- Boost-users - lists.boost.org

Re: [Boost-users] [enable_if] using enable_if to extend std::numeric_limits?
by james.jones＠firstinvestors.com 07 Mar '07

07 Mar '07

From: me22 <me22.ca(a)gmail.com> > I don't think you can "extend" classes that already exist. Note, for > example, that boost has integer_traits, and doesn't try to modify > numeric_limits. > > Why do you want the concept to be in numeric_limits? Can you not > simply use is_decimal_type? I should be more precise - by "extend" I really mean "specialize". I'd like to provide specializations for numeric_limits for my new decimal types. I can certainly just provide my own decimal_limits class (and have implemented this already), but the idea is to create a drop-in replacement for, say, double - and existing code likely uses numeric_limits<T> with T=double already, so being able to provide this with T=my_decimal_type would be ideal. - James Jones Administrative Data Mgmt. Webmaster 375 Raritan Center Pkwy, Suite A Data Architect Edison, NJ 08837

2 1

Re: [Boost-users] [enable_if] using enable_if to extend std::numeric_limits?
by james.jones＠firstinvestors.com 07 Mar '07

07 Mar '07

From: "Ovanes Markarian" <om_boost(a)keywallet.com> > Sorry, > > I ment enable_if< is_decimal_type<T> >::type. The ::type will be expanded to > either void or a derivable type. Maybe I'm misunderstanding, but how can I derive std::numeric_limits from enable_if, since std::numeric_limits is already defined? I can't do this: namespace std { template <typename T> class numeric_limits : private boost::enable_if<boost::decimal::is_decimal_type<T> >::type { public: static T min() { return boost::decimal::decimal_limits<T>::min(); } }; } and I can't do this: namespace std { template <typename T> class numeric_limits<T> : private boost::enable_if<boost::decimal::is_decimal_type<T> >::type { public: static T min() { return boost::decimal::decimal_limits<T>::min(); } }; } because this is not a specialization. - James Jones Administrative Data Mgmt. Webmaster 375 Raritan Center Pkwy, Suite A Data Architect Edison, NJ 08837

2 1

[boost-thread] memory leak (according to Boost.Test)
by n.torrey.pines＠gmail.com 07 Mar '07

07 Mar '07

This little program, compiled on VC++ 8.0 leaks memory, apparently. #define BOOST_AUTO_TEST_MAIN #include <boost/test/auto_unit_test.hpp> #include <boost/thread/thread.hpp> void f() {} BOOST_AUTO_TEST_CASE(threads_leak) { boost::thread thrd(&f); thrd.join(); } Output: Running 1 test case... *** No errors detected Detected memory leaks! Dumping objects -> {218} normal block at 0x00357458, 24 bytes long. Data: < 9 > 80 39 15 00 FF FF FF FF 00 00 00 00 00 00 00 00 {217} normal block at 0x00357398, 8 bytes long. Data: <Xt5 > 58 74 35 00 01 CD CD CD Object dump complete. Versions: Boost-1.33.1 VC++8.0 (debug mode) Windows XP extra lib: libboost_unit_test_framework-vc80-mt-gd.lib (not auto-linked) Can anyone using VC++ 8.0 and the release version of Boost confirm this? With Google, I found some past references to Boost.Thread leaking memory.

3 4

iostreams use with asio tcp stream?
by Jeff 07 Mar '07

07 Mar '07

Everyone, Will I be able to use an asio tcp stream in an iostream filtering pipeline, so that I can first compress, then encrypt, then send the data over a tcp or udp connection? Thanks, Jeff

2 1

[enable_if] using enable_if to extend std::numeric_limits?
by james.jones＠firstinvestors.com 07 Mar '07

07 Mar '07

Suppose I have a set of templates that have a common interface (each is a model of the same concept, to use that lingo). I'd like to extend std::numeric_limits to support this concept. I'm trying to find a way to use enable_if to do this, but I don't see how. Is this possible? Is there some other trick to doing this? Example with code (not real code): template< class U, class V > class model1; tempalte< class U, class V > class model2; template< class U > is_decimal_type { static const bool value = false; }; template< class U, class V > is_decimal_type < model1<U, V> > { static const bool value = true; }; template< class U, class V > is_decimal_type < model2<U, V> > { static const bool value = true; }; Would like to do this: template< class T > class numeric_limits< T > { // extensions go here }; But only if is_decimal_type<T>::value is true (so that the partial specialization only applies to my concept). Thanks, - James Jones Administrative Data Mgmt. Webmaster 375 Raritan Center Pkwy, Suite A Data Architect Edison, NJ 08837

3 3

[BGL] In what order does out_edge_iterator iterate?
by Shufei Fan 07 Mar '07

07 Mar '07

Hi, In my application, the order in which edges are retreaved is also important to me. So my question is, when I iterate through an vertex's out_edges, do they show up in the order in which the edges are inserted, or any other principals are used in order of iteration? Example: ...... typename graph_traits<Graph>::out_edge_iterator out_i, out_end; typename graph_traits<Graph>::edge_descriptor e; for (tie(out_i, out_end) = out_edges(v, g); out_i != out_end; ++out_i) { e = *out_i; Vertex src = source(e, g), targ = target(e, g); std::cout << "(" << get(vertex_id, src) << "," << get(vertex_id, targ) << ") "; } ...... How are these edges ordered in the output? And is it right all those iterators iterates in the same principals/order? Thanks, Shufei

4 4

[boost::wave] Finding out if a macro is defined
by Ovanes Markarian 07 Mar '07

07 Mar '07

Hello All, using boost::wave (with context initialized as explained in the docs example) and trying to find out if some certain macro is defined leads to 2 errors: 1. Doc for context state: bool is_defined_macro(std::string const &name) const; But the source has no such function definition. I assume the signature was changed to: template <typename IteratorT2> bool is_defined_macro(IteratorT2 const &begin, IteratorT2 const &end) { return macros.is_defined(begin, end); } Trying to fix it and using the std::string::begin() and std::string::end() members to specify the begin and end range result in the error: c:\boost\include\boost-1_33_1\boost\wave\util\cpp_macromap.hpp(331) : error C2228: left of '.get_value' must have class/struct/union type is 'char' c:\boost\include\boost-1_33_1\boost\wave\cpp_context.hpp(150) : see reference to function template instantiation 'bool boost::wave::util::macromap<ContextT>::is_defined<IteratorT2>(const IteratorT &,const IteratorT &)' being compiled with [ ContextT=boost::wave::context<token_groups::str_iter_type,token_groups::lex_iter_type>, IteratorT2=char *, IteratorT=char * ] e:\projects\inpecting_wave\inpecting_wave\header_processor.cpp(177) : see reference to function template instantiation 'bool boost::wave::context<IteratorT,LexIteratorT>::is_defined_macro<char*>(const IteratorT2 &,const IteratorT2 &)' being compiled with [ IteratorT=token_groups::str_iter_type, LexIteratorT=token_groups::lex_iter_type, IteratorT2=char * ] This happens in cpp_macromap.hpp line 331: function source: template <typename ContextT> template <typename IteratorT> inline bool macromap<ContextT>::is_defined(IteratorT const &begin, IteratorT const &end) { // in normal mode the name under inspection should consist of an identifier // only token_id id = token_id(*begin); if (T_IDENTIFIER != id && !IS_CATEGORY(id, KeywordTokenType) && !IS_EXTCATEGORY(id, OperatorTokenType|AltExtTokenType)) { BOOST_WAVE_THROW(preprocess_exception, invalid_macroname, impl::get_full_name(begin, end).c_str(), main_pos); } IteratorT it = begin; error line >>>>>> string_type name ((*it).get_value()); <<<<<< error line typename defined_macros_type::iterator cit(current_macros -> find(name)); if (++it != end) { // there should be only one token as the inspected name BOOST_WAVE_THROW(preprocess_exception, invalid_macroname, impl::get_full_name(begin, end).c_str(), main_pos); } return cit != current_macros -> end(); } I assume compiler dereferences iterator to char and char has no member to get_value. Does anyone have suggestions? With Kind Regards, Ovanes Markarian

2 6

[program_options] ignoring unknown options
by martux 07 Mar '07

07 Mar '07

Hi, I am having some trouble getting program_options to ignore unknown options. As suggested in the documentation I set allow_unregistered() but po::store(po::command_line_parser(argc, argv).options(pd.get()).allow_unregistered().run(), vm); po::notify(vm); still throws boost::program_options::unknown_option. Could anyone please make clear how to do it correctly? -- View this message in context: http://www.nabble.com/-program_options--ignoring-unknown-options-tf3361505.… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

regex: seem to be missing something here
by Noah Roberts 06 Mar '07

06 Mar '07

fs::path upper("c:/differences", fs::native); fs::path checked(upper/"checked"); assert(fs::exists(upper)); assert(fs::exists(checked)); fs::directory_iterator end; rg::regex rgx("*ps", rg::regex::basic); for (fs::directory_iterator it(upper); it != end; ++it) { if (rg::regex_match(it->string(), rgx)) std::cout << "MATCH: "; else std::cout << "NOMATCH: "; std::cout << it->string() << std::endl; } Supposed to be looking for ps files. tried *\\.ps, which should work but also didn't. Doesn't match any of the strings. One example line of output: NOMATCH: c:/differences/Adj. speed test case 5 (metric units).ps Supposedly regex is greedy, no? It should accept . as a * element and continue to match everything up to ps and ps also...why no worky? Thanks.

1 1

[Boost.Condition]
by Manuel Jung 06 Mar '07

06 Mar '07

Hi, I would like to know the waiting threads for a condition. Is there a way to calculate the number of "wait()"s which would be affected by a notifiy_all()? Maybe nobody needed this before? Or isn't this possible to implement such a thing with boost::condition and i have to count the waiting threads on the application level? Greetings Manuel Jung

3 3