- Boost-users - lists.boost.org

[String algo] Breaking change between 1.41 and 1.44 in split
by Christophe Henry 13 Oct '10

13 Oct '10

Hi, I noticed a breaking change in the split function when the input string is empty. Consider the following example: #include <string> #include <vector> #include <iostream> #include <boost/algorithm/string/split.hpp> #include <boost/algorithm/string/classification.hpp> int main() { std::vector<std::string> result; std::string empty_string; boost::algorithm::split( result, empty_string,boost::algorithm::is_any_of( "\t" ),boost::algorithm::token_compress_off); std::cout << "result size: " << result.size() << std::endl; return 0; } With boost 1.41 I get: result_size: 1 With 1.44 I get: result_size: 0 Meaning that splitting an empty string now has a new meaning. All I found to this subject was that 1 seems to be the correct answer (http://lists.boost.org/Archives/boost/2005/07/90266.php) Furthermore, there has been, according to the release notes, no change to string algo since the 1.41. This brings me to the question, before I start updating my code, is this change a design decision or an accident? Thanks, Christophe

1 0

[Accumulators] How to define the initial value ?
by Adrien Todeschini 13 Oct '10

13 Oct '10

Hi everyone, As the code of weighted_sum_impl constructor is [ in header <boost/accumulators/statistics/weighted_sum.hpp> ] template<typename Args> weighted_sum_impl(Args const &args) : weighted_sum_( args[parameter::keyword<Tag>::get() | Sample()] * numeric::one<Weight>::value ) { } and the template parameter Tag is defined as tag::sample namespace tag { struct weighted_sum : depends_on<> { typedef accumulators::impl::weighted_sum_impl<mpl::_1, mpl::_2, tag::sample> impl; }; } How can I define this tagged argument when initializing the accumulator_set ? ( Indeed, weighted_sum_impl constructor is called at this point ) Thanks in advance for your help. Adrien

2 6

[serialization] Deserialization of a float whose value is infinite throws an archive_exception ("stream error")
by Luca 12 Oct '10

12 Oct '10

Hello, using the Boost.Serialization library version 1.43, I saved to an XML archive a float number which has a value of infinite; then loading that XML file leads to a boost::archive::archive_exception, exception.what() return "stream error". The source code to reproduce the problem is the following: #include <boost/serialization/serialization.hpp> #include <sstream> #include <boost/archive/xml_woarchive.hpp> #include <boost/archive/xml_wiarchive.hpp> #include <iostream> struct XXX { friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & BOOST_SERIALIZATION_NVP(mValue); } float mValue; };//struct int main() { std::wstringstream lSS; boost::archive::xml_woarchive oa(lSS); XXX xxx; float zero = 0.f; xxx.mValue = 1 / zero; oa << BOOST_SERIALIZATION_NVP(xxx); std::wcout<<"saved:"<<lSS.str() << std::endl; boost::archive::xml_wiarchive ia(lSS); ia >> BOOST_SERIALIZATION_NVP(xxx); std::wcout<<"loaded:"<<lSS.str() << std::endl; return 0; } Using gcc 4.4.4 on Linux/Debian the mValue member is saved as: <mValue>inf</mValue> while on Visual Studio 2005 on Windows XP sp2 the mValue member is saved as: <mValue>1.#INF</mValue> I see that the value of infinite has a different representation, (and this could be the main motivation behind the choice to not support them), but this lead to a situation in which a properly saved archive is not loadable by the library itself. As far I can see, the issue is at this line 209 of \libs\serialization\src\basic_xml_grammar.ipp parse_info<BOOST_DEDUCED_TYPENAME std::basic_string<CharType>::iterator> result = boost::spirit::parse(arg.begin(), arg.end(), rule_); return result.hit; where the parse function fails since result.hit is equal to false. I resolved the issue by checking all the data preventing accurately any infinite value before it is saved to an xml archive, but I wonder if this problem belongs to the library itself or not. Greetings, Luca

2 1

[asio]How to insure async handles not be called after calling deadline_timer::cancel()?
by 鑫炜廖 12 Oct '10

12 Oct '10

The document says, If the timer has already expired when cancel() is called, then the handlers for asynchronous wait operations will: have already been invoked; or have been queued for invocation in the near future. These handlers can no longer be cancelled, and therefore are passed an error code that indicates the successful completion of the wait operation. Does this mean that async handlers may be called after calling deadline_timer::cancel() (say, the second curcumstance) ? If yes, it seems make things unintuitive, eg. delete the resources that associate with handlers after cancel(). Is there any way to suppress or identify these handlers? thx.

3 2

Re: [Boost-users] boost 1.44.0 undefined reference to boost::filesystem2::no_check
by Asif Fayyaz 12 Oct '10

12 Oct '10

No, that issue is a bit different and I have already included <boost/filesystem.hpp> I am again posting the code: --------------------------------Sample application source: #include <stdio.h> #include <boost/filesystem.hpp> int main() { std::string m_FileName(boost::filesystem::path("testfile1.txt", boost::filesystem::no_check).native_file_string()); printf("Testboost.....%s\n", m_FileName.c_str()); return 0; } -------------------------------- Asif ---------- Replied message ---------- From: Igor R <boost.li...(a)gmail.com> Date: Oct 11, 5:57 pm Subject: boost 1.44.0 undefined reference to boost::filesystem2::no_check To: BOOST Archives > Any one?? > any idea? > Waiting for response in this regard. There was a similar issues here: http://permalink.gmane.org/gmane.comp.lib.boost.user/62789 Doesn't it apply your case? do you include <boost/filesystem.hpp>? _______________________________________________ Boost-users mailing list Boost-us...@lists.boost.orghttp:// lists.boost.org/mailman/listinfo.cgi/boost-users

3 2

Distance/size of boost ranges for std::set.
by Matthew Gwynne 12 Oct '10

12 Oct '10

Hi, I'm working within a C++ library which makes extensive use of the range concept, and also of boost::size. The current version of the library makes use of boost 1_38, however, updating to the newer versions of boost, I've had to replace boost::size with boost::distance as some of the ranges involved, such as std::set don't have random access iterators. However, this presents a problem - it's not really acceptable within our application to have linear time complexity for the size operation on sets when the size member function of the set container is constant time, so I'm looking for some advice. Are we misusing the range concept here and should be using something else? Everything other than this matches up nicely, and it's only the time complexity with regards to boost::distance that will cause problems. Should we simply be extending the boost::size function with additional overloaded versions specific to std::set if we wish to use it in this way (is this reasonable?)? Thanks in advance! Matthew Gwynne

3 3

[shared_ptr] Better support for allocators
by Alex Dubov 12 Oct '10

12 Oct '10

Greetings. shared_ptr<> can be constructed with both custom allocator and deleter. However, there's an unfortunate shortcoming: if the held object needs to do some unusual housekeeping on itself during destruction, it needs to store an additional copy of the allocator or rely on some other custom functionality. It would be considerably more convenient, if shared_ptr<> could invoke a custom deleter with allocator (optionally) supplied: virtual void sp_counted_impl_pda::dispose() { d_(p_, a_); } Another handy alternative is to be able to obtain the allocator out of shared_ptr<> explicitly, in a manner similar to get_deleter(). Tr1 implementation actually constructs a proxy deleter object around the user supplied one, which inherits from allocator, but it's impossible to obtain a reference to it from the outside.

1 0

Variadic Macro Data library in the Boost sandbox
by Edward Diener 12 Oct '10

12 Oct '10

This may interest end-users who use macros/Boost PP in their own code to do preprocessor metaprogramming. I have put in the Boost sandbox the Variadic Macro Data library, or VMD. This library supplies variadic macro functionality and integrates variadic macros with the Boost PP library without changing the latter in any way. The library in the sandbox is in the variadic_macro_data folder. The library is a header-only library of macros. There is documentation and tests and the appropriate jamfiles for both. To build the documentation you need doxygen but I have also provided documentation in an html subfolder. Access to the Boost sandbox is explained on the Boost web site at http://www.boost.org/community/sandbox.html. The main use of the library is to be able to pass variadic data to macros, and in particular to extract variadic data elements or convert variadic data to Boost PP data types for use by Boost PP. The library also enhances the functionality of Boost PP tuple data. Installation is simply to take the boost and lib subfolders of the library and copy them to a Boost installation, preferable the trunk where support for variadic macros has been added in Boost config in the form of a BOOST_NO_VARIADIC_MACROS which is defined for compilers/versions which do NOT support variadic macros. To use the macros in the library, just "#include <boost/variadic_macro_data/VariadicMacroData.hpp>". Compilers which support variadic macros are gcc 3.0 and above and VC8 and above among a few others. You could also use the library on any installed Boost installation if the compiler you are using supports variadic macros. There is a discussion in the doc of why you might choose to use variadic macros rather than a Boost PP data type in a macro tailored for end-users. Comments, suggestions, and bug reports are welcome. Edward Diener

1 0

[array][bind] Binding boost array to a boost function
by Ryan McConnehey 11 Oct '10

11 Oct '10

I'm trying to bind a boost array to be called at a later time. This is giving me a lot of errors. Is what I'm attempting possible? Ryan template<typename T, std::size_t U> class Sorter { private: typedef boost::function<void(boost::array<T, U> & a)> t_SortFunction; t_SortFunction m_SortFunction; public: Sorter(t_SortFunction func) : m_SortFunction(func) {} bool sort(boost::array<T, U> & a) { m_SortFunction(a); } }; template<typename T, std::size_t U> void quicksort(boost::array<T, U> & a) { } int main(int argc, char* argv[]) { Sorter shell(boost::bind(quicksort, _1)); return 0; }

3 4

[filesystem][v3] remove_all and system::error_code error?
by valblade 11 Oct '10

11 Oct '10

Hi everyone, First of all, I want to thanks boost::filesystem contributors for their amazing work. I have a question in the recent filesystem V3. when using boost::uintmax_t remove_all(const path& p, system::error_code* ec=0); like this : ----------- listing ----------------------- remove_all("C:\\test_path_to_delete"); ------------- end listing --------------- the "C:\\test_path_to_delete" (and its descendent) is successfully deleted (as described in documentation) but when I use it like this : ----------- listing ----------------------- boost::system::error_code ec; remove_all("C:\\test_path_to_delete", ec); ------------- end listing --------------- The directory (and it descendent) aren't removed at all. I have investigated and, I found the following : boost::uintmax_t remove_all_aux(const path& p, fs::file_status sym_stat, error_code* ec) { boost::uintmax_t count = 1; if (!fs::is_symlink(sym_stat)// don't recurse symbolic links && fs::is_directory(sym_stat)) { for (fs::directory_iterator itr(p); itr != end_dir_itr; ++itr) { fs::file_status tmp_sym_stat = fs::symlink_status(itr->path(), *ec); HERE=> if (ec != 0 && ec) return count; count += remove_all_aux(itr->path(), tmp_sym_stat, ec); } } remove_file_or_directory(p, sym_stat, ec); return count; } This function is called by remove_all. Here, ec is passed by address. When remove_all("C:\\test_path_to_delete"); is called, &ec = 0 (default argument) so the check if (ec != 0 && ec) will never hit. But when called with remove_all("C:\\test_path_to_delete", ec); ec is allocated and &ec != 0, the check if (ec != 0 && ec) is allways true, and the function returns at the first iteration, I'm maybe wrong, but, isn't this what you meant to do? if (ec != 0 && *ec) return count; Sorry if I have misunderstood something... Thanks again.

1 0