March 2006 - Boost-users - lists.boost.org

[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

What's best practice to incorporate Boost.Test in own project?
by Mateusz Łoskot 08 Mar '06

08 Mar '06

Hi, I'm going to add Unit Tests to one of projects. It's quite big project, so there will be many tests suits and cases. The project build system is based on automake for Unix and nmake for Windows. I wonder what is the best practice to include Boost.Test frameworks to my own project? Currently I consider only Unit Test Framework. I see following possibilities: 1. Boost libraries are installed from official release (precompiled or compiled by user) Makefiles for my project try to detect if Boost exists and link to Unit Test Framework librar. Here Boost has to be listed in compilation requirements of my project. 2. Boost.Test libraries in binary form are included to my project tree and shipped together. So there is no requirement to have complete Boost installed, because test can link to UTF library already present 3. Include Boost.Test source code and allow user of my project to build UTF library on his own. Here I expect some problems with bjam/makefiles incompatibility. So, bjam would be required on user's platform. Simply, this solution seems to be quite confusing for user. I'd like to know your comments about which solution is best and what are possible traps hidding. Regarding the 1 solution I think there may be possible problem with using incorrect Boost.Test version. The second one involves additional work for overall projct maintenance. May be there are more possibilities I've not listed above, so please share your ideas. Thanks in advance for your comments. Best regards -- Mateusz Łoskot http://mateusz.loskot.net

3 7

Boost.Preprocessor with __LINE__ in MSVC71
by Wu Yinghui, Freddie 08 Mar '06

08 Mar '06

Hi all, I'm observing an interesting problem on MSVC71 regarding the use of __LINE__ with Boost.Preprocessor. The code looks like this ==== snip ==== #include <boost/preprocessor/cat.hpp> #define UNIQUE_ID BOOST_PP_CAT(id, __LINE__) #define UNIQUE_ID2 BOOST_PP_CAT(id, _MSC_VER) int main() { int UNIQUE_ID2 = 321; int UNIQUE_ID = 123; bool UNIQUE_ID = false; } ==== snip ==== The compiler works fine with UNIQUE_ID2, but not UNIQUE_ID. The only different in them is only the use of __LINE__ instead of _MSC_VER. Does anyone on the list has more insight on this issue? Or, what're the common ways to generate an unique identifier automatically within certain scope? Cheers, Freddie -- Wu Yinghui, Freddie Research & Development Software Engineer Volume Interactions Pte Ltd 1 Kim Seng Promenade, #12-01 Great World City East Tower Singapore 237994 Tel: +65 62226962 (Ext 216) Fax: +65 62226215 Email: yhwu(a)volumeinteractions.com URL: http://www.volumeinteractions.com Important: This message is intended for the recipient(s) addressed above. It contains privileged and confidential information. If you are not the intended recipient, please notify the sender immediately by replying to this message and then delete it from your system. You must not read, copy, use, or disseminate this communication in any form. Thank you.

3 2

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

08 Mar '06

There is a known bug with MSVC71 where if you use __LINE__ too much (not sure what the limits are or under what circumstances though), you get a "Fatal Error: Out of Keys" or something similar. Instead of this, for generating unique variable names etc I have seen the __COUNTER__ macro used instead, which provides similar functionality, but doesn't seem to suffer from this issue. Yours, Stewart -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org]On Behalf Of Wu Yinghui, Freddie Sent: 08 March 2006 09:00 To: boost-users(a)lists.boost.org Subject: [Boost-users] Boost.Preprocessor with __LINE__ in MSVC71 Hi all, I'm observing an interesting problem on MSVC71 regarding the use of __LINE__ with Boost.Preprocessor. The code looks like this ==== snip ==== #include <boost/preprocessor/cat.hpp> #define UNIQUE_ID BOOST_PP_CAT(id, __LINE__) #define UNIQUE_ID2 BOOST_PP_CAT(id, _MSC_VER) int main() { int UNIQUE_ID2 = 321; int UNIQUE_ID = 123; bool UNIQUE_ID = false; } ==== snip ==== The compiler works fine with UNIQUE_ID2, but not UNIQUE_ID. The only different in them is only the use of __LINE__ instead of _MSC_VER. Does anyone on the list has more insight on this issue? Or, what're the common ways to generate an unique identifier automatically within certain scope? Cheers, Freddie -- Wu Yinghui, Freddie Research & Development Software Engineer Volume Interactions Pte Ltd 1 Kim Seng Promenade, #12-01 Great World City East Tower Singapore 237994 Tel: +65 62226962 (Ext 216) Fax: +65 62226215 Email: yhwu(a)volumeinteractions.com URL: http://www.volumeinteractions.com Important: This message is intended for the recipient(s) addressed above. It contains privileged and confidential information. If you are not the intended recipient, please notify the sender immediately by replying to this message and then delete it from your system. You must not read, copy, use, or disseminate this communication in any form. Thank you. ________________________________________________________________________ This e-mail has been scanned for viruses by MessageLabs. ________________________________________________________________________ This e-mail has been scanned for viruses by MessageLabs.

1 0

Call for_each with boost::lambda library
by yinglcs2＠yahoo.com 08 Mar '06

08 Mar '06

Hi, I am trying to Call for_each with boost::lambda library to print out the return value of a function: vector<A> aList; for_each(aList.begin(), aList.end(), cout << bind(&A::getPriority, _1) ); This can't compile. can yo please tell me what did I do wrong? __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

[Format] Cost of using format
by Scott Meyers 08 Mar '06

08 Mar '06

I've been trying to get a handle on cost of using Boost.format by looking at the interface, reading the docs, and, until my head exploded, looking at the source code. What I'd really like to know is whether simple output without reordering incurs the cost of an allocation (for a string or maybe a stringstream) for each argument. For example, should I, as a user, expect the following to potentially cost me three dynamic memory allocations? cout << format("%1% %2% %3%") % 86 % 100.55 % "Hello"; The doc says this: > Since common stream implementations eventually call functions of the printf family for the actual formatting of numbers, in general printf will be noticeably faster than the direct stream operations And due to to the reordering overhead (allocations to store the pieces of string, stream initialisation at each item formatting, ..) the direct stream operations would be faster than boost::format, (one cas expect a ratio ranging from 2 to 5 or more) This suggests that per-argument buffering to support reordering may cost an allocation per argument, but if there is no reordering, does one pay this cost? I could figure this out on my own given enough time, but I'm hoping that others familiar with this library already know the answer and can save me the trouble. Thanks, Scott

1 0

multi_array - problems with resize
by Craig Glasgow 08 Mar '06

08 Mar '06

Hi, I'm trying to write serialization load and save methods for multi-array. For the load method I have a function with the prototype: template<class Archive, class T, std::size_t NumDims> inline void load(Archive& ar, boost::multi_array<T, NumDims>& arr, const unsigned int file_version) This function gets called when the array is serialized and should populate the multi-array arr. However this requires re-sizing arr which apparently can only be done using the extents construct e.g. arr.resize(boost::extents[3][4][2]), but I want to be able to do this for an arbitray number of dims as with reshape. Any idea how this can be done? Regards Craig

2 1