June 2011 - Boost-users - lists.preview.boost.org

Serialization: how to inject context into save/load free functions?
by Michael Schuerig 20 Jun '11

20 Jun '11

I have a POD struct that I de/serialize through appropriately overloaded save and load free functions[*]. In my case, de/serialization involves translating between different representations (uid <-> username, gid <-> groupname). This translation is encapsulated in a class Etc and currently I just have a local variable of class Etc in both methods. I'd prefer a way to inject the Etc instance into save and load. The ideal solution would meet these criteria: * The Etc instance is created externally to save/load. * No global or quasi-global (static or singleton) variables. * No changes to the serialized struct (i.e. not an additional pointer to an Etc instance). * Usable with existing archive classes without extension. As far as I can tell, so far, I can't get all of these. I'm hoping that there is a way that I have overlooked. Michael [*] https://github.com/mschuerig/preserfs/blob/master/src/DirectoryMetadata.cpp… -- Michael Schuerig mailto:michael@schuerig.de http://www.schuerig.de/michael/

3 3

Boost.Preprocessor: expansion stops, result is further expandable
by Roman Neuhauser 20 Jun '11

20 Jun '11

Hello, I'd like to thank everybody involved in creation of Boost.Preprocessor for this cool piece of software, it does wonders. I have a possibly naive question (I'm an autodidact) which involves Boost.PP. I have stumbled upon strange behavior where the preprocessor stops expanding but the output expands further if fed to the preprocessor again. Given these directives: #include <boost/preprocessor.hpp> # #define DAPHP_REGISTER_CLASS(class_, props) \ CXXEXT_REGISTER_CLASS( \ class_ \ , BOOST_PP_SEQ_TRANSFORM(DAPHP_MAKE_PROPERTY, ~, props) \ ) # #define DAPHP_MAKE_PROPERTY(_, __, name) \ (name, IS_LONG, 0) # #define CXXEXT_REGISTER_CLASS(class_, props) \ object< class_ >::register_(); \ CXXEXT_DECLARE_PROPERTIES(class_, props) # #define CXXEXT_DECLARE_PROPERTIES(class_, props) \ BOOST_PP_SEQ_FOR_EACH( \ CXXEXT_DECLARE_PROPERTY_SEQ \ , ~ \ , BOOST_PP_SEQ_TRANSFORM(CXXEXT_PROPERTY_ADD_CLASS, class_, props) \ ) # #define CXXEXT_PROPERTY_ADD_CLASS(_, class_, prop) \ BOOST_PP_ARRAY_DATA(BOOST_PP_ARRAY_PUSH_FRONT((3, prop), class_)) # #define CXXEXT_DECLARE_PROPERTY_SEQ(_, __, elem) \ BOOST_PP_EXPAND(CXXEXT_DECLARE_PROPERTY elem) # #define CXXEXT_DECLARE_PROPERTY(native_type, name, ztype_, default_) \ object< native_type >::declare_property( \ ZEND_ACC_PUBLIC \ , ztype_ \ , BOOST_PP_STRINGIZE(name) \ , default_ \ ); # #define DAPHP_PROPS_DATE (year) (month) (day) This: DAPHP_REGISTER_CLASS(DapiDate, DAPHP_PROPS_DATE) Expands to (newlines added): object< DapiDate >::register_(); object< DapiDate >::declare_property( ZEND_ACC_PUBLIC , IS_LONG , "year" , 0 ); object< DapiDate >::declare_property( ZEND_ACC_PUBLIC , IS_LONG , "month" , 0 ); object< DapiDate >::declare_property( ZEND_ACC_PUBLIC , IS_LONG , "day" , 0 ); With these definitions added to the mix: #define DAPHP_REGISTER_CLASSES(classes) \ BOOST_PP_SEQ_FOR_EACH( \ DAPHP_REGISTER_CLASS_SEQ \ , ~ \ , classes \ ) # #define DAPHP_REGISTER_CLASS_SEQ(_, __, classprops) \ DAPHP_REGISTER_CLASS( \ BOOST_PP_TUPLE_ELEM(2, 0, classprops) \ , BOOST_PP_TUPLE_ELEM(2, 1, classprops) \ ) This: DAPHP_REGISTER_CLASSES( ((DapiDate, DAPHP_PROPS_DATE)) ) Expands to (newlines added): object< DapiDate >::register_(); BOOST_PP_SEQ_FOR_EACH( CXXEXT_DECLARE_PROPERTY_SEQ , ~ , ((DapiDate, year, IS_LONG, 0)) ((DapiDate, month, IS_LONG, 0)) ((DapiDate, day, IS_LONG, 0)) ) Which is an intermediate expansion of the above listed DAPHP_REGISTER_CLASS(DapiDate, DAPHP_PROPS_DATE). g++ 4.5.0 (MinGW) and MSVC 9 Express behave the same, which makes me believe this is a PEBKAC, but I just don't see it. Why doesn't DAPHP_REGISTER_CLASSES expand all the way as DAPHP_REGISTER_CLASS does? Please enlighten me. -- roman

2 2

Request for an addition in Boost Coding standard
by Gokulakannan Somasundaram 19 Jun '11

19 Jun '11

Hi, We have a custom memory allocator and we don't do global operator new/delete override for the obvious reasons. While the container classes being provided by Boost get the allocator as a template arguments, there are a lot of classes which do contain string and vector inside them, but don't provide the allocator as template arguments. May i request to have a coding standard to provide allocators as template arguments in all the classes that maintain private heap directly /indirectly( through the other containers )? To name a few, i saw the implementation of Boost Format, which provides the allocator for string, but not for the vectors. I also looked at the Boost filesystem, which has a string inside, but restricts itself to std::allocator. Thanks, Gokul.

7 11

[preprocessor] using ENUM_PARAMS to generate macro call arguments
by Krzysztof Czainski 18 Jun '11

18 Jun '11

Hello, I'll try to illustrate what I am trying to do on this example: #define VER 1 // given alternative sets of macros: #if VER==1 #define M0() #define M1( arg0 ) #define M2( arg0, arg1 ) #define M3( arg0, arg1, arg2 ) #define M4( arg0, arg1, arg2, arg3 ) #elif VER==2 #define V0( x ) #define V1( x, arg0 ) #define V2( x, arg0, arg1 ) #define V3( x, arg0, arg1, arg2 ) #define V4( x, arg0, arg1, arg2, arg3 ) #endif // I'm trying to write alternative versions of INVOKE for the above sets of macros, // so that the interface below needs no #ifs #if VER==1 #define INVOKE( n_args ) \ BOOST_PP_CAT(M,n_args)( BOOST_PP_ENUM_PARAMS(n_args,arg) ) #elif VER==2 #define INVOKE( n_args ) \ BOOST_PP_CAT(V,n_args)( dummy BOOST_PP_ENUM_TRAILING_PARAMS(n_args,arg) ) #endif // the interface implemented in terms of INVOKE #define A0() INVOKE(0) #define A1( arg0 ) INVOKE(1) #define A2( arg0, arg1 ) INVOKE(2) #define A3( arg0, arg1, arg2 ) INVOKE(3) #define A4( arg0, arg1, arg2, arg3 ) INVOKE(4) int main() { A2(1,2) } My goal here is to keep INVOKE as simple and non-redundant as possible, because alternative versions of INVOKE for alternative sets of given macros will be added later on. Now the call A2(1,2) doesn't compile. If I understand correctly, the reason is ENUM_PARAMS(n_args,arg) is passed to M2, while I need the result of macro expansion of ENUM_PARAMS(...) to be passed to M2. So my question is: Is there a way around this? I attached a file ready to compile with the presented code. Thanks in advance Kris

2 2

[boost] [review] [heap] Summary
by Andrew Sutton 18 Jun '11

18 Jun '11

The formal review of Tim Blechmann's Heap library has concluded. Only four votes were cast, although more people commented on issues either on the mailing list or though the Code Collaborator tool. However, from the comments received, the review identified no fundamental problems in the design or implementation of the library that would entail rejection. A such, the Boost.Heap library is ACCEPTED into Boost Votes for the library were cast by: Phil Endecott Daniel Russel Votes against: Thomas Klimpel (provisionally) Andrew Sutton Other comments received during the course of the review were generally positive, with a number of technical issues identified that need to to be addressed. The final version of the library can be fully accepted once Tim has addressed the technical issues identified during the review. Below is a summary of technical concerns. - The pairing heap has a known issue, regarding stack overflows, making it potentially unusable for production code. If the recursion issue can't be adequately solved, then the data structure should carry a much stronger warning label (i.e., put it in an experimental directory). - The performance of b_heap vs. d_ary_heap vs. priority_queue needs to be better characterized. Under what circumstances does one outperform the other. If no convincing scenario can be found where b_heap outperforms the others, then it should also be placed in an experimental directory. - The heap concepts need to be revisited as per comments on the review tool. The concepts should also describe semantics of the operations, not just the interfaces. - There were several comments on the implementation of heap stability. Three alternatives were identified: “naturally stable” heaps (of which no examples were given), the current counter based approach, and a chaining approach. No consensus on the best implementation strategy. Although the counter-based approach is probably adequate for most applications, it is not universally desirable. A more thorough investigation of alternative strategies is encouraged. It would be acceptable to document the investigation in a “Future Work” section of the library documentations (with any experiments also housed in the library). - The notion of Mergeable heaps needs to be better addressed, at least to differentiate between heaps that merge in O(n) or slower time and those that can merge in O(lg n) time. The proposed solution is to write merge as a general heap algorithm and specialize it for Mergeable heaps by calling the member function merge. Also, merge should be destructive, so merge and clear should not be needed. - The documentation must be substantially improved with respect to the overall design of heap data structures (i.e., concepts), especially mutable heaps and best practices. It also needs to provide both theoretical and performance comparisons for the different heaps. - Include documentation of the benchmarking tool in the documentation. It could be instructive for users to evaluate different heap implementations on their host or target architectures. - Document the complexity of == and < for heaps. - Address the comments and issues in the Review Tool. There were other recommendations that should be considered in the long term for the library. - Thorsten Ottosen suggest making b_heap and d_ary_heap container adaptors along the lines of std::stack, std::queue, and std::priority_queue, allowing parameterization over (e.g., stack-based container). Subsequent discussion indicates that there are some impracticalities in doing so, namely that it could invalidate other policies (e.g., stability). - Phil Endecott suggested rewriting some heaps (b_heap and d_ary_heap) in terms of more general algorithms operating on a (Random Access) Range in order to support the formulation of heap structures on other memory structures. Doing this would also help support the definition of heaps that satisfied Thorsten's request. I strongly recommend investigating Phil's suggestion, but this is not a barrier to acceptance. This feature could be added in future revisions of the library. I'd like to thank the community members who took the time to review Tim's work and especially those that submitted formal reviews. I'd also like to thank Tim (again) for his contributions and patiently answering our questions and addressing our comments. Andrew _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

1 0

about using boost on dev cpp
by prasad jondhale 18 Jun '11

18 Jun '11

hi all, i am trying to install boost library on dev cpp but not able to do it as i am new to boost so please give me step by step procedure to install it on dev cpp i am trying from last 5 days thanx

2 1

Iterators for the cartesian product
by kelvSYC 18 Jun '11

18 Jun '11

So that the wheel doesn't have to be reinvented if it's already out there, is there something in boost that will take a bunch of sequences (lists, say) and return a sequence containing the cartesian product all of them, either as a tuple or after applying a function to them? For example, is there something in boost that will, given a bunch of lists of integers, return a list of integers formed by multiplying together one member from each list?

3 6

[bind] member function as signal handler
by Tim Odenthal 17 Jun '11

17 Jun '11

Hi, all! I'd like to get a member function of an object to handle signals - is that possible with boost::bind? I tried the following, which doesn't work: SomeClass{ ... bool done_; public: void handler_done(int signal){ std::cout << "Recieved signal " << signal <<" ..." <<std::endl; done_ = true; } }; then, in the main: #include <signal.h> #include <boost/bind.hpp> ... SomeClass someClass; signal( SIGINT,boost::bind( &SomeClass::handler_done, &someClass,_1) ); ... I get the following error (g++ (Ubuntu 4.4.3-4ubuntu5) 4.4.3): error: cannot convert ‘boost::_bi::bind_t<void, boost::_mfi::mf1<void, SomeClass, int>, boost::_bi::list2<boost::_bi::value<SomeClass*>, boost::arg<1> > >’ to ‘void (*)(int)’ for argument ‘2’ to ‘void (* signal(int, void (*)(int)))(int)’ Is this generally impossible, or am I just making some stupid error from not really understanding boost::bind properly? Thanks in advance! Tim

2 1

Intel Compiler: "warning #1170: invalid redeclaration of nested class" with bimaps
by Oleksiy Atramentov 17 Jun '11

17 Jun '11

Dear Experts. I am getting "invalid redeclaration of nested class" warnings when I include "bimap/bimap.hpp" and use intel compiler(v12). Actually, these warnings are still present when the source has nothing but the #include and an empty main(). I noticed that I don't get these warnings with gcc (v4.1.2). I tried to debug the warning but I am still not sure whether the compiler (and/or boost settings for it) is beeing extra carefull and the warning is benign. Thank you, Oleksiy 1>boost/bimap/tags/support/tag_of.hpp(57): warning #1170: invalid redeclaration of nested class 1> BOOST_BIMAP_STATIC_ERROR( USING_TAG_OF_WITH_AN_UNTAGGED_TYPE, (Type) ); 1> ^ 1> 1boost/bimap/relation/support/member_with_tag.hpp(73): warning #1170: invalid redeclaration of nested class 1> BOOST_BIMAP_STATIC_ERROR( MEMBER_WITH_TAG_FAILURE, (Relation,Tag) ); 1> ^ 1> 1>boost/bimap/relation/support/value_type_of.hpp(57): warning #1170: invalid redeclaration of nested class 1> BOOST_BIMAP_SYMMETRIC_METADATA_ACCESS_BUILDER 1> ^ 1> 1>boost/bimap/relation/support/pair_type_by.hpp(48): warning #1170: invalid redeclaration of nested class 1> BOOST_BIMAP_SYMMETRIC_METADATA_ACCESS_BUILDER 1> ^ 1>

1 0

Re: [Boost-users] [Serialization] [Ptr_container] Serializing a boost::ptr_vector fails
by Philipp Münzel 17 Jun '11

17 Jun '11

I'm sorry, I simply forgot to #include <boost/ptr_container/serialize_ptr_vector.hpp> please excuse the disturbance. Regards, Philipp On 17.06.2011 16:18 "Philipp Münzel" <mailing-lists(a)philippmuenzel.de> wrote: > Dear list, > > I want to use boost::ptr_vector<> in conjunction with > boost::serialization. > The ptr_cotainer documentation says: > > > As of version 1.34.0 of Boost, the library supports serialization > > via Boost.Serialization. > > Of course, for serialization to work it is required that the stored > > type itself is serializable. For maps, both the key type and the > > mapped type must be serializable. > > However, when I create a class like this: > > #include <boost/ptr_container/ptr_vector.hpp> > #include <boost/serialization/access.hpp> > > > class Foo { > boost::ptr_vector<int> the_vector_ > friend class boost::serialization::access; > > template<class Archive> > void serialize(Archive & ar, const unsigned int) > { > ar & the_vector_; > } > } > > and then try to serialize it: > > std::ofstream ofs("filename"); > Foo f; > boost::archive::text_oarchive oa(ofs); > oa << f; > > I'm getting the following error, using gcc 4.2.1 and boost 1.46.1: > > > /usr/local/include/boost/serialization/access.hpp:118: error: 'class > boost::ptr_vector<Waypoint, boost::heap_clone_allocator, > std::allocator<void*> >' > has no member named 'serialize' > > How can I serialize a boost::ptr_container ? > > Regards, > Philipp >

1 0