- Boost-users - lists.boost.org

Re: [Boost-users] Boost.Preprocessor with __LINE__ in MSVC71
by Wu Yinghui, Freddie 09 Mar '06

09 Mar '06

>>> -----Original Message----- >>> That's what I feel. The error message indicate that the >>> compiler somehow considers "id" and __LINE__ two identifier >>> when the error occurs. I tried using the /P parameter to dump >>> the preprocessed source, and that looked fine, though. So I >>> guess it is the problem with MSVC compiler mixing >>> preprocessing and parsing phases together... > > Yes. There is nothing that the pp-lib can do about it. I agree. But I pointed this out after much effort trying to debug the mysterious bug. Maybe it'd be better to document this in pp-lib's manual as "known issues?" That'd make future victims' lives easier. :) >>> > What do you mean by "certain scope"? Or, more precisely, >>> given a > finite scope such as that defined by a function, >>> why do you need > unique identifiers? Most of the time when >>> people want unique > identifiers, they want an identifier >>> that is unique across an entire > program or, at minimum, >>> inside a translation unit. >>> >>> Well... The scope can be the whole program, or in a >>> namespace, or in a translation unit. The unique ID is used in >>> a macro in the library for some tidious code that can be >>> automatically generated using preprocessor. > >For the record, doing this sort of thing is almost guaranteed to produce ODR >violations. With that said... Well... I think I know what I'm doing now. The use of __LINE__ (or __COUNTER__ on MSVC) is working for me after combining with some salt as keyword. Thanks for your concern, but the current solution seems suffice for me purpose. >>> Anyway, thanks for Stewart Tootill's suggestion. I had to >>> resort to using __COUNTER__ at the meantime, which seemed >>> working so far. (But one thing that worries me is that >>> __COUNTER__ is only unique within a translation unit. So what >>> if I need a unique identifier throughout the >>> program?) > >The pp-lib contains a similar construct, but it must be manually >updated. E.g. >#include BOOST_PP_UPDATE_COUNTER() >BOOST_PP_COUNTER() // 1 >#include BOOST_PP_UPDATE_COUNTER() >BOOST_PP_COUNTER() // 2 >BOOST_PP_COUNTER() // 2 >#include BOOST_PP_UPDATE_COUNTER() >BOOST_PP_COUNTER() // 3 But this is not documented. :( I didn't know its existence until you told me and then I did a grep.) >I could provide the means to seed this value for a translation unit, but I'm not >sure if this is really a good idea. Thoughts? I don't think there is a general solution for "all cases" here. (At least I don't have one with me right now.) But given some restricted usage, this solution actually works out quite well. BTW, I think I'm still sticking to __LINE__/__COUNTER__ solution for now, since it is not possible for me to encode the "#include ..." line into my macro. >Regards, >Paul Mensonides Thanks a lot for your efforts and insight! Cheers, Freddie

2 1

[Serialization] Status of bug in 1.33?
by Andrew Lin 09 Mar '06

09 Mar '06

Hi everyone, Some time back I wrote asking about a problem I was having with the 1.33 serialization library which turned out to be a bug: http://lists.boost.org/boost-users/2005/12/15969.php In short, custom archives had some kind of compilation failure which I never really understood, but which didn't fail in 1.32. Does anyone know what ever came of this? Is there a bug tracker where I can follow its status? (I didn't see it in the 1.33.1 release notes, so I'm assuming it's still broken.) I thought I'd be able to use 1.32 until 1.34 came out, but something has come up in the project which would be perfect for the multi_index library with hashing, new in 1.33. Thanks, Andrew ============================================================================================= Email transmissions can not be guaranteed to be secure or error-free, as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message which arise as a result of email transmission. In addition, the information contained in this email message is intended only for use of the individual or entity named above. If the reader of this message is not the intended recipient, or the employee or agent responsible to deliver it to the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication, disclosure of the parties to it, or any action taken or omitted to be taken in reliance on it, is strictly prohibited, and may be unlawful. If you are not the intended recipient please delete this email message. ==============================================================================================

2 1

[Lambda] Help with if_then_else
by Scott Meyers 09 Mar '06

09 Mar '06

Consider this (working) code using Boost.FileSystem: fs::directory_iterator di(dir); fs::directory_iterator end; while (di != end) { if (is_directory(*di)) cout << "D\n"; else cout << "~D\n"; ++di; } This looks like it should be lambda-able. Here's my attempt: for_each(fs::directory_iterator(dir), fs::directory_iterator(), if_then_else(bind(&fs::is_directory, _1), cout << "D\n", cout << "~D\n")); VC7.1 doesn't like this -- the diagnostics are below. Any idea how I can make this work? Thanks, Scott fs.cpp(39) : error C2784: 'const boost::lambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::ifthenelse_action,boost::tuples: :tuple<boost::lambda::lambda_functor<Arg1>,boost::lambda::lambda_functor<Arg2>,boost::lambda::lambda_functor<Arg3>>>> boost::lambda::if_then_else (const boost::lambda::lambda_functor<Arg1> &,const boost::lambda::lambda_functor<Arg2> &,const boost::lambda::lambda_functor<Arg3> &)' : could no t deduce template argument for 'const boost::lambda::lambda_functor<T2> &' from 'std::basic_ostream<_Elem,_Traits>' with [ _Elem=char, _Traits=std::char_traits<char> ] D:\C++\Boost\Current\boost\lambda\if.hpp(100) : see declaration of 'boost::lambda::if_then_else' fs.cpp(39) : error C2784: 'const boost::lambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::ifthenelse_action,boost::tuples: :tuple<boost::lambda::lambda_functor<Arg1>,boost::lambda::lambda_functor<Arg2>,boost::lambda::lambda_functor<Arg3>>>> boost::lambda::if_then_else (const boost::lambda::lambda_functor<Arg1> &,const boost::lambda::lambda_functor<Arg2> &,const boost::lambda::lambda_functor<Arg3> &)' : could no t deduce template argument for 'const boost::lambda::lambda_functor<T2> &' from 'std::basic_ostream<_Elem,_Traits>' with [ _Elem=char, _Traits=std::char_traits<char> ] D:\C++\Boost\Current\boost\lambda\if.hpp(100) : see declaration of 'boost::lambda::if_then_else' fs.cpp(39) : error C2784: 'const boost::lambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::ifthenelse_action,boost::tuples: :tuple<boost::lambda::lambda_functor<Arg1>,boost::lambda::lambda_functor<Arg2>,boost::lambda::lambda_functor<Arg3>>>> boost::lambda::if_then_else (const boost::lambda::lambda_functor<Arg1> &,const boost::lambda::lambda_functor<Arg2> &,const boost::lambda::lambda_functor<Arg3> &)' : could no t deduce template argument for 'const boost::lambda::lambda_functor<T2> &' from 'std::basic_ostream<_Elem,_Traits>' with [ _Elem=char, _Traits=std::char_traits<char> ] D:\C++\Boost\Current\boost\lambda\if.hpp(100) : see declaration of 'boost::lambda::if_then_else' fs.cpp(39) : error C2784: 'const boost::lambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::ifthenelse_action,boost::tuples: :tuple<boost::lambda::lambda_functor<Arg1>,boost::lambda::lambda_functor<Arg2>,boost::lambda::lambda_functor<Arg3>>>> boost::lambda::if_then_else (const boost::lambda::lambda_functor<Arg1> &,const boost::lambda::lambda_functor<Arg2> &,const boost::lambda::lambda_functor<Arg3> &)' : could no t deduce template argument for 'const boost::lambda::lambda_functor<T2> &' from 'std::basic_ostream<_Elem,_Traits>' with [ _Elem=char, _Traits=std::char_traits<char> ] D:\C++\Boost\Current\boost\lambda\if.hpp(100) : see declaration of 'boost::lambda::if_then_else'

4 6

[Smart_Ptr] Error in intrusive_ptr ctors' exception specification?
by Scott Meyers 08 Mar '06

08 Mar '06

Consider this constructor from the intrusive_ptr doc (http://boost.org/libs/smart_ptr/intrusive_ptr.html) intrusive_ptr(T * p, bool add_ref = true); Effects: if(p != 0 && add_ref) intrusive_ptr_add_ref(p);. Note in particular that nothing is said about exceptions being thrown. Presumably intrusive_ptr_add_ref may throw, as I can't find any restriction to the contrary. Now consider these constructors from the same document: intrusive_ptr(intrusive_ptr const & r); // never throws template<class Y> intrusive_ptr(intrusive_ptr<Y> const & r); // never throws Effects: if(r.get() != 0) intrusive_ptr_add_ref(r.get());. They are documented not to throw, yet both call intrusive_ptr_add_ref. How can it be that the top constructor above may throw, but these constructors may not? FWIW, it seems to me that all three should be allowed to throw, unless some constraint is imposed on the implementation of intrusive_ptr_add_ref. For example, perhaps intrusive_ptr_add_ref wants to throw if its reference count will overflow -- maybe because it's using only an 8-bit counter. Isn't that possible? Scott

3 2

Re: [Boost-users] Boost.Preprocessor with __LINE__ in MSVC71
by Wu Yinghui, Freddie 08 Mar '06

08 Mar '06

> What is going wrong? I tried it on VC7, VC7.1, and VC8, and it appeared to work correctly on all three versions of the compiler. >> Does anyone on the list has more insight on this issue? > I've seen strange problems related to __LINE__ on VC before, but I > can't remember off the top of my head what the issue was. > (Thinking...) It has something to do with the generation of debug > information. Okay, yes, I see the error that you are getting now. It > only happens in during a debug build. You are right. It occurs only in debug builds. (Sorry for forgetting to mention this.) > This is just one of the *many* ways in which VC's preprocessor is > wacky. It suffices to say it isn't even in the ballpark of > implementing the phases of translation correctly. That's what I feel. The error message indicate that the compiler somehow considers "id" and __LINE__ two identifier when the error occurs. I tried using the /P parameter to dump the preprocessed source, and that looked fine, though. So I guess it is the problem with MSVC compiler mixing preprocessing and parsing phases together... >> Or, what're the common ways to generate an unique identifier >> automatically within certain scope? > What do you mean by "certain scope"? Or, more precisely, given a > finite scope such as that defined by a function, why do you need > unique identifiers? Most of the time when people want unique > identifiers, they want an identifier that is unique across an entire > program or, at minimum, inside a translation unit. Well... The scope can be the whole program, or in a namespace, or in a translation unit. The unique ID is used in a macro in the library for some tidious code that can be automatically generated using preprocessor. Anyway, thanks for Stewart Tootill's suggestion. I had to resort to using __COUNTER__ at the meantime, which seemed working so far. (But one thing that worries me is that __COUNTER__ is only unique within a translation unit. So what if I need a unique identifier throughout the program?) > Regards, > Paul Mensonides Thanks for your comments. Cheers, Freddie

2 1

Newbie question
by Tom Nabhani 08 Mar '06

08 Mar '06

Hi, I'm having some difficulty using the boost libraries in my programs (I have an iBook running Panther). My boost install passes all the regression tests. My files are all stored in the appropriate places under /usr/local. But when I try to #include <boost> libraries in my programs, the compiler doesn't seem to find them. I'm assuming that there is some very simple stupid problem with the way my computer is set up, but I don't know enough about using c++ libraries to know what to do. I'm a pretty experienced Java programmer, but I'm very new to c++. Unfortuntately I don't have anyone to ask how to solve this problem. I hope this doesn't try your patience too much, but could anybody suggest how I might get this problem fixed? Thanks, Tom

3 2

[foreach] reverse foreach?
by Andy Little 08 Mar '06

08 Mar '06

FWIW Heres some sample code. Its seems a bit lopsided! std::ofstream out("toolpath_out.txt"); BOOST_FOREACH(boost::shared_ptr<section> sp,vect){ if (sp->m_action == section::reverse){ for( std::vector<std::string>::reverse_iterator iter = sp->m_vector.rbegin(); iter != sp->m_vector.rend(); ++iter){ out << *iter << '\n'; } } else { BOOST_FOREACH(std::string str,sp->m_vector){ out << str << '\n'; } } } Are there any plans for a BOOST_REVERSE_FOREACH? regards Andy Little

3 4

[MPL] Need help crossing runtime boundary (lengthy)
by Ames Andreas 08 Mar '06

08 Mar '06

Hi all, I'm new to using the MPL and I want to write a generic class, that makes it easy to provide other classes with a mechanism to notify their clients about member changes. I'll paste in some code that illustrates how I imagine this thing could look like. I've marked three places with comments ((1), (2), (3)) that contain the specific questions but I also welcome more general suggestions. For example I'd appreciate alternatives to the std::map within the Notifier. <code> template <class OBJECT_TYPE, typename MEMBER_ID_TYPE, typename ID_TO_MEMBER_MAP_TYPE> class Notifier { // an alternative combiner for signals struct notifier_combiner { typedef boost::optional< int > result_type; template <typename INPUT_ITERATOR> result_type operator()(INPUT_ITERATOR pFirst, INPUT_ITERATOR pLast) const { result_type res; for (; pFirst != pLast; ++pFirst) res = *pFirst; return res; } }; template <MEMBER_ID_TYPE id, typename MEMBER_TYPE> struct get_signal_type { typedef boost::signal< int (OBJECT_TYPE const*, MEMBER_TYPE const&), notifier_combiner > type; }; // (1) How can I get SOMETHING? It should be a sequence of all // possible signal types, most probably an mpl::set, so that each // signal type is only present once. I guess I should iterate // over the elements in ID_TO_MEMBER_MAP_TYPE and then call // get_signal_type< elem::first::value, elem::second::type >, but // I don't see exactly how to use one of the mpl-algorithms or // generic iteration. typedef boost::make_variant_over< SOMETHING >::type signals_t; // maps member ids to a variant containing the 'right' signal type typedef std::map< MEMBER_ID_TYPE, signals_t > signal_map_t; signal_map_t signal_map; public: Notifier() { // (2) How do I initialise signal_map so that all ids map to a // signals_t instance with the right default constructed // signal type assigned? I guess the answer is somewhere in // chapter 9 of "C++ template metaprogramming" but that's the // chapter that is the hardest one to grasp for me :-) } template <MEMBER_ID_TYPE id, typename MEMBER_TYPE> boost::connection connect(get_signal_type< id, MEMBER_TYPE >::type::slot_type slot) { return boost::get< get_signal_type< id, MEMBER_TYPE >::type >(signal_map[id]).connect(slot); } template <MEMBER_ID_TYPE id, typename MEMBER_TYPE> int notify(OBJECT_TYPE const *pObj, MEMBER_TYPE const &member_val) const { boost::optional< int > res = boost::get< get_signal_type< id, MEMBER_TYPE >::type >(signal_map[id])(pObj, member_val); return res ? res.get() : 0; } }; // how I'd like to use it: strcut sth_ { int i; long j; } class RandomClass { bool yesno; sth_ sth; enum MemberId { YESNO , STH }; typedef mpl::map< mpl::pair< mpl::integral_c< MemberId, YESNO >, bool > , mpl::pair< mpl::integral_c< MemberId, STH >, sth_ > > id_to_member_type_t; template < typename INTEGRAL_CONST > struct get_type_by_id { typedef mpl::at< id_to_member_type_t, INTEGRAL_CONST >::type type; }; Notifier< RandomClass // I hope incomplete type is ok here // (because Notifier only uses // RandomClass const*; OTOH, Notifier // uses a std::map which somehow // refers to RandomClass)? Anyway, // this is not crucial. , MemberId , mpl::map< mpl::pair< mpl::integral_c< MemberId, YESNO >, bool > , mpl::pair< mpl::integral_c< MemberId, YESNO >, sth_ > > > m_notifier; public: RandomClass() : yesno(false) , sth({1, 2}) { } // a client connects by specifying the member id and a callback template < MemberId id > boost::connection notify(int (pF*)(RandomClass const* , get_type_by_id< mpl::integral_c< MemberId, id > >::type const&)) { // (3) Can the compiler deduce the second template argument of // Notifier::connect? I guess not, so how could I change the // method's declaration to make that happen? return m_notifier.connect< id >(pF); } void setYesNo(bool b) { yesno = b; m_notifier<YESNO>(this, b); } void setSthI(int i) { sth.i = i; m_notifier<STH>(this, sth); } }; </code> Obviously this doesn't yet compile and there may be all sorts of errors. I hope that's ok, because I still need very basic help. TIA, aa -- Andreas Ames | Programmer | Comergo GmbH | Voice: +49 69 7505 3213 | andreas . ames AT comergo . com "I imagine Sisiphus as happy ..." Albert Camus

1 0

Re: [Boost-users] Boost.Preprocessor with __LINE__ in MSVC71
by Sohail Somani 08 Mar '06

08 Mar '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Wu > Yinghui, Freddie > Sent: Wednesday, March 08, 2006 5:45 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Boost.Preprocessor with __LINE__ in MSVC71 > > > What is going wrong? I tried it on VC7, VC7.1, and VC8, and it > appeared to work correctly on all three versions of the compiler. > > >> Does anyone on the list has more insight on this issue? > > > I've seen strange problems related to __LINE__ on VC before, but I > > can't remember off the top of my head what the issue was. > > (Thinking...) It has something to do with the generation of debug > > information. Okay, yes, I see the error that you are > getting now. It > > only happens in during a debug build. > > You are right. It occurs only in debug builds. (Sorry for > forgetting to > mention this.) When your debug information is for edit and continue iirc.

1 0

[mpl] random sequence generator, whats wrong?
by Mark Nevill 08 Mar '06

08 Mar '06

Hi, I am trying to make a template for building sequences of random numbers using Boost.Mpl, but apparently I haven't quite got the hang of it yet ;). Below is my code (adapted from the Boost.Random docs) and the compiler output (GCC v 3.4.2), I seem to have misunderstood the usage of mpl::first and mpl::second in lambdas or something. Can someone clarify what I've done wrong? The whole construct should be invoked with something like typedef utils::make_random_sequence<vector0_c<unsigned long>, utils::minstd_rand, 10>::type random_vector; to generator a vector_c of 10 random numbers. Thanks, Mark #include <cstdlib> #include <boost/static_assert.hpp> #include <boost/mpl/integral_c.hpp> #include <boost/mpl/plus.hpp> #include <boost/mpl/times.hpp> #include <boost/mpl/modulus.hpp> #include <boost/mpl/iter_fold.hpp> #include <boost/mpl/range_c.hpp> #include <boost/mpl/pair.hpp> #include <boost/mpl/erase.hpp> #include <boost/mpl/begin.hpp> #include <boost/mpl/end.hpp> #include <boost/mpl/push_back.hpp> #include <boost/mpl/lambda.hpp> #include <boost/type_traits/is_integral.hpp> namespace utils { using namespace boost::mpl; template<class T, class Formula, T seed> struct random : integral_c< T, template Formula::apply<seed>::value > { typedef random<T, Formula, value> next; template<T new_seed> struct reseed { typedef random<T, Formula, new_seed> type; }; }; template<class IntType, IntType a, IntType c, IntType m> struct linear_congruential { BOOST_STATIC_ASSERT(boost::is_integral<IntType>::value); typedef IntType value_type; template<IntType seed> struct apply : integral_c<IntType, (a * seed + c) % m> { }; }; typedef random< long, linear_congruential<long, 16807L, 0, 2147483647L>, 1043618065L > minstd_rand0; typedef random< long, linear_congruential<long, 48271L, 0, 2147483647L>, 399268537L > minstd_rand; template<class integral_sequence, class generator, size_t size> struct make_random_sequence { typedef typename iter_fold< range_c<size_t, 1, size>, pair< typename erase< integral_sequence, begin<integral_sequence>, end<integral_sequence> >::type, generator >, pair< push_back< first<_1>, second<_1> >::type, second<_1>::next > >::type type; }; } 28: error: missing `>' to terminate the template argument list 28: error: template argument 2 is invalid 28: error: expected `{' before "template" 28: error: expected unqualified-id before "template" 28: error: expected `;' before "template" C:/Boost/include/boost-1_33/boost/mpl/pair.hpp: In instantiation of `boost::mpl::first<mpl_::_1>': C:/Boost/include/boost-1_33/boost/mpl/aux_/has_tag.hpp:20: instantiated from `boost::mpl::aux::has_tag<boost::mpl::first<mpl_::_1>, mpl_::bool_< false> >' C:/Boost/include/boost-1_33/boost/mpl/sequence_tag.hpp:115: instantiated from `boost::mpl::sequence_tag<boost::mpl::first<mpl_::_1> >' C:/Boost/include/boost-1_33/boost/mpl/push_back.hpp:32: instantiated from `boost::mpl::push_back<boost::mpl::first<mpl_::_1>, boost::mpl::second<mpl_::_1> >' 85: instantiated from here C:/Boost/include/boost-1_33/boost/mpl/pair.hpp:43: error: no type named `first' in `struct mpl_::_1' 85: error: `type' is not a member of `boost::mpl::push_back<boost::mpl::first<mpl_::_1>, boost::mpl::second<mpl_::_1> >' 85: error: `type' is not a member of `boost::mpl::push_back<boost::mpl::first<mpl_::_1>, boost::mpl::second<mpl_::_1> >' C:/Boost/include/boost-1_33/boost/mpl/pair.hpp: In instantiation of `boost::mpl::second<mpl_::_1>': 86: instantiated from here C:/Boost/include/boost-1_33/boost/mpl/pair.hpp:56: error: no type named `second' in `struct mpl_::_1' 86: error: `next' is not a member of `boost::mpl::second<mpl_::_1>' 86: error: `next' is not a member of `boost::mpl::second<mpl_::_1>' 88: error: template argument 1 is invalid 88: error: template argument 2 is invalid 88: error: template argument 3 is invalid 88: error: `type' does not name a type

3 2