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

error C2976 in python tutorial?
by Guido Ziliotti 24 Aug '07

24 Aug '07

Probably Has anybody faced error C2976: 'boost::python::class_' : too few template arguments? Compiler msvc-8.0. BOOST_PYTHON_MODULE(hello){ class_<World>("World") .def("greet", &World::greet) .def("set", &World::set) ; } produced that. Source code follows. #include <string> char const* greet() { return "hello, world"; } struct World { void set(std::string msg) { this->msg = msg; } std::string greet() { return msg; } std::string msg; }; #include <boost/python/module.hpp> #include <boost/python/def.hpp> using namespace boost::python; //nota che la funzione globale va prima BOOST_PYTHON_MODULE(hello) { def("greet", greet); class_<World>("World") .def("greet", &World::greet) .def("set", &World::set) ; }

1 0

[test] Uncaught exception in BOOST_CHECK_THROW() causing SIGABRT
by Richard Dingwall 24 Aug '07

24 Aug '07

Hi there, I thought I'd post this because A) it might save someone a lot of head-scratching and B) there aren't any posts on it already. The short of it is, that if you have a setup similar to the following (i.e., an empty throw() clause on a template specialisation): template<class output_type> output_type do_something(int input_value) throw(std::invalid_argument); template<> int do_something(int input_value) throw(std::invalid_argument); template<> long do_something(int input_value) throw(); And attempt to test throwing a std::invalid_argument from within the long specialised function, e.g.: // note: an argument less than zero will throw a std::invalid_argument BOOST_CHECK_THROW(do_something<long>(-1), std::invalid_argument); You will see something the following: Running 4 test cases... terminate called after throwing an instance of 'std::invalid_argument' what(): Input value too low. unknown location(0): fatal error in "long_throw": signal: SIGABRT (application abort requested) sigabrt_test_suite.cpp(58): last checkpoint *** 1 failure detected in test suite "sigabrt tests" My function is throwing a std::invalid_argument and my test case been programmed to catch it. But it doesn't. This had me scratching my head for a few days, until I noticed the empty throw() clause. Completing it resolved the problem immediately. I'm not sure if this is expected behaviour, but if anyone's interested, a test case is attached. I'm using gcc version 4.0.1 (Apple Computer, Inc. build 5367) on a G4 Powerbook with whatever was in Boost's SVN trunk on August 6. Cheers, Richard

2 1

Multi Array Question
by chris dion 24 Aug '07

24 Aug '07

Hi I'm having a problem with the boost/multi_array and I'm hoping someone here can help, or at least explain why the following doesn't work. The problem: I have a multi_array of ints that I want to be able to view as being either row major or column major to simplify a number algorithms. I don't want to change the internal representation of the array (the default c++ row major storage layout is fine), I just want to be able to interchange the indexing order from array[row][col] to array[col][row] for the same multi_array. To do so, I attempted to define 2 array_views of the same multi_array with different stride values, like this: #include <iostream> using namespace std; #include <boost/multi_array.hpp> using namespace boost; int main(void) { typedef multi_array<int, 2> intarray2; typedef intarray2::array_view<2>::type intarray2view2; typedef intarray2::array_view<1>::type intarray2view1; typedef intarray2::index_range intarray2range; intarray2 arr(extents[3][3]); // test data int a = 0; for (int r = 0; r < 3; r++) { for (int c = 0; c < 3; c++) { arr[r][c] = a++; } } intarray2view2 rows = arr[indices[intarray2range().stride(3)][intarray2range().stride(1)]]; intarray2view2 cols = arr[indices[intarray2range().stride(1)][intarray2range().stride(3)]]; /* 0 1 2 3 4 5 6 7 8 */ cout << rows[2][1]; // i expect this to print 7 cout << cols[2][1]; // i expect this to print 5 } Unfortunately, this doesn't work at all (an assert fails when indexing the views). I debugged (using Visual C 2005), and found that the 'rows' array_view has a stride collection of [9, 1], while the 'cols' array_view stride collection is [3, 3]. Both are different than what I specified during construction. It appears that they're being multiplied by the extents of the first dimension of the multi_view array (in this example: 3). I can produce a column major view of a single column of the array using a 1 dimensional view constructed using a degenerate range, like this: intarray2view1 col= arr[indices[1][intarray2range()]]; // column 1 in arr but that's not what I'm hoping to achieve. Can someone explain the behavior above and/or suggest a way of creating a 2d column major array_view of a 2d multi_array? Thanks - Chris Dion _________________________________________________________________ Show Your Messenger Buddies How You Really Feel http://www.freemessengeremoticons.ca/?icid=EMENCA122

2 1

Construction/assignment inconsistency with boost::function on MSVC2005
by Josef Grahn 23 Aug '07

23 Aug '07

Hello, I encountered the following inconsistency constructing a boost::function object from a pointer to a template function. template <typename T> void foo(T) {} int main(int argc, char **argv) { boost::function<void (int)> f1; f1 = foo<int>; boost::function<void (int)> f2 = foo<int>; f1(0); // ok f2(0); // error, function pointer is zero! return 0; } I am compiling on MSVC2005 SP1, using boost version 1.34.1. It seems to work properly with GCC 4.1.2 and boost 1.33.1 on Ubuntu, however. Is this a known problem? Best Regards, Josef Grahn

3 2

Re: [Boost-users] [boost] ScopeExit Review Begins Today
by Jody Hagins 23 Aug '07

23 Aug '07

The discussion threads included comments from several people who did not offer an actual review and/or YES/NO. I have filtered every message with either scope or exit in the subject, and I want to make sure I have all the reviews. If your name is in this list, that means I have a YES/NO review from you. Otherwise, I may see some discussion from you, but no YES/NO review. If your name is NOT in this list, but you feel you provided a review, please point me to your review posting, or provide appropriate clarification. The fact that you participated in the discussion does not mean that I know your opinion on the acceptance of ScopeExit. In addition, I have received a few reviews via private email. For the sake of fairness and openness, I asked the authors to send their review to the list. I will not consider them unless they appear in the open mailing list. Thanks for your cooperation! Kim Barrett Aaron LaFramboise Steven Watanabe Ben Kuppers Andrey Semashev Peder Holt Goran Mitrovic Johan Nilsson Pavel Vozenilek Tom Brinkman Maxim Yanchenko

1 0

Re: [Boost-users] [boost] ScopeExit Review Begins Today
by Jody Hagins 23 Aug '07

23 Aug '07

Thanks for all the input on Alexander's ScopeExit submission. There are some good discussions, so please continue them over the next day or so if necessary. I am very busy this weekend (I am HC of a high school football team, and our first game is tomorrow night), but I will render a conclusion some time next week. Again, thanks for the input.

1 0

ptr_container problem
by Aljaz 23 Aug '07

23 Aug '07

Hello I'm trying to use ptr_vector but I'm having headaches because I have no idea why I'm getting this error: error C2248: 'boost::ptr_container_detail::reversible_ptr_container<Config,CloneAllocator>::operator =' : cannot access private member declared in class 'boost::ptr_container_detail::reversible_ptr_container<Config,CloneAllocator>' with [ Config=boost::ptr_container_detail::sequence_config<boost::asio::dns::base_nameserver_resource,std::vector<void *,std::allocator<void *>>>, CloneAllocator=boost::heap_clone_allocator ] C:\Program Files\boost\boost_1_34_0\boost/ptr_container/detail/reversible_ptr_container.hpp(287) : see declaration of 'boost::ptr_container_detail::reversible_ptr_container<Config,CloneAllocator>::operator =' with [ Config=boost::ptr_container_detail::sequence_config<boost::asio::dns::base_nameserver_resource,std::vector<void *,std::allocator<void *>>>, CloneAllocator=boost::heap_clone_allocator ] This diagnostic occurred in the compiler generated function 'boost::ptr_sequence_adapter<T,VoidPtrSeq,CloneAllocator> &boost::ptr_sequence_adapter<T,VoidPtrSeq,CloneAllocator>::operator =(const boost::ptr_sequence_adapter<T,VoidPtrSeq,CloneAllocator> &)' with [ T=boost::asio::dns::base_nameserver_resource, VoidPtrSeq=std::vector<void *,std::allocator<void *>>, CloneAllocator=boost::heap_clone_allocator ] My code (simplified): class base1 { public: void somefunc() { }; }; class main : public base1 { public: virtual void somefunc() { } }; class derived : public main { public: virtual void somefunc() { } }; boost::ptr_vector<main> m_vector; Any Idea what am I doing wrong? I tried to add do_clone functions/members without success. Any help is highly appreciated Regards A

2 1

ScopeExit Review ends tomorrow (Wednesday)
by Jody Hagins 23 Aug '07

23 Aug '07

The ScopeExit review ends on Wednesday, August 22. If you want to have input on its acceptance into Boost, please provide a review by tomorrow. BTW, we need more input ;-) Thanks! >From the original announcement... The review period for Scope Exit Time Series, submitted by Alexander Nasonov, runs from Monday, August 13 until Wednesday, August 22. >From the documentation: Nowadays, every C++ developer is familiar with RAII technique. It binds resource acquisition and release to initialization and destruction of a variable that holds the resource. But there are times when writing a special class for such variable is not worth the effort. This is when ScopeExit macro comes into play. You put resource acquisition directly in your code and next to it you write a code that releases the resource. Note that ScopeExit uses PP macros and relies on Boost.Typeof. The latest version can is here: http://boost-consulting.com/vault/index.php?action=downloadfile&filename=sc… It's available also at http://tinyurl.com/2z6qg2 Online documentation is here: http://194.6.223.221/~nasonov/scope_exit-0.04/libs/scope_exit/doc/html/inde… >From the review guidelines here: http://boost.org/more/formal_review_process.htm ... Boost mailing list members are encouraged to submit Formal Review comments: * Publicly on the mailing list. * Privately to the Review Manager. Private comments to a library submitter may be helpful to her or him, but won't help the Review Manager reach a decision, so the other forms are preferred. What to include in Review Comments Your comments may be brief or lengthy, but basically the Review Manager needs your evaluation of the library. If you identify problems along the way, please note if they are minor, serious, or showstoppers. Here are some questions you might want to answer in your review: * What is your evaluation of the design? * What is your evaluation of the implementation? * What is your evaluation of the documentation? * What is your evaluation of the potential usefulness of the library? * Did you try to use the library? With what compiler? Did you have any problems? * How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? * Are you knowledgeable about the problem domain? And finally, every review should answer this question: * Do you think the library should be accepted as a Boost library? Be sure to say this explicitly so that your other comments don't obscure your overall opinion. Thanks in advance to every for your participation in this review. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

3 5

Boost fixed point data type or other support for monetary values?
by Jeff Dunlap 23 Aug '07

23 Aug '07

Hello, I'm having some rounding problems using the double data type on some financial calculations. Does Boost have a fixed point data type or is there other functionality that may be helpful to me? Any suggestions would also be appreciated. Thanks very much

1 0

basic file io in boost 1_34_1. A compilable 'hello world' example for large (>2^31byte) files?
by Peter Leopold 23 Aug '07

23 Aug '07

Dear Boost Users, I've used boost on and off for years for its string parsing. Now I'd like to use it to replace std::ifstream for file reading. The immediate requirement is a file > 2^31 bytes. Since my requirement is to replace ifstream, I was hoping for a fairly simple transition, perhaps even something I could manage with an emacs global search-and-replace. I started with kubuntu-feisty's boost 1.33.1. but was unable to execute the first main in the iostreams tutorial Overview (2.1.1). The showstopper was the compiler error <file_sink> not defined in scope. % g++ test.cpp test.cpp: In functin 'int main()': test.cpp::9: error: 'file_sink' was not declared in this scope . . . several more errors. This is an error without precedent, at least according to google. Thinking this is a kubuntu-apt-get issue, I downloaded and rebuilt and installed boost 1_34_1 from scratch. Alas, <file_sink> is still not declared. Am I doing something obvious wrong? I'm using g++ 4.1.2. Does anyone have a few lines of code that will * open a file? * seekg() to the beginning, end and arbitrary location in the file? * tellg() the present file pointer location in the file? * read(address, number of bytes) from the file? These are the basic STL features I'm trying to replace. Many thanks for any pointers to useful documentation (but not to boost.org/libs/iostreams/doc/index.html, please) or code snippets. Once I have my "hello world" working, I might start making sense of <file_sink> templated streams, and why my install can't find <file_sink> when it is sitting right there in <boost/iostreams/device/file.hpp>!?!? -- Peter Leopold Peter dot Leopold at BioAnalyte dot com <http://www.bioanalyte.com>

2 1