- Boost-users - lists.boost.org

yet another logging library
by Mojmir Svoboda 10 Jun '08

10 Jun '08

Hello, i'm trying to create a logging library framework. It is my second attempt, the first can be seen at http://www.codeplex.com/flog . There is some documentation in the doc directory that is perhaps worth reading. The first version used boost::iostreams and not very smart locking mechanism in mt environment (in terms of performance mainly), but served more as a verification 'it can be done and it is not that ugly as i thought' than a real library candidate. Now i'd like to focus on performance and code size, so I made a sketch of the next version. It is really small and is rather a model than anything usable, but before I start implementation, I'd like to know opinion of library writers and users - what do you think about the concept? is it worth continuing, or i'd better go doing something better? :) The sketch of the next version can be seen at http://www.codeplex.com/fflog It's meant to be illustration of the interface and it's usage itself. Short summary of the design: User specifies compile-time vector of requested filters that apply to the logged text. Each filter requiring user's input is tagged and when logging function is invoked, the supplied types are checked against expected type, set to filter and later written to a buffer. Buffers are managed by the sink_manager and likely is one of few places that has to be mt-safe, as the rest is on the stack. I thought about some lock-free queues of buffers that are assigned to caller and filled returned back to write. The actual write of the buffer can be performed in dedicated lower-priority thread. Run-time filtering like level or context is possible, but not implemented yet. The main idea can be seen in the 1st version in runtime_filtering.hpp. Again, runtime values are shared among threads, so they have to be protected in mt environment. Many thanks for your attention, Mojmir

6 13

parallel-iteration
by e r 10 Jun '08

10 Jun '08

Hi, suppose i have container_type<range_type> seqs; i'm looking for iterator, whose instance I call i, such that if(*i == seq [j][k] ){ if(j<seqs.size()){ ++i; *i == seq[j+1][k] }else{ ++i; *i == seq[0][k+1] } if anyone has this or anything similar, could he/she kindly share it? the above is not quite the same as zip_iterator, as the parallel ranges all have the same type. also i'm aware that multi_arrays deal with this, but here i'm interested in arbitrary type. thanks for reading this thread, e.

1 0

is there strided iterator already available somewhere?
by e r 10 Jun '08

10 Jun '08

hello, is there strided iterator already available somewhere? iter = vec.begin(); std::advance(iter,offset) for(unsigned int i=0; i<(vec.size()-offset)/stride; ++i){ /* do something with iter */ std::advance(iter,stride) } alternatively, a permutation iterator does not seem very practical here... thanks, e.

2 2

metafunction invocation performance
by Hicham Mouline 09 Jun '08

09 Jun '08

Hello, I submitted a post a few days ago http://lists.boost.org/boost-users/2008/06/36852.php and I went to investigate on my own. Having not reached a solution to speed up the compilation, I looked again at the post and realized the code didn't compile due to a wrong paste. Apologies (if anyone tried), Anyways, here is the version that is too slow and runs out of mem during compilation on a P4 512Mb minimal redhat box with g++4.1 while, on core2duo 2Gb vista64 with intel 10.1, it compiles in 40s and runtime is correct. The point is to generate all template "curve" specializations for various tmpl arguments (constant, linear....) (cross product of 3 mpl sequences interpolator_types, extrapolator_types and integrator_types , each containing the possible tmpl type arguments 1, 2 and 3), and insert the fully specialized templates curve<T1, T2, T3> in a mpl vector "curve_types" to be used by make_variant_over. Regards, -------------------- #include <iostream> #include <boost/mpl/pair.hpp> #include <boost/mpl/placeholders.hpp> #include <boost/mpl/fold.hpp> #include <boost/mpl/vector/vector0.hpp> #include <boost/mpl/vector.hpp> #include <boost/mpl/joint_view.hpp> #include <boost/mpl/transform.hpp> #include <boost/mpl/for_each.hpp> #include <boost/mpl/size.hpp> #include <boost/variant.hpp> class constant {}; class linear {}; class polynomial {}; class cubic_spline {}; class simpson {}; template <typename inter, typename extra, typename integ> class curve {}; template <typename inter, typename extra, typename integ> struct curve_metafunction { typedef curve<inter,extra,integ> type; }; template<typename T> struct pair_with : boost::mpl::pair<T,boost::mpl::_> {}; template<typename pair> struct triple_with : curve_metafunction<typename pair::first, typename pair::second, boost::mpl::_> {}; template<typename Sequence1,typename Sequence2> struct sequence_product : boost::mpl::fold< Sequence1, boost::mpl::vector0<>, boost::mpl::joint_view< boost::mpl::_1, boost::mpl::transform< Sequence2, pair_with<boost::mpl::_2> > > > {}; template<typename SequencePair, typename Sequence> struct seq_seqpair_product : boost::mpl::fold< SequencePair, boost::mpl::vector0<>, boost::mpl::joint_view< boost::mpl::_1, boost::mpl::transform< Sequence, triple_with<boost::mpl::_2> > > > {}; template<typename Sequence1,typename Sequence2,typename Sequence3> struct seq_product : seq_seqpair_product< sequence_product<Sequence1,Sequence2>::type, Sequence3 > {}; typedef boost::mpl::vector<constant, linear, polynomial, cubic_spline> interpolator_types; typedef interpolator_types extrapolator_types; typedef boost::mpl::vector<simpson> integrator_types; typedef seq_product<interpolator_types, extrapolator_types,integrator_types>::type curve_types; struct type_printer { template <typename T> void operator()( T ) { std::cout<< typeid(T).name() <<std::endl; } }; int main() { std::cout<< boost::mpl::size<curve_types>::type::value <<std::endl; boost::mpl::for_each<curve_types>( type_printer() ); }

4 7

Using boost 1.35.0 on Mac OS X with Xcode 3
by Phil Hystad 09 Jun '08

09 Jun '08

I am rebuilding a library that was last compiled a year ago using an earlier version of boost and an earlier version of Xcode (2.4) and an earlier version of Mac OS X (10.4). I downloaded boost_1_35_0 distro for my Mac and built it using the standard configuration and build procedures. That seemed to go OK. I am compiling C++ source code that formerly compiled, without errors or warnings, but now reports the following errors in many files: "error: 'snprintf' is not a member of 'std'" And, the offending line is the following from the c++locale.h file (the error is reported on the line referencing snprintf): #ifdef _GLIBCXX_USE_C99 const int __ret = std::snprintf(__out, __size, __fmt, __prec, __v); #else const int __ret = std::sprintf(__out, __fmt, __prec, __v); #endif Does anyone know the source of this problem. There is absolutely no changes in my source that is being compiled. I am dual posting this on both the boost-users and the Mac Xcode lists because I am not sure which is the more appropriate forum.

1 0

Re: [Boost-users] [iterators][range] size compile failure in 1.35
by Thorsten Ottosen 09 Jun '08

09 Jun '08

Jeff Flinn skrev: > David Abrahams wrote: >> in 1.35 I'm pretty sure that iterator_range<T>::size() only works when T >> is a random access iterator, and filtered iterators are by necessity >> bidirectional. I suggest you use >> >> std::distance(r.begin(), r.end()) >> >> if you're willing to pay for O(N) execution. >> > > Hmm, it looks like that's what iterator_range<T>::size() did in 1.34.1 > while 1.35 returns m_Begin - m_End. Any idea why this was done? boost::distance(r) could also work. I did it as part of the general clean-up of the range lib. size() is expected to be O(1) by users, distance() is not. -Thorsten

1 0

multi-threaded safe circular_buffer
by Selçuk Giray Özdamar 09 Jun '08

09 Jun '08

Hi everyone, I specialized the boost::circular_buffer template class for concurrent access. But I have no idea about how the locking strategy must be for copy ctor, and also assignment operator. I inserted question marks at the end of the related lines. Thanks in advance... template <class T> class CircularBuffer { typedef boost::circular_buffer<T> internal_buffer; typedef boost::shared_mutex ReadWriteMutex; typedef boost::shared_lock<boost::shared_mutex> ReadLock; typedef boost::unique_lock<boost::shared_mutex> WriteLock; public: // Constructors explicit CircularBuffer(capacity_type capacity = 10) : m_capacity(capacity) , m_buffer(capacity) { } // copy ctor CircularBuffer(const CircularBuffer& other) { //WriteLock w_lock(rw_mutex); ??????????? m_capacity = other.m_capacity; m_buffer = other.m_buffer; } size_type size() const { ReadLock r_lock(rw_mutex); return m_buffer.size(); } void push_back(param_value_type item = value_type()) { WriteLock w_lock(rw_mutex); m_buffer.push_back(item); } // assignment operator CircularBuffer& operator= (const CircularBuffer& other) { //WriteLock w_lock(rw_mutex); ?????????? if (this != &other) { m_capacity = other.m_capacity; m_buffer = other.m_buffer; } return *this; } private:// member variables capacity_type m_capacity; internal_buffer m_buffer; // rw_mutex for internal buffer mutable ReadWriteMutex rw_mutex; };

3 3

serialize shared_array
by Hansi 09 Jun '08

09 Jun '08

Hello, Is it possible to serialize a shared_array?How? Best regards Hansjörg

1 0

Converting seconds to minutes and seconds using boost::datetime
by Sean Farrow 09 Jun '08

09 Jun '08

Hi: Does boost::datetime allow me to convert seconds (such as 164) to minutes and seconds? If yes wha function does this. Sean. __________ Information from ESET NOD32 Antivirus, version of virus signature database 3167 (20080609) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com

2 3

serialization, 1.34/1.35 difference: serializing a vector
by Filip Peters 08 Jun '08

08 Jun '08

I just tried serializing a simple vector, std::vector<unsigned int> vectortest; If I serialize out in 1.34, I can't read it in in 1.35. xml result from 1.35: <vectortest> <count>3</count> <item>1</item> <item>2</item> <item>3</item> </vectortest> from 1.34: <vectortest> <count>3</count> <item_version>0</item_version> <item>1</item> <item>2</item> <item>3</item> </vectortest> Notice the item_version, this causes an exception when reading it in with 1.35.... I need a way to get the behavior from 1.34 back.... Otherwise my files aren't readable anymore. small testprogram: #include <boost/config.hpp> #include <boost/archive/xml_oarchive.hpp> #include <boost/archive/xml_iarchive.hpp> //#include <boost/archive/binary_iarchive.hpp> //#include <boost/archive/binary_oarchive.hpp> #include <boost/serialization/vector.hpp> #include <boost/serialization/export.hpp> #include <fstream> #include <iomanip> int _tmain(int argc, _TCHAR* argv[]) { std::vector<unsigned int> testvector; testvector.push_back(1.); testvector.push_back(2.); testvector.push_back(3.); //this in 1.34 std::ofstream ofs( "testing2.txt", std::ios::binary ) ; boost::archive::xml_oarchive oa(ofs); oa << boost::serialization::make_nvp("vectortest", testvector); ofs.close(); //do this in 1.35 try{ std::ifstream ofs( "testing2.txt", std::ios::binary ) ; boost::archive::xml_iarchive ia(ofs); ia >> boost::serialization::make_nvp("vectortest", testvector2); ofs.close(); } catch(boost::archive::archive_exception testexception){ //reading a file from 1.34, will end up here std::string error = testexception.what(); } return 0; } ____________________________________________________________________________________ You rock. That's why Blockbuster's offering you one month of Blockbuster Total Access, No Cost. http://tc.deals.yahoo.com/tc/blockbuster/text5.com

5 6