) (perl_matcher_common.hpp:73) ==25074== by 0x4E856D2: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::perl_matcher(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::_cxx::__normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, __gnu_cxx::__normal_iterator<char const*, std::string>) (regex_search.hpp:55) ==25074== by 0x4E85BEF: boost::regex_token_iterator_implementation<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::init(__gnu_cxx::__normal_iterator<char const*, std::string>) (regex_token_iterator.hpp:109) ==25074== by 0x4E8AC07: boost::regex_token_iterator<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::regex_token_iterator(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, int, boost::regex_constants::_match_flags) (regex_token_iterator.hpp:200) ==25074== by 0x442E4F: application_ut_suite::(anonymous namespace)::Test_Setup::Test_Setup(std::istream&, unsigned long) (application_ut.cpp:96) ==25074== by 0x45D99C: application_ut_suite::test_logon::test_method() (application_ut.cpp:139) ==250ub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::construct_init(boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags) (perl_matcher_common.hpp:73) ==25074== by 0x4E856D2: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::perl_matcher(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, __gnu_cxx::__normal_iterator<char const*, std::string>) (perl_matcher.hpp:366) ==25074== by 0x4E8583A: bool boost::regex_search<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > >&, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, boost::regex_constants::_match_flags, __gnu_cxx::__normal_iterator<char const*, std::string>) (regex_search.hpp:55) ==25074== by 0x4E85BEF: boost::regex_token_iterator_implementation<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::init(__gnu_cxx::__normal_iterator<char const*, snstants::_match_flags) (regex_token_iterator.hpp:200) ==25074== by 0x442E4F: application_ut_suite::(anonymous namespace)::Test_Setup::Test_Setup(std::istream&, unsigned long) (application_ut.cpp:96) ==25074== by 0x45D99C: application_ut_suite::test_logon::test_method() (application_ut.cpp:139) ==25074== by 0x45F620: _ZN20application_ut_suiteL18test_logon_invokerEv (application_ut.cpp:133) ==25074== Address 0x6a64598 is not stack'd, malloc'd or (recently) free'd **25074** new/new[] failed and should throw an exception, but Valgrind cannot throw exceptions and so is aborting instead. Sorry. ==25074== at 0x4C22E23: (within /usr/lib64/valgrind/amd64-linux/vgpreload_memcheck.so) ==25074== by 0x4C250C8: operator new(unsigned long) (in /usr/lib64/valgrind/amd64-linux/vgpreload_memcheck.so) ==25074== by 0x4E84680: __gnu_cxx::new_allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >::allocate(unsigned long, void const*) (new_allocator.h:92) ==25074== by 0x4E846AF: std::_Vector_base<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> >, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > >::_M_allocate(unsigned long) (stl_vector.h:144) ==25074== by 0x4E8A4AE: std::vector<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> >, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*tor.tcc:395) ==25074== by 0x4E8A61E: std::vector<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> >, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > >::insert(__gnu_cxx::__normal_iterator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> >*, std::vector<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> >, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > > >, unsigned long, boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > const&) (stl_vector.h:792) ==25074== by 0x4E8A73B: boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > > >::set_size(unsigned long, __gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>) (match_results.hpp:399) ==25074== by 0x4E8B1BF: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::find_imp() (perl_matcher_common.hpp:263) ==25074== by 0x5810D0A: boost::re_detail::perl_matcher<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::find() (perl_matcher_common.hpp:229) ==25074== by 0x4E85846: bool boost::regex_search<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<char const*, std::string> > >, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::match_results<__gnu_cxx::__normal_iterator<char const*, std::string>, std::allocator<boost:::regex_traits<char, boost::cpp_regex_traits<char> > >::init(__gnu_cxx::__normal_iterator<char const*, std::string>) (regex_token_iterator.hpp:109) ==25074== by 0x4E8AC07: boost::regex_token_iterator<__gnu_cxx::__normal_iterator<char const*, std::string>, char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::regex_token_iterator(__gnu_cxx::__normal_iterator<char const*, std::string>, __gnu_cxx::__normal_iterator<char const*, std::string>, boost::basic_regex<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > > const&, int, boost::regex_constants::_match_flags) (regex_token_iterator.hpp:200) ==25074== ==25074== ERROR SUMMARY: 3 errors from 3 contexts (suppressed: 3 from 1) ==25074== malloc/free: in use at exit: 64,694 bytes in 500 blocks. ==25074== malloc/free: 590 allocs, 89 frees, 103,110 bytes allocated. ==25074== For counts of detected errors, rerun with: -v ==25074== searching for pointers to 500 not-freed blocks. ==25074== checked 402,856 bytes. ==25074== ==25074== LEAK SUMMARY: ==25074== definitely lost: 0 bytes in 0 blocks. ==25074== possibly lost: 7,480 bytes in 190 blocks. ==25074== still reachable: 57,214 bytes in 310 blocks. ==25074== suppressed: 0 bytes in 0 blocks. ==25074== Rerun with --leak-check=full to see details of leaked memory.

2 2

[1.41.0] Release branch open for merges
by Beman Dawes 04 Sep '09

04 Sep '09

The release branch (https://svn.boost.org/svn/boost/branches/release) is open for merging all stable changes from trunk, including bug fixes and major upgrades to existing libraries. Breaking changes should be coordinated via the mailing list with the developers of the other libraries affected. New libraries may be added, but require permission of a release manager. In other words, except for new libraries, merging changes *does not* need permission from a release manager. I.e. just do it! The release schedule is more aggressive than usual on the front end so that we can maintain our quarterly release schedule: * September 13, 2009: branches/release closed for new libraries and breaking changes to existing libraries. Still open for bug fixes and other routine changes to all libraries. * September 20, 2009: Release managers start QA checks. * October 4, 2009: branches/release closed for major code changes. Still open for serious problem fixes and docs changes. * October 11, 2009: branches/release closed for all library changes except when specific permission given by release manager(s). * October 18, 2009: Beta release target date. Further betas and/or release candidates as feedback dictates. * November 1, 2009: Release ship date. Links relating to this release: * Google Calendar with the release schedule: <http://www.boost.org/development/index.html>. * Release schedule explanation: <http://svn.boost.org/trac/boost/wiki/ReleaseSchedule>. * Release practices for developers: <http://svn.boost.org/trac/boost/wiki/ImprovingPractices>. -- The release managers

3 3

Re: [boost] question about TMP
by DE 04 Sep '09

04 Sep '09

for those who is still interested i thought about types that a library user can actually use and finally came to this decision one can use something like //don't look at syntax, check out the concept matrix<double> m; //general rectangular dense matrix matrix<double, sparse> sp; //rectangular sparse matrix square<double> sq; //square matrix, although one might square<double, symmetric> sym; //use the following types which square<double, upper_triang> ut; //model corresponding concepts square<double, lower_triang> lt; //i.e. symmetric, upper- & lower //triangular matrices square<double, symmetric | sparse> sparse_sym; //also this square<double, lower_triang | sparse> sparse_lt;//might be useful band<double> b; //guess... band<double, symmetric> bs; //why not? //etc... all these types provide strict explicit interfaces e.g. - 'square' provides 'resize(size_t)' member but not 'resize(size_t, size_t)' - 'band' provides useful 'bandwidth()' member and 'diagonal(int)' member which returns diagonal (might be useful for implementing a band solver) -- Pavel

1 0

optional<optional<T>>
by Andrei Alexandrescu 04 Sep '09

04 Sep '09

Hello, I'm defining an "optional" type for D's standard library modeled similarly to Boost.optional. An interesting question came up - should optional<optional<T>> fold itself into optional<T>, or is "double optional" an interesting concept of its own? I thought I'd ask here because by now there's a significant body of experience with optional<T>. I perused the online documentation and the forum and couldn't find information about that specific detail. Thanks, Andrei

14 22

Re: [boost] [ublas] Extending type_traits for use with Boost Units
by Matthias Schabel 04 Sep '09

04 Sep '09

> Am Wednesday 26 August 2009 schrieb Manoj Rajagopalan: >> Hi UBLAS developers, >> >> I have begun using Boost Units which was introduced in boost v >> 1.36. I am working on extending UBLAS to support physically >> dimensioned types and I have provided a short example below. >> >> This can be achieved efficiently if a few extensions to UBLAS >> type_traits are allowed and I was wondering if anyone would object to >> this. An example of such an extension would be the introduction of >> "sqrt_type" as the type for the square root operation. This would be >> the return type for type_traits<T>::type_sqrt(). For arithmetic types >> like float and double the template can be partially specialized so >> that sqrt_type is float or double but for physically dimensioned >> types like quantity<length,double>, sqrt_type can be set to >> quantity<length^(1/2), double>. > > This is a good idea. Could you prepare a patch and a small example > program? > >> I have already adapted most of vector_expression.hpp for use with >> these dimensioned quantities with work-arounds for these extensions >> but this is getting ugly and unwieldy. > > is it possible to deduct the square_type from operator*()? E.g., using > return type deduction (see detail/returntype_deduction and > promote_traits from traits.hpp) You might also want to get in touch with Steven Watanabe. I believe he has done some work toward making UBLAS and Boost.Units interoperable. Matthias

2 1

Re: [boost] [Polygon] What does this error mean?
by Phil Endecott 03 Sep '09

03 Sep '09

Dear Luke (& others), Here's some test code, followed by the errors that I get when I try to compile it. It boils down to some confusion about whether set is a type or a function. Am I doing something that isn't allowed? Thanks for any suggestions. #include <boost/polygon/polygon.hpp> template<typename COORD_T> struct Box { typedef COORD_T coord_t; coord_t x0; coord_t y0; coord_t x1; coord_t y1; }; typedef int screen_coord_t; typedef Box<screen_coord_t> screen_box_t; namespace boost { namespace polygon { template <> struct geometry_concept<screen_box_t> { typedef rectangle_concept type; }; template <> struct rectangle_traits<screen_box_t> { typedef screen_coord_t coordinate_type; typedef std::pair<screen_coord_t,screen_coord_t> interval_type; static inline interval_type get(const screen_box_t& rectangle, orientation_2d orient) { return (orient==HORIZONTAL) ? std::make_pair(rectangle.x0,rectangle.x1) : std::make_pair(rectangle.y0,rectangle.y1); } }; }; }; typedef boost::polygon::polygon_90_set_data<screen_coord_t> markers_t; markers_t markers; void draw_marker(screen_box_t sb) { using namespace boost::polygon::operators; markers |= sb; } /usr/lib/gcc/i486-linux-gnu/4.1.2/../../../../include/c++/4.1.2/bits/stl_set.h: In function ‘typename boost::enable_if<typename boost::polygon::gtl_and_3<boost::polygon::y_r_assign, typename boost::polygon::is_mutable_rectangle_concept<typename boost::polygon::geometry_concept<T>::type>::type, typename boost::polygon::is_rectangle_concept<typename boost::polygon::geometry_concept<T2>::type>::type>::type, rectangle_type_1>::type& boost::polygon::assign(rectangle_type_1&, const rectangle_type_2&) [with rectangle_type_1 = boost::polygon::rectangle_data<int>, rectangle_type_2 = Box<int>]’: /usr/local/src/gtl/boost/polygon/detail/iterator_geometry_to_set.hpp:34: instantiated from ‘boost::polygon::iterator_geometry_to_set<boost::polygon::rectangle_concept, rectangle_type>::iterator_geometry_to_set(const rectangle_type&, boost::polygon::direction_1d, boost::polygon::orientation_2d, bool) [with rectangle_type = Box<int>]’ /usr/local/src/gtl/boost/polygon/polygon_90_set_data.hpp:137: instantiated from ‘void boost::polygon::polygon_90_set_data<T>::insert(const geometry_type&, bool, boost::polygon::orientation_2d) [with geometry_type = Box<int>, T = int]’ /usr/local/src/gtl/boost/polygon/polygon_90_set_data.hpp:113: instantiated from ‘void boost::polygon::polygon_90_set_data<T>::insert(iT, iT, boost::polygon::orientation_2d) [with iT = const Box<int>*, T = int]’ /usr/local/src/gtl/boost/polygon/detail/polygon_90_set_view.hpp:318: instantiated from ‘typename boost::enable_if<typename boost::polygon::gtl_and<boost::polygon::operators::y_ps90_pe, typename boost::polygon::is_polygon_90_set_type<polygon_set_type_2>::type>::type, boost::polygon::polygon_90_set_data<coordinate_type> >::type& boost::polygon::operators::operator+=(boost::polygon::polygon_90_set_data<coordinate_type>&, const geometry_type_2&) [with coordinate_type_1 = int, geometry_type_2 = Box<int>]’ /usr/local/src/gtl/boost/polygon/detail/polygon_90_set_view.hpp:329: instantiated from ‘typename boost::enable_if<typename boost::polygon::gtl_and<boost::polygon::operators::y_ps90_be, typename boost::polygon::is_polygon_90_set_type<polygon_set_type_2>::type>::type, boost::polygon::polygon_90_set_data<coordinate_type> >::type& boost::polygon::operators::operator|=(boost::polygon::polygon_90_set_data<coordinate_type>&, const geometry_type_2&) [with coordinate_type_1 = int, geometry_type_2 = screen_box_t]’ testpoly.cc:41: instantiated from here /usr/lib/gcc/i486-linux-gnu/4.1.2/../../../../include/c++/4.1.2/bits/stl_set.h:71: error: ‘template<class _Key, class _Compare, class _Alloc> class std::set’ is not a function, /usr/local/src/gtl/boost/polygon/point_concept.hpp:77: error: conflict with ‘template<class T, class coordinate_type> void boost::polygon::set(T&, boost::polygon::orientation_2d, coordinate_type, typename boost::enable_if<typename boost::polygon::is_mutable_point_concept<typename boost::polygon::geometry_concept<T>::type>::type, void>::type*)’ /usr/local/src/gtl/boost/polygon/rectangle_concept.hpp:192: error: in call to ‘set’ /usr/local/src/gtl/boost/polygon/detail/iterator_geometry_to_set.hpp:34: instantiated from ‘boost::polygon::iterator_geometry_to_set<boost::polygon::rectangle_concept, rectangle_type>::iterator_geometry_to_set(const rectangle_type&, boost::polygon::direction_1d, boost::polygon::orientation_2d, bool) [with rectangle_type = Box<int>]’ /usr/local/src/gtl/boost/polygon/polygon_90_set_data.hpp:137: instantiated from ‘void boost::polygon::polygon_90_set_data<T>::insert(const geometry_type&, bool, boost::polygon::orientation_2d) [with geometry_type = Box<int>, T = int]’ /usr/local/src/gtl/boost/polygon/polygon_90_set_data.hpp:113: instantiated from ‘void boost::polygon::polygon_90_set_data<T>::insert(iT, iT, boost::polygon::orientation_2d) [with iT = const Box<int>*, T = int]’ /usr/local/src/gtl/boost/polygon/detail/polygon_90_set_view.hpp:318: instantiated from ‘typename boost::enable_if<typename boost::polygon::gtl_and<boost::polygon::operators::y_ps90_pe, typename boost::polygon::is_polygon_90_set_type<polygon_set_type_2>::type>::type, boost::polygon::polygon_90_set_data<coordinate_type> >::type& boost::polygon::operators::operator+=(boost::polygon::polygon_90_set_data<coordinate_type>&, const geometry_type_2&) [with coordinate_type_1 = int, geometry_type_2 = Box<int>]’ /usr/local/src/gtl/boost/polygon/detail/polygon_90_set_view.hpp:329: instantiated from ‘typename boost::enable_if<typename boost::polygon::gtl_and<boost::polygon::operators::y_ps90_be, typename boost::polygon::is_polygon_90_set_type<polygon_set_type_2>::type>::type, boost::polygon::polygon_90_set_data<coordinate_type> >::type& boost::polygon::operators::operator|=(boost::polygon::polygon_90_set_data<coordinate_type>&, const geometry_type_2&) [with coordinate_type_1 = int, geometry_type_2 = screen_box_t]’ testpoly.cc:41: instantiated from here /usr/local/src/gtl/boost/polygon/rectangle_concept.hpp:192: error: no matching function for call to ‘set(boost::polygon::rectangle_data<int>&, boost::polygon::orientation_2d_enum, std::pair<int, int>)’ /usr/local/src/gtl/boost/polygon/detail/iterator_geometry_to_set.hpp:34: instantiated from ‘boost::polygon::iterator_geometry_to_set<boost::polygon::rectangle_concept, rectangle_type>::iterator_geometry_to_set(const rectangle_type&, boost::polygon::direction_1d, boost::polygon::orientation_2d, bool) [with rectangle_type = Box<int>]’ /usr/local/src/gtl/boost/polygon/polygon_90_set_data.hpp:137: instantiated from ‘void boost::polygon::polygon_90_set_data<T>::insert(const geometry_type&, bool, boost::polygon::orientation_2d) [with geometry_type = Box<int>, T = int]’ /usr/local/src/gtl/boost/polygon/polygon_90_set_data.hpp:113: instantiated from ‘void boost::polygon::polygon_90_set_data<T>::insert(iT, iT, boost::polygon::orientation_2d) [with iT = const Box<int>*, T = int]’ /usr/local/src/gtl/boost/polygon/detail/polygon_90_set_view.hpp:318: instantiated from ‘typename boost::enable_if<typename boost::polygon::gtl_and<boost::polygon::operators::y_ps90_pe, typename boost::polygon::is_polygon_90_set_type<polygon_set_type_2>::type>::type, boost::polygon::polygon_90_set_data<coordinate_type> >::type& boost::polygon::operators::operator+=(boost::polygon::polygon_90_set_data<coordinate_type>&, const geometry_type_2&) [with coordinate_type_1 = int, geometry_type_2 = Box<int>]’ /usr/local/src/gtl/boost/polygon/detail/polygon_90_set_view.hpp:329: instantiated from ‘typename boost::enable_if<typename boost::polygon::gtl_and<boost::polygon::operators::y_ps90_be, typename boost::polygon::is_polygon_90_set_type<polygon_set_type_2>::type>::type, boost::polygon::polygon_90_set_data<coordinate_type> >::type& boost::polygon::operators::operator|=(boost::polygon::polygon_90_set_data<coordinate_type>&, const geometry_type_2&) [with coordinate_type_1 = int, geometry_type_2 = screen_box_t]’ testpoly.cc:41: instantiated from here /usr/lib/gcc/i486-linux-gnu/4.1.2/../../../../include/c++/4.1.2/bits/stl_set.h:71: error: ‘template<class _Key, class _Compare, class _Alloc> class std::set’ is not a function, /usr/local/src/gtl/boost/polygon/point_concept.hpp:77: error: conflict with ‘template<class T, class coordinate_type> void boost::polygon::set(T&, boost::polygon::orientation_2d, coordinate_type, typename boost::enable_if<typename boost::polygon::is_mutable_point_concept<typename boost::polygon::geometry_concept<T>::type>::type, void>::type*)’ /usr/local/src/gtl/boost/polygon/rectangle_concept.hpp:193: error: in call to ‘set’ /usr/local/src/gtl/boost/polygon/rectangle_concept.hpp:193: error: no matching function for call to ‘set(boost::polygon::rectangle_data<int>&, boost::polygon::orientation_2d_enum, std::pair<int, int>)’

6 12

[1.41.0] New libraries?
by Beman Dawes 03 Sep '09

03 Sep '09

Are there any new libraries that will be ready for 1.41.0? There is a New Library Checklist that needs to be gone over with the release managers before merging new libraries into branches/release. See http://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#NewL… Thanks, --Beman

2 1

[graph] quickbook reference docs and formatting
by Andrew Sutton 03 Sep '09

03 Sep '09

Jeremiah, Nick, others with opinions.... Two questions: First, I need a second opinion on some formatting alternatives for the quickbook docs w.r.t the documentation of member/non-member functions of the graph classes. These should be mostly buildable - there are some errors about missing links, but they aren't fatal. Let me know if there are issues building them. I got the doc toolchain working a while ago and I'm afraid to mess with it. I may not have the same set up as everybody else. Hopefully I do. The HTML docs give a flat listing with each function separated by a HR element, which is effective (although not especially navigable). Other lib**raries (e.g., Boost.Filesystem) present a similar listing although without the HR's to divide the content. I've tried putting these in a table to improve readability, and it works a little but, but I don't have enough control over the table to prevent auto-wrapping, which can break a function declaration line in weird places. So what's better: a flat listing or a table? Is there a third option? Generating reference docs via Doxygen (like ProgramOptions) probably isn't a viable solution since there's a lot of metaprogramming and atypical class construction going on in the BGL. Second, all of the reference docs for graph classes cite their associated types via the graph_traits and property_map templates. For most of these types, this strikes me as superfluous at best and slightly disingenuous at worst. Traits classes are only required in a generic context, not when working with a concrete type. If your graph class defines a nested type, vertex_iterator, why not use it directly? I feel that the reference documentation for data structures would be better served by focusing on the specifics of the data structure, not its generic interfaces. That's what the concepts are for. Thoughts? Andrew Sutton andrew.n.sutton(a)gmail.com

2 1

[date_time, Boost 1.40.0]
by Christoph 03 Sep '09

03 Sep '09

including boost/date_time/posix_time/posix_time.hpp or boost/date_time/gregorian/gregorian.hpp gives lots of warning for the following line(s): (boost/date_time/gregorian/conversion.hpp; posix_time, tooo) line 44 std::tm datetm = {}; // zero initialization is needed for extension members, like tm_zone I admit I don't understand the comment. Is it possible to fix that (correctly) such that including the header does not produce like 20 warnings in each .cpp file including the header? (I am using linux running gcc 4.1.2) Best regards Christoph

4 5

[serialization] using boost::serialization::collection_size_type and backwards compatibility
by joaquin＠tid.es 03 Sep '09

03 Sep '09

An observant colleague has noticed that Boost.MultiIndex serialization uses std::size_type instead of the adviced boost::serialization::collection_size_type: https://svn.boost.org/trac/boost/ticket/3365 My question is: if I just change std::size_type with boost::serialization::collection_size_type in Boost.MultiIndex serialization code, will backwards compatibility be preserved, i.e. will users be able to load archives saved with the old std::size_type variant? If not, what's the suggested way to handle this issue? Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

3 3

compile-time for-loop
by klaus triendl 03 Sep '09

03 Sep '09

I wonder whether a compile-time for-loop is useful enough so that it would get included in the mpl library; it's very similar to what fold does but it can be used for computations with types other than sequence containers. The loop construct takes a start value, a loop condition and a function calculating the next start value for the next loop recursion and additionally a state together with a binary operation serving as loop body: namespace boost { namespace mpl { template< typename Start, typename UnaryLoopTest, typename BinaryNext , typename State, typename BinaryOp > struct for_: eval_if< // loop as long as the loop condition yields true typename apply< UnaryLoopTest, typename Start::type >::type, // next loop iteration: recursive call for_< // calculate next loop value typename apply< BinaryNext, State, typename Start::type >::type, // pass on UnaryLoopTest and pass on BinaryNext UnaryLoopTest, BinaryNext, // calculate next state typename apply< BinaryOp, State, typename Start::type >::type, // pass on BinaryOp BinaryOp >, // loop termination if loop condition yields false State > {}; } } // That's an example how to use the for-loop: using namespace boost; template<typename T, T N> struct cross_total: mpl::for_< // start mpl::integral_c<T, N>, // loop condition mpl::not_equal_to<mpl::_1, mpl::integral_c<T, 0> >, // calculate next value mpl::divides<mpl::_2, mpl::int_<10> >, // state = current cross sum of already iterated digits mpl::int_<0>, // binary function is loop body: crosstotal += digit mpl::plus< mpl::_1, mpl::modulus<mpl::_2, mpl::int_<10> > > >::type {}; </code> More useful may be to calculate a number's digits for any base, which can be used to calculate the required maximum storage size for a type (I know that there's std::numeric_limits<> but this one would be more general). Klaus Triendl

4 6

[Polygon] performance
by Simonson, Lucanus J 03 Sep '09

03 Sep '09

Performance has become a critical aspect of this review with the revelation that Barend claims significant performance advantage with his recent implementation. I would like to address this issue directly and, unfortunately, I must do so by addressing Barend's implementation and comparing it to my own, which I would have preferred to avoid. I realize that this comparison of the two algorithms may appear confrontational. My intention is to focus on technical issues to allow people to understand Barend's performance claims in the context of what the two algorithms have in common and how they differ. My understanding of Barend's algorithm is based solely on what he has told me, and he refused to tell me enough to allow me to implement it myself, so if I am in error in my description of it I trust he will promptly correct me. First, I do not believe that Barend's implementation is numerically robust when instantiated with float or double. Successfully passing tests is not proof that his algorithm will not generate self intersecting polygons at its output. Specifically, it is my understanding that his algorithm snaps intersection points to the floating point grid at the end of computation when he writes out output polygons. This means that edges may move laterally a distance dependent on the exponent of the floating point value of the result. In floating point coordinates a vertex may lie arbitrarily close to a nearby edge. Even a very small lateral movement of that edge due to loss of precision in the floating point representation has the potential to cross a nearby vertex and interduce self intersection in the output polygons. Since "simple polygons" is a common precondition to many algorithms (including Barend's own) it is simply unacceptable for there to be any chance of this post condition not being satisfied. Note that infinite precision floating point calculations don't solve this problem if you convert the output to regular floating point at some later stage, which is the common way the algorithm's output would be consumed. Second, Barend has stated that his algorithm does the same thing as mine. This is simply not true. Barend's algorithm imposes quite a few preconditions on its inputs (non-self intersecting being just one example) while mine imposes none. The two arguments of a boolean operation in my library may be any number of intersecting and overlapping polygons, Barend requires that all polygons be simple and non-overlapping. My algorithm accepts a broader range of inputs, which makes it more general and more powerful than Barend's algorithm and a better algorithm to use in a generic interface where preconditions become concept requirements for input geometry data. Furthermore, my algorithm has more output modes such as trapezoidization and keyholing as well as the flexibility to perform n-layer operations such as connectivity extraction and map overlay. Barend's algorithm does not. Comparing Barend's intersection algorithm to my own is not an apples to apples comparision, it is an apple to fruit basket which happens to contain an apple comparision. For these two reasons I think we should take Barend's performance claims with a grain of salt. When compared to other libraries that are known to be numerically robust (GPC, CGAL, Geos) my library is sometimes slower, sometimes faster in Barend's own benchmarking. Only Barend's algorithm is always faster in his benchmarking. However, if Barend's library has the potential to produce self intersecting output it is not even in the same class as these others. My benchmark data can be reproduced using the forcgal.cpp file in the sandbox/gtl directory. I provided it when Andreas asked. I admit that my algorithm could be faster. For the benchmarks I performed I estimated it could be up to 10X faster. For Barend's benchmarks with many pointed star polygons which result in O(n^2) pathological performance of my admittedly suboptimal line segment intersection algorithm that number might be higher sill. I would prefer that the focus of the review be the concept interfaces (function declarations) and how the typesystem is implemented so that we can decide whether it is a good model for a generic geometry API. Changes to eg. the line intersection algorithm or internals of polygon_set_data class can be made without breaking code that depends on the library API. If someone implements a true drop in replacement for my polygon intersection algorithm (or just for line segment intersection) and licenses it under the boost license we can drop it in if it is faster and all any user should notice is that the library is faster in a new release of boost. I don't think there are any decisions I made in the design or implementation of the concept type system or library interfaces that prevent or even make difficult such a change to its implementation details. Overall, I don't think we should view this review as an either or Boost.Polygon now or GGL later decision. There is no reason both libraries cannot be accepted in due time. Barend has made quite a bit a progress since boostcon and I am genuinely happy for him for what he has achieved. Thanks, Luke

3 2

Formal Review of Boost.Polygon extended one more week: until September 9, 2009
by Fernando Cacciola 02 Sep '09

02 Sep '09

Dear developers, The review of Boost.Polygon was initially scheduled to end today. However, the discussions are in their best moment only now so the review period was extended by one week, until September 9, 2009 (that is next Wednesday). Best Fernando Cacciola Review Manager

1 0

[pool] patch to make pool support zero sized allocations.
by Steven Watanabe 02 Sep '09

02 Sep '09

AMDG The patch at https://svn.boost.org/trac/boost/attachment/ticket/386/pool.patch allows the pool classes to support zero sized allocation, by returning a null pointer. std::vector sometimes allocates a zero sized block, so this is necessary for pool_allocator. Returning a null pointer is the simplest solution and I'm that the standard allows an allocator to return the null pointer when asked for a block of zero elements. Okay to commit? In Christ, Steven Watanabe

1 1

Re: [boost] [mpl, units] rational and static_rational
by Matthias Schabel 02 Sep '09

02 Sep '09

> I am looking for a template library that can handle compile time > rationals, ideally with a similar interface to mpl::int_ and that > plays well with boost::rational (the dynamic type). > > First I found boost/mpl/math/rational_c.hpp but soon I realized that > it doesn't work with basic operations like mpl::plus, it doesn't even > simplify reducible fractions (i.e. 2/4 -> 1/2), in fact it looks like > it is just the storage of two (static) integers. Is this something > that will be developed in the future or am I using it in the wrong > way? > > Then I found boost::units::static_rational which is more promising, > because it simplifies fractions and works with mpl::plus, but it is > not integrated with boost::rational nor has an interface similar to > that of mpl metatypes. > > Is there a way of extending mpl::math::rational_c with the existing > libraries in boost, or a simple way to make > boost::units::static_rational have an interface similar to mpl and > also have a value type that returns it runtime boost::rational value? I certainly would have no objection to repackaging the static_rational type used in Boost.Units as an MPL type that plays nicely with Boost.Rational. My advice would be to put together a functional implementation and propose it for a mini-review. We could easily typedef units::static_rational -> mpl::rational_ in the units library. Matthias

1 0

[polygon] Overload resolution for operators fails
by Joachim Faulhaber 02 Sep '09

02 Sep '09

Hi Luke and polygonists, trying to set up some law based tests for boost::polygon I faced difficulties with the overload resolution of operators under msvc-9. This is the code sniplet where the problems occur: ----------------------------------------------------------- #include <boost/validate/laws/law.hpp> #include <boost/polygon/polygon.hpp> // Here I map a PolygonSet class to geometry_concepts according to // the docs: #include <boost/itl_xt/custom_polygon.hpp> namespace boost{namespace itl { template <typename Type> class PolygonCommutativity : public Law<PolygonCommutativity<Type>, LOKI_TYPELIST_2(Type,Type), LOKI_TYPELIST_2(Type,Type)> { // a o b == b o a . . . bool holds() { using namespace boost::polygon; using namespace boost::polygon::operators; Type value_a = this->template getInputValue<operand_a>(); Type value_b = this->template getInputValue<operand_b>(); Type left = value_a; left += value_b; // Compiler fails to find the appropriate overload // but // this works self_assignment_boolean_op<Type,Type,0>(left, value_b); // and also this works boost::polygon::operators::operator+=<Type,Type>(left, value_b); // and this boost::polygon::operators::operator+=(left, value_b); . . . return equivalence(left, right); } }; ----------------------------------------------------------- moreover, if my custom PolygonSet conainer implements an operator += this operator is *silently* chosen by the compiler, so the wrong addition function is called! Contrary to the traditional understanding of a concept, that defines *requirements* for a parameter class it can use, we have to take care here, that the client class *does not* implement certain operations. I am afraid this is not a good thing. I hope this is fixable, may be the error is on my side but I can't see it yet. Regards Joachim

2 2