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

Re: [Boost-users] [Containers Library Review] The review of the Containers library by Ion Gaztanaga starts today
by Ion Gaztañaga 03 Aug '11

03 Aug '11

El 03/08/2011 14:41, Mathias Gaunard escribió: > I think it might be good for the people not quite ready to hop onto the > C++0x bandwagon yet. > Plus sometimes it's good to be able to rely on a single portable > reference implementation that actually works with stateful allocators > etc., while it is implementation-defined whether the standard one does. Now that the new standard is nearly finished, the idea is to make them fully compliant, backport to C++03 all possible features and also have a library to experiment new proposals (N1953: Upgrading the Interface of Allocators using API Versioning and N2045: Improving STL allocators), enhancement functions, etc. Best, Ion

2 1

[multi-index][foreach] Iterating over a sequence index (worked in 1.46, broke in 1.47)
by Maxime van Noppen 03 Aug '11

03 Aug '11

Hi, I use BOOST_FOREACH to iterate of a BMI with a sequence index. The code used to work with Boost 1.46 and before, but broke with Boost 1.47. Am I doing something wrong ? ------------------------ #include <iostream> # include <boost/multi_index/hashed_index.hpp> # include <boost/multi_index/member.hpp> # include <boost/multi_index/sequenced_index.hpp> # include <boost/multi_index_container.hpp> # include <boost/foreach.hpp> using namespace ::boost::multi_index; typedef multi_index_container< int, indexed_by< sequenced<> > > bmi_t; typedef bmi_t::nth_index<0>::type seq_idx; const seq_idx& ret(const seq_idx& idx) { return idx; } int main() { bmi_t b; seq_idx& seq = b.get<0>(); // Works BOOST_FOREACH (int i, seq) std::cout << i << std::endl; // Doesn't compile BOOST_FOREACH (int i, ret(seq)) std::cout << i << std::endl; } ------------------------ Note that the offending part is the 'ret(seq)', without it, the code compiles. The error with g++ 4.6.1 is: > In file included from bmi.cc:9:0: > /home/yabo/dev/boost/usr/include/boost/multi_index/sequenced_index.hpp: In destructor ‘boost::foreach_detail_::simple_variant<T>::~simple_variant() [with T = boost::multi_index::detail::sequenced_index<boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, boost::mpl::vector0<mpl_::na> >]’: > /home/yabo/dev/boost/usr/include/boost/foreach.hpp:256:8: instantiated from here > /home/yabo/dev/boost/usr/include/boost/multi_index/sequenced_index.hpp:561:3: error: ‘boost::multi_index::detail::sequenced_index<SuperMeta, TagList>::~sequenced_index() [with SuperMeta = boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, TagList = boost::mpl::vector0<mpl_::na>]’ is protected > /home/yabo/dev/boost/usr/include/boost/foreach.hpp:586:13: error: within this context > /home/yabo/dev/boost/usr/include/boost/multi_index/sequenced_index.hpp: In constructor ‘boost::foreach_detail_::simple_variant<T>::simple_variant(const T&) [with T = boost::multi_index::detail::sequenced_index<boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, boost::mpl::vector0<mpl_::na> >]’: > /home/yabo/dev/boost/usr/include/boost/foreach.hpp:648:95: instantiated from ‘boost::foreach_detail_::auto_any<boost::foreach_detail_::simple_variant<T> > boost::foreach_detail_::contain(const T&, bool*) [with T = boost::multi_index::detail::sequenced_index<boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, boost::mpl::vector0<mpl_::na> >]’ > bmi.cc:28:3: instantiated from here > /home/yabo/dev/boost/usr/include/boost/multi_index/sequenced_index.hpp:549:3: error: ‘boost::multi_index::detail::sequenced_index<SuperMeta, TagList>::sequenced_index(const boost::multi_index::detail::sequenced_index<SuperMeta, TagList>&) [with SuperMeta = boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, TagList = boost::mpl::vector0<mpl_::na>]’ is protected > /home/yabo/dev/boost/usr/include/boost/foreach.hpp:571:9: error: within this context > /home/yabo/dev/boost/usr/include/boost/multi_index/sequenced_index.hpp: In copy constructor ‘boost::foreach_detail_::simple_variant<T>::simple_variant(const boost::foreach_detail_::simple_variant<T>&) [with T = boost::multi_index::detail::sequenced_index<boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, boost::mpl::vector0<mpl_::na> >, boost::foreach_detail_::simple_variant<T> = boost::foreach_detail_::simple_variant<boost::multi_index::detail::sequenced_index<boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, boost::mpl::vector0<mpl_::na> > >]’: > /home/yabo/dev/boost/usr/include/boost/foreach.hpp:256:8: instantiated from ‘boost::foreach_detail_::auto_any<boost::foreach_detail_::simple_variant<T> > boost::foreach_detail_::contain(const T&, bool*) [with T = boost::multi_index::detail::sequenced_index<boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, boost::mpl::vector0<mpl_::na> >]’ > bmi.cc:28:3: instantiated from here > /home/yabo/dev/boost/usr/include/boost/multi_index/sequenced_index.hpp:549:3: error: ‘boost::multi_index::detail::sequenced_index<SuperMeta, TagList>::sequenced_index(const boost::multi_index::detail::sequenced_index<SuperMeta, TagList>&) [with SuperMeta = boost::multi_index::detail::nth_layer<1, int, boost::multi_index::indexed_by<boost::multi_index::sequenced<> >, std::allocator<int> >, TagList = boost::mpl::vector0<mpl_::na>]’ is protected > /home/yabo/dev/boost/usr/include/boost/foreach.hpp:578:13: error: within this context Note that with clang++ 2.9 the code doesn't compile at all either with Boost 1.46 or Boost 1.47. Thanks! -- Maxime

6 12

[Concept] Ignoring return types
by Robert Jones 03 Aug '11

03 Aug '11

I have a template that I instantiate with a type that has to be nullary callable, so template <typename NullaryCallable> struct X { /*....*/ }; I'd like to be a good boy scout here, so I should constrain the template parameter, so template <typename NullaryCallablle> struct X { BOOST_CONCEPT_ASSERT(( Generator<NullaryCallable, void> )); // ... }; However, that constrains it to be nullary callable and returning void, and I'd prefer to avoid the void condition. Is all that sensible, true, and fixable? Thx - Rob.

1 0

#define BOOST_ASSERT_MSG_OSTREAM std::cout - message is not displayed
by asif saeed 03 Aug '11

03 Aug '11

Hi, After I have #defined BOOST_ASSERT_MSG_OSTREAM std::cout, using BOOST_ASSERT_MSG with a customized message does not display the message and, instead, I get the following message from Visual C++ 2010 Express: R6010 Abort() has been called What can I do to display the customized message? Best, Asif

1 0

Boost Book - comments?
by Robert Jones 03 Aug '11

03 Aug '11

Morning All Does anyone have any knowledge of this book? http://www.amazon.co.uk/Boost-C-Libraries/dp/3639942310/ref=sr_1_7?s=books&… <http://www.amazon.co.uk/Boost-C-Libraries/dp/3639942310/ref=sr_1_7?s=books&…>It's genesis, currency, provenance, and possibly a few review-like comments? Thx - Rob.

2 2

Re: [Boost-users] boost.regex with ICU memory cleanup on windows
by Alex Perry 03 Aug '11

03 Aug '11

John Maddock wrote on Tue, 2 Aug 2011 12:14:53 +0100 > There are a number of issues here: > > 1) Your call to u_cleanup() occurs while a u32regex object is still in existence, > so any ICU data being cached is actually still being used at that point. I couldn't get u_cleanup() to work anywhere - ie BOOST_AUTO_TEST_CASE( u32_regex ) { { boost::u32regex re( boost::make_u32regex("\\d{3}")); } u_cleanup(); } Or having u_cleanup() in a fixture teardown (which is how it was in the original code) doesn't seem to make any difference > 2) Boost.Regex does cache some internal data for improved performance - > this is relatively small and gets cleared after main() exits. > 3) Point (2) means that a call to _CrtDumpMemoryLeaks() before main exits > will report leaks, where as using something like: > > struct cleanup > { > ~cleanup() > { > _CrtDumpMemoryLeaks(); > } > }; > > cleanup c; The boost unit test framework does appear to do the memory dump in a destructor of a static object - I guess whether the regex library statics are destructed first depends on the order of loading - not sure if there is any easy way of ensuring that the unit test framework is destructed last (perhaps if I link to UTF as a static library?). > 4) Other leak detectors such as valgrind and VLD report no errors (and give > much more useful reports than the brain-dead MS ones IMO). > > HTH, John. :-) yes I agree - my current solution is --detect_memory_leaks=0 when running the unit tests - which at least stops the report. I was hoping that this was a reasonably common issue and there was some clean up function (like u_cleanup()) specifically for the regex library but if not can live with switching off memory loss detection. Thanks a lot for the help here though - it is much appreciated. Alex

2 1

Re: [Boost-users] [Repost][Boost][Serialization] ASSERTtriggered during reading 1.40 binary archive w/ Boost 1.47
by marc.viala＠sfr.fr 03 Aug '11

03 Aug '11

> De la part de Pfligersdorffer, Christian > > We were facing similar problems jumping from 1.36 to 1.45. > Binary archive backwards compatibility is broken in a few places. > > But: it is fixable. We may provide a patch Thanks Christian for your reply to this annoying issue. If you confirm that the backward compatibility is broken for Boost.Serialization #1.40 to #1.47 then *Yes* I would really appreciate any king of patch to address this issue. Looking the Boost.Serialization code, my first feeling would be to patch the implementation of basic_binary_iarchive::load_override(class_id_type & t, int version) for the library_version_type(5) ('5' stands for the library version for the release #1.40) but maybe this patch is too naïve and insufficient. If you can send me back a patch for this issue, I'll use it ASAP. > is there a TRAK ticket? I'm not sure but the ticket #5567 (https://svn.boost.org/trac/boost/ticket/5567) seems to address this bug. Thanks in advance and best regards, Marc Viala

1 0

Re: [Boost-users] [Repost][Boost][Serialization] ASSERT triggeredduring reading 1.40 binary archive w/ Boost 1.47
by Pfligersdorffer, Christian 03 Aug '11

03 Aug '11

We were facing similar problems jumping from 1.36 to 1.45. Binary archive backwards compatibility is broken in a few places. But: it is fixable. We may provide a patch - is there a TRAK ticket? Regards, -- Christian Pfligersdorffer Software Engineering www.eos.info > -----Ursprüngliche Nachricht----- > Von: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] Im Auftrag von > Marc Perso > Gesendet: Mittwoch, 3. August 2011 07:30 > An: boost-users(a)lists.boost.org > Betreff: [Boost-users] [Repost][Boost][Serialization] ASSERT triggeredduring reading 1.40 binary archive > w/ Boost 1.47 > > Let me forward this message posted few days ago since nobody has confirmed > that the issue described hereafter exists. I would really appreciate if > someone could take a look (maybe Mr Robert Ramey) and give some advice on > this subject. > > Best regards, > > Marc Viala > > ------------------------------------------------------------------------ > > Hello Booster, > > We are trying to jump from Boost 1.40 to Boost 1.47 but we are facing an > issue: an ASSERT is raised in debug mode when reading back binary archive > generated by Boost.Serialization 1.40 with the Boost.Serialization 1.47. > This ASSERT is at line 116 in basic_archive.hpp: > > BOOST_ASSERT(t_ <= boost::integer_traits<base_type>::const_max) > > And it occurs with the following platform: MSVC 8.0 and Windows XP/32bits or > Win7/64bits. > > Looking at the load_override and save_override implementations for > class_id_type in #1.40, they seem loading and saving an 'int_least16_t' > (equiv. to a short for our platform) while load_override(class_id_type) for > #1.47 attempts to read an 'int' (file: basic_binary_iarchive.hpp @ line 98). > > My questions are: > - This assert suggests that we can't read back previous binary archives > (#1.40 at least) with Boost.Serialization 1.47, is this a known pending > issue for Boost.Serialization? > > - If yes, how we can manage that and how we can complete the migration to > 1.47 in our case (it should be noted that we must be able to read back our > 1.40 binary archives)? Is there any known patch for this problem? > > Any assistance will be greatly appreciated. > > Best regards, > > Marc Viala > > PS: I'm not sure but ticket #5567 seems already opened on this issue but has > not been yet fixed (https://svn.boost.org/trac/boost/ticket/5567). > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

Boost.Tokenizer / Sep - Compile-time errors ...
by asif saeed 03 Aug '11

03 Aug '11

Hi, Here is a piece of code that _only_ works when I provide it a hard-coded constant ";": typedef boost::char_separator<char_type_t> sep_type_t; typedef boost::tokenizer<sep_type_t> tokenizer_t; sep_(pa_->separator_.c_str()), t_(messageString,sep_), i_(t_.begin()) It wouldn't work with the separator passed in a variable (pa_->separator.c_str()). Tokenizer wouldn't work even if I initialize a const char* with a ";" and then pass that const char *variable to sep_ above. It gives me the following errors on if I pass separator in a variable: error C2780: 'bool boost::char_separator<Char>::operator ()(InputIterator &,InputIterator,Token &)' : expects 3 arguments - 1 provided Any idea how I can pass it a separator in a variable? Best, Asif

2 2

[Repost][Boost][Serialization] ASSERT triggered during reading 1.40 binary archive w/ Boost 1.47
by Marc Perso 03 Aug '11

03 Aug '11

Let me forward this message posted few days ago since nobody has confirmed that the issue described hereafter exists. I would really appreciate if someone could take a look (maybe Mr Robert Ramey) and give some advice on this subject. Best regards, Marc Viala ------------------------------------------------------------------------ Hello Booster, We are trying to jump from Boost 1.40 to Boost 1.47 but we are facing an issue: an ASSERT is raised in debug mode when reading back binary archive generated by Boost.Serialization 1.40 with the Boost.Serialization 1.47. This ASSERT is at line 116 in basic_archive.hpp: BOOST_ASSERT(t_ <= boost::integer_traits<base_type>::const_max) And it occurs with the following platform: MSVC 8.0 and Windows XP/32bits or Win7/64bits. Looking at the load_override and save_override implementations for class_id_type in #1.40, they seem loading and saving an 'int_least16_t' (equiv. to a short for our platform) while load_override(class_id_type) for #1.47 attempts to read an 'int' (file: basic_binary_iarchive.hpp @ line 98). My questions are: - This assert suggests that we can't read back previous binary archives (#1.40 at least) with Boost.Serialization 1.47, is this a known pending issue for Boost.Serialization? - If yes, how we can manage that and how we can complete the migration to 1.47 in our case (it should be noted that we must be able to read back our 1.40 binary archives)? Is there any known patch for this problem? Any assistance will be greatly appreciated. Best regards, Marc Viala PS: I'm not sure but ticket #5567 seems already opened on this issue but has not been yet fixed (https://svn.boost.org/trac/boost/ticket/5567).

1 0