April 2007 - Boost-users - lists.preview.boost.org

string_algo no erase_if
by gast128 03 Apr '07

03 Apr '07

Dear all, I was trying to remove (replace, find...) characters from a string through the handy classification class. Unfortunately there is no erase_if fct which can be fed with this predicate functions. There is a trim_if function, but this trims only. It would be handy if all these '_if' variants where implemented for find, repalce and erase. Also the 'any' (or contains_if with predicate) is something we need regularly. But maybe this functionality can be achieved in another way. Anyone an idea? Remove_if/erase combination does of course work, but these are quite cumbersome. Wkr, me

1 1

checked_delete, shared_ptr, and incomplete types...
by Graham Reitz 03 Apr '07

03 Apr '07

I was toying with the example in "Beyond the C++ Standard Library An Introduction to Boost", Library 3:Utility. Specifically, the section on using checked_delete(). On page 85 it mentions that, "It can be easy to get into this kind of trouble with smart pointers". I tried the example presented with raw and shared_ptr. The destructor doesn't get called with the raw pointer example (the undefined behavior) but when using shared_ptr with the same example the destructor gets called. What was the author (Bjorn Karlsson) referring to when he mentioned this can be an issue with smart pointers? Thanks, Graham Code: // to_be_deleted.h #include <iostream> class to_be_deleted { public: ~to_be_deleted() { std::cout << "Help me get called.\n"; } }; // deleter.h #include "boost/shared_ptr.hpp" class to_be_deleted; using namespace boost; class deleter { public: //void delete_it(shared_ptr<to_be_deleted> p); void delete_it(to_be_deleted* p); }; // deleter.cpp #include "deleter.h" #include "boost/checked_delete.hpp" void deleter::delete_it(to_be_deleted* p) //void deleter::delete_it(shared_ptr<to_be_deleted> p) { delete p; //boost::checked_delete(p); } // main.cpp #include "boost/shared_ptr.hpp" #include "deleter.h" #include "to_be_deleted.h" using namespace boost; // Forces the type being deleted to be known at compile time int main(int argc, char* argv[]) { { //shared_ptr<to_be_deleted> p(new to_be_deleted()); to_be_deleted* p = new to_be_deleted; deleter d; d.delete_it(p); } return 0; }

1 0

Boost Con
by Robert Marion 02 Apr '07

02 Apr '07

So, attending the Boost Con in May is no trivial task. Aside from the $599 (early bird) conference cost and the hotel, you have to fly to Denver and then drive about 200 miles to Aspen -- because the airport in Aspen is closed in May (no kidding). I'm tempted to go just because the place will be nearly uninhabited. Dudes, what gives? Robert

4 3

Tokenizer - remainder of parsed string
by Gleske, Paul GOSS 02 Apr '07

02 Apr '07

I have a situation where I want to parse off the first few fields of a message and then return the rest of the message. The message format is: <STX>field1,field2,field3,field4,field5,field6,field7,field8,field9<ETX> What I want to do is pull off the first 5 fields and then get at the remainder of the string (fields 6-9) as a single string. I've pieced together something that mostly works It leaves me with a remainder of ",field6,field7,field8,field9". Note the leading comma. ----------------- std::string str = "\x02" "field1,field2,field3,field4,field5,field6,field7,field8,field9" "\x03"; typedef boost::tokenizer<boost::char_separator<char> > tokenizer; boost::char_separator<char> sep(",\x02\x03"); tokenizer tokens(str, sep); tokenizer::iterator tok_iter = tokens.begin(); for (int i=1; tok_iter != tokens.end(); tok_iter++) { std::cout << "<" << *tok_iter << "> "; ++i; if (i > 5) break; } std::cout << "\n"; std::cout << "current token <" << tok_iter.current_token() << ">"; std::cout << "\n"; boost::char_separator<char> sep2("\x03"); tokenizer remainder(tok_iter.base(), str.end(), sep2); std::cout << "<" << *remainder.begin() << ">" << "\n"; ----------------- What I don't understand is what the tok_iter.base() returns, or how the tokenizer constructor that uses it works (see the line for remainder). How can that constructer take tok_iter.base() and str.end()? Is the string iterator converted? Is there a better way to get the remainder of the tokenized string? I would very much like to not have the leading comma on the remainder. I know I can use substr on the remainder, but it seems that there should be a better way. Thanks, -pkg -------------------------------------------------------- Goss ... Innovation for Business NOTICE: This e-mail and any attachment(s) may contain confidential and proprietary information of Goss International Corporation and/or its subsidiaries and may be legally privileged. This e-mail is intended solely for the addressee. If you are not the addressee, dissemination, copying or other use of this e-mail or any of its content is strictly prohibited and may be unlawful. If you are not the intended recipient please inform the sender immediately and destroy the e-mail and any copies. All liability for viruses is excluded to the fullest extent permitted by law. Any views expressed in this message are those of the individual sender. No contract may be construed by this e-mail.

1 0

Boost and WXwidgets
by Lucas de Sales Joner 02 Apr '07

02 Apr '07

Hi, I have a program that use boost library (matrix, vector and others) and return a vector and I need plot this vector with wxwidgets. The problem is that I can´t use wxwidgets with boost library because return a lot of problems. My plot program work well, but when I use "include<boost/numeric/ublas/matrix.hpp>" in that program, return this errors: C:\Arquivos de programas\Dev-Cpp\include\boost\numeric\ublas\storage.hpp:25, from C:\Arquivos de programas\Dev-Cpp\include\boost\numeric\ublas\vector.hpp In file included from C:/Arquivos de programas/Dev-Cpp/include/boost/numeric/ublas/storage.hpp:25, from C:/Arquivos de programas/Dev-Cpp/include/boost/numeric/ublas/vector.hpp C:\Arquivos de programas\Dev-Cpp\include\boost\numeric\ublas\exception.hpp In member function `void boost::numeric::ublas::bad_argument::raise()': C:\Arquivos de programas\Dev-Cpp\include\boost\numeric\ublas\exception.hpp exception handling disabled, use -fexceptions to enable How can I use wxwidgets and boost library? Someone know other way to plot points in using boost? LUCAS DE SALES JONER Engenharia de Controle e Automação Industrial - UFSC Laboratório de Robotica - UFSC ionner_rs(a)hotmail.com ionner.rs(a)gmail.com (48) 9148 4732 _________________________________________________________________ Veja só alguns dos novos serviços online no Windows Live Ideas — são tão novos que ainda não foram disponibilizados oficialmente. http://ideas.live.com

2 1

Iostreams - GZip problem
by Jabin Reinhold 02 Apr '07

02 Apr '07

Im trying to use the iostreams gzip compression on xml serializations produced using boost::serialization and Im having a bit of trouble with it. The variable to be serialized is savedClass. As shown below, I create the serialization, save to a file with gzip_compressor, then read it from the file with gzip_decompressor. My issue is that at the end of the code, test1.xml(~10k) and test2.xml(~12k) are not identical. Test2.xml has some extraneous characters added to the end (~2k worth). I assume these are leftovers from a memcopy somewhere, but Im not sure why. I have the same effect when I use zlib_compressor as well. What is wrong with my usage of the library(s)? //////////////////////////////////////////////////////////////// std::strstream sStream; if(true) { //Scope used to ensure that the archive is fully processed boost::archive::xml_oarchive oArchive(sStream); oArchive << BOOST_SERIALIZATION_NVP(savedClass); } std::ofstream oStreamXML1( "c:\\test1.xml", std::ios::out ); oStreamXML1 << sStream.str(); if(true) { //Scope used to ensure that the archive is fully processed std::ofstream file("c:\\test.gz", std::ios_base::out | std::ios_base::binary); boost::iostreams::filtering_streambuf<boost::iostreams::output> outStream; outStream.push(boost::iostreams::gzip_compressor()); outStream.push(file); boost::iostreams::copy(sStream, outStream); } std::strstream sStream2; if(true) { //Scope used to ensure that the archive is fully processed std::ifstream file2("c:\\test.gz", std::ios_base::in | std::ios_base::binary); boost::iostreams::filtering_streambuf<boost::iostreams::input> inStream; inStream.push(boost::iostreams::gzip_decompressor()); inStream.push(file2); boost::iostreams::copy(inStream, sStream2); } std::ofstream oStreamXML2( "c:\\test2.xml", std::ios::out ); oStreamXML2 << sStream2.str(); ////////////////////////////////////////////////////////////////

2 2

Intrusive library review results: library accepted
by "JOAQUIN LOPEZ MU?Z" 02 Apr '07

02 Apr '07

I'm pleased to announce that Ion Gaztañaga's Intrusive library has been accepted into Boost. Prior to its inclusion in Boost, the author is encouraged to address several issues raised during the review, a summary of which is given below. A total of 14 full reviews were received, all of them positive. Additionally, other members did not submit full reviews but expressed their support for the library. One Boost member raised some concerns about the usefulness of the library; I hope these concerns have been satisfactorily addressed by those review participants who explained their current and potential usage scenarios for Boost.Intrusive. This has shown that more movitational material need to be provided in the documentation introduction and in the form of full-fledged examples. Thank you to Ion Gaztañaga for all the hard work involved in creating a Boost library, to Olaf Krzikalla, who wrote the first versions of this piece of work, and to all the participants in the review. The list of issues to be addressed follows. Feel free to comment if you detect some error or omission from my part. List of issues to be addressed prior to inclusion of the Intrusive library into Boost: * Documentation - Provide a motivating introduction about usefulness of Boost.Intrusive. - Add some hello world motivating examples. - Class names linked to their section in the reference (except in code due to Quickbook limitations). - Document iterator and const_iterator as part of the container interfaces. - Add overview of iterator validity rules. - Document usage before concepts. - Improve concepts part. - Provide container concepts à la STL. - Add performance section. - Add quick guide to docs. - Make the tests more comprehensive. - Add examples section. - State clearly the purpose of the lib in the introduction. - Add more examples using the different hooks. - State in docs the possibility that tags be incomplete types. - State in docs the fact that Boost.Intrusive does not support shared_ptr's. - Document the fact that member hooks rely on non-standard implementation techniques. * Naming - Change Destroyer to a more suitable name (e.g. Sink or Disposer), rename member functions accordingly. - Rename current() as iterator_to(). - Rename [s]list_algorithms as circular_[s]list_algorithms. - Drop the 'i' prefix from the names of containers. - Rename linked() as is_linked(). - Rename assign_and_destroy() as destroy_and_assign(). * Interface improvements - Study ways to simplify the usage syntax of hooks. - Add a default value to Tag. - Member hooks shouldn't be parameterized on the host class type. - Make hooks assignment operator a no-op. - bucket_size() y bucket() should be const. - Provide cbegin(), cend() etc. - Add *_fwd.hpp headers. - Fuse autolink and regular/safe hooks into one class whose second parameter is a category class. * Implementation improvements - Iterators to be implemented with iterator_facade. - Use Boost.Concept for concept checking if compilation times are not severely affected. - Extend bucket primes table to 64 bits. - Use template trick to avoid direct use std::size_t in template parms of containers, thus pulling the std namespace in via ADL. - Use Watanabe's version of swap_nodes() for ilist. - Use insert_after() in clone_from(). - Reimplement clone_from() to not delete partially constructed data on exception. - Investigate bug with member hooks and virtual inheritance. * Extensions and issues to be considered in the future - Order-statistics tree. - Study other libraries for ideas: DOL, PTL, jjLib. - Study feasibility of implementing a "destroy multihooked element when completely unlinked" mode. - Investigate use of two-way pointers with Boost.Intrusive. - Consider a shift() member function for lists and slists. - Additional slist-like container with a pointer-to-last member to make swapping constant time. The Review Manager, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

Typetraits for decomposing member function pointers?
by Tom Bachmann 02 Apr '07

02 Apr '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, is there a way to get more information about a member function pointer, using type traits, than is_member_function_pointer<T>::value? I'm looking for information about e.g. the constness and function_traits-like information. - -- - -ness- -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFGBs++vD/ijq9JWhsRApJNAJ9nbbZnRMnTMBi0l09yqBaCIiYGmgCfTjRO I12viJTyF3sUp0sICPR/nDg= =ym+h -----END PGP SIGNATURE-----

2 1

[mpl] find_all type vectors in a type matrix that match a predicate.
by Christian Henning 02 Apr '07

02 Apr '07

Hi there, is it possible to find all type vectors in a type matrix ( 2D type vector ) that match a predicate? In the end I would like to have them in another type matrix. Can somebody point out how I would put together such a metafunction? I'm not sure if the mpl lib supports my feature request. Consider the following code: #include "boost/mpl/vector.hpp" #include <boost/mpl/find_if.hpp> #include <boost/mpl/at.hpp> using namespace std; using namespace boost; namespace fields { struct id {}; struct name {}; struct address {}; struct birthday {}; } namespace tags { struct primary {}; struct foreign {}; struct no_key {}; } typedef mpl::vector<fields::id , tags::primary> id_field_t; typedef mpl::vector<fields::name , tags::foreign> name_field_t; typedef mpl::vector<fields::address , tags::foreign> address_field_t; typedef mpl::vector<fields::birthday, tags::no_key> birthday_field_t; typedef mpl::vector< id_field_t , name_field_t , address_field_t , birthday_field_t > table; // aggregate all foreign keys in the following vector typedef mpl::vector0<> foreign_keys; /// ??? metafunction for aggregating all foreign keys ??? int main() { return 0; } BTW, I'm not particular interested in a certain order of the aggregated type vectors. Reagrds, Christian

3 7

newbie's questions about multi_index
by Guillaume Lazzara 02 Apr '07

02 Apr '07

Hi, I'm currently using multi_index and I have few questions: - I have a multi_index with composed keys, sometimes I need to get a range of items, iterate over them and modify their key. How can I do this properly? sample code: EquivSet::iterator tmp; std::pair<EquivSet::iterator, EquivSet::iterator> es = contains.equal_range(*it); for (EquivSet::iterator it_equiv = es.first; it_equiv != es.second;) if (!X[it_equiv->id]) { tmp = it_equiv++; contains.modify_key(tmp, update_equiv(nbEns)); } else it_equiv++; In this example, the second iterator in the pair, normaly used as end iterator, can be changed and, sometimes, the loop iterates beyond the latter. - If I can choose between a std::set and a multi_index used as a set, which one should I use? Thank you very much for your answers! -- Guillaume Lazzara Epita CSI 2008

2 2