October 2008 - Boost-users - lists.preview.boost.org

Collections versus std::map
by Benson Margulies 17 Oct '08

17 Oct '08

I'm looking for a fast sparse array that permits ordered traversal of the occupied slots. std::map is much too slow. boost multi-index with a sequenced index on an unsigned int is significantly slower even that the std::map, which isn't too surprising given the writeup. Is there something else in boost that fills this space?

2 3

serialize a template class working with MSVC2005 but not gcc
by professor rumsdiegeige 17 Oct '08

17 Oct '08

Hello, I've got the following class definition: template< class T > class ProxyClass { boost::ptr_vector<T> &e_; ProxyClass( boost::ptr_vector<T> &f ) : e_(f) {;}; template<class Archive> void save(Archive & ar, const unsigned int version) const { size_t o=0; boost::ptr_vector< T >::iterator it = e_.begin(); /*for ( ; it != e_.end(); ++it, ++o ) { const T &t = *( it ); ar & t; }*/ } }; Where it says "boost::ptr_vector< T >::iterator it", the gcc compiler on Linux complains "error: expected ';' before 'it'", while Visual Studio 2005 compiles it fine without any errors. What can I do here? Thanks in advance for your help! S. Lorentz __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

cpp code and Boost.Interprocess compile error
by hamsinv 17 Oct '08

17 Oct '08

Hello All, I was looking for using stl maps in shared memory, when i came across Boost library. I have downloaded boost_1_36_0 version. In the examples directory, there is a cpp file for sharing stl maps named doc_map.cpp. When i compile the same with the following command: c++ -I $BOOST_ROOT doc_map.cpp -o shmmap I got the following errors: /tmp/ccEYJnov.o: In function `boost::interprocess::shared_memory_object::priv_open_or_create(boost::interprocess::detail::create_enum_t, char const*, boost::interprocess::mode_t)': doc_map.cpp:(.text._ZN5boost12interprocess20shared_memory_object19priv_open_or_createENS0_6detail13create_enum_tEPKcNS0_6mode_tE[boost::interprocess::shared_memory_object::priv_open_or_create(boost::interprocess::detail::create_enum_t, char const*, boost::interprocess::mode_t)]+0x163): undefined reference to `shm_open' /tmp/ccEYJnov.o: In function `boost::interprocess::detail::mutexattr_wrapper::mutexattr_wrapper(bool)': doc_map.cpp:(.text._ZN5boost12interprocess6detail17mutexattr_wrapperC1Eb[boost::interprocess::detail::mutexattr_wrapper::mutexattr_wrapper(bool)]+0x14): undefined reference to `pthread_mutexattr_init' doc_map.cpp:(.text._ZN5boost12interprocess6detail17mutexattr_wrapperC1Eb[boost::interprocess::detail::mutexattr_wrapper::mutexattr_wrapper(bool)]+0x2b): undefined reference to `pthread_mutexattr_setpshared' doc_map.cpp:(.text._ZN5boost12interprocess6detail17mutexattr_wrapperC1Eb[boost::interprocess::detail::mutexattr_wrapper::mutexattr_wrapper(bool)]+0x4d): undefined reference to `pthread_mutexattr_settype' /tmp/ccEYJnov.o: In function `boost::interprocess::detail::mutexattr_wrapper::~mutexattr_wrapper()': doc_map.cpp:(.text._ZN5boost12interprocess6detail17mutexattr_wrapperD1Ev[boost::interprocess::detail::mutexattr_wrapper::~mutexattr_wrapper()]+0xd): undefined reference to `pthread_mutexattr_destroy' /tmp/ccEYJnov.o: In function `boost::interprocess::shared_memory_object::remove(char const*)': doc_map.cpp:(.text._ZN5boost12interprocess20shared_memory_object6removeEPKc[boost::interprocess::shared_memory_object::remove(char const*)]+0x33): undefined reference to `shm_unlink' collect2: ld returned 1 exit status Can anyone please tell me what i'm missing in the compilation. Thank you. -- View this message in context: http://www.nabble.com/cpp-code-and-Boost.Interprocess-compile-error-tp20012… Sent from the Boost - Users mailing list archive at Nabble.com.

2 3

boost::numeric::interval - what are interval_holder and number_holder for?
by Stephan Puchegger 17 Oct '08

17 Oct '08

Dear boost members! Perhaps the main boost developer mailing list would be the better place to ask this question, but I'll try it here first: I cannot find any references to interval_holder and number_holder, which are defined in boost/numerics/interval/interval.hpp in any other files of the boost library. Are those two structures deprecated remains of a previous boost release? (The last time I checked they were still in trunk of the svn repository.) regards Stephan

1 0

[format] crash on windows / inquiry
by Tore Halvorsen 17 Oct '08

17 Oct '08

Hi, I just wondering about two things with the format library. 1) The following program crashes on windows - is this intentional? #include <boost/format.hpp> int main(int argc, char* argv[]) { boost::format f("%s %s"); f.exceptions(boost::io::no_error_bits); // not needed, but just in case char const* null_cstr = 0; char const* empty_cstr = ""; f % empty_cstr; // triggers debug assertion failure f % null_cstr; // triggers access violation return EXIT_SUCCESS; } The debug assertion failure only happens on a debug build, of course. I can understand the access violation on a null string, but that feature makes it unsuitable for a sprintf replacement. 2) Is there a way to determine the number of expected arguments in a format object? Looks like the basic_format::num_args_ is exactly what I'm looking for, but I can't find this exposed anywhere. Going through a loop and catching exceptions is one way of getting the number of expected arguments, but it feels wrong. -- Eld på åren og sol på eng gjer mannen fegen og fjåg. [Jøtul] <demo> 2008 Tore Halvorsen || +052 0553034554

1 0

shared_ptr singleton
by Alessandro Re 17 Oct '08

17 Oct '08

Hello, I'm getting some troubles with shared_ptr when trying to implement a singleton pattern. Actually, it's not the singleton itself (which seems to work, in a separate program), but in the context I need it, doesn't work. I tried to figure out why (and actually I found an alternative implementation which works), but still I'm trying to understand what goes wrong. I'm trying to do this: having a hierarchy of classes where there is one Base class and a number of derived ones, each class is in a different cpp file. All these classes must be registered in a Register class, which is itself derived from Object. Register is the singleton, implemented like this: class Register; typedef shared_ptr<Register> RegisterP; class Register: public Object { public: Register(); void add(string name) { // Stuff with name } static RegisterP global() { if (!_global) _global = RegisterP(new Register); // * return _global; } private: static RegisterP _global; }; To register the various classes, does exist a special class, not in hierarchy, like this: class ClassRegisterer { public: ClassRegisterer::ClassRegisterer(string name) { Register::global()->register(name); // ** } } #define REGISTER(C) ClassRegisterer _registerClass_ ## C ( #C) In each file of the of the project, this macro is used when a class is created: Foo.cpp class Foo: public Object {...}; REGISTER(Foo); Bar.cpp class Bar: public Object {...}; REGISTER(Bar); This is also used for the Register class itself: Register.cpp class Register: public Object {...}; REGISTER(Register); The problem is that global() is called N times, but the new Register (*) is executed twice, instead of once. Investigating with some couts, I found out that - dunno why - the pointer _global is set to 0 after the first execution, in fact at the time of second execution it's created another object, which isn't deleted and is kept. Trying quickly with a debugger, it seems that at (**) a destructor for shared_ptr is called, and I think that it may be the cause of the issue, but I'm having an hard time figuring this out and *why* a destructor is called (maybe it's just for a temporary object). Looking at boost docs, i found out this: http://www.boost.org/doc/libs/1_35_0/libs/smart_ptr/sp_techniques.html#stat… so I tried to use that null deleter, like this: struct nullDeleter { void operator()(void const *p) const { cout << "Trying to delete " << p << endl; } }; and used like: if (!_global) _global = RegisterP(new Register, nullDeleter()); The problem is that this is never called! The object is never destroyed by this, but somehow the static _global is set to 0 (only once), and thus allocated twice. I temporary solved this using a static raw pointer, which is allocated once and a shared_ptr is built with it: class Register { ... static RegisterP global() { if (!_global) _global = new Register; return RegisterP(_global, nullDeleter()); } ... static Register *_global; }; Which works. Any suggestion is welcome :) Thanks -- ~Ale

4 6

1.36.0 boost/random/uniform_01 Infinite[?] Loop
by Dave Steenburgh 17 Oct '08

17 Oct '08

I'm sure everyone will think I'm crazy, but here goes... Background: I came across this problem while developing a proof of concept for my master's project. It entails generating random numbers according to either a uniform or a normal distribution. I wrote a few wrapper classes to simplify dealing with the generators and distributions. Those wrapper classes let me do this: randomGenerator *rg; if(uniformDistribution) rg = new uniformGenerator(); else rg = new normalGenerator(); (I also have constructors that take arguments to alter the range of values, but I'm not using them presently.) Then I can retrieve values like so: (*rg)() uniformGenerator and normalGenerator inherit from randomGenerator, and are wrappers for these: boost::variate_generator<boost::minstd_rand&, boost::uniform_real<> > boost::variate_generator<boost::minstd_rand&, boost::normal_distribution<> > I currently have two programs that use those wrapper classes. One is fairly minimal; it simply parses some command line arguments that affect the distribution and quantity, then retrieves the values and prints them to stdout. The other program is my proof of concept; it does the same argument parsing (with a few more options), then makes use of the values to run my simulation, which is mostly manipulating STL containers. Problem: Now we come to the crazy part. I work mostly on one of two Windows machines, using g++ under Cygwin. On either machine, both of those programs run to completion every time, if I compile with -O1, -O2, or -O3. However, if I use -O0 (or just no -O argument), then the simulation locks up. The other program---that just prints random numbers---never locks up. (Incidentally, this program was my attempt to create a minimal program that replicated the behavior.) By single-stepping with gdb, I determined that it gets stuck in boost/random/uniform_01.hpp: 58 for (;;) { 59 result_type result = result_type(_rng() - (_rng.min)()) * _factor; 60 if (result < result_type(1)) 61 return result; 62 } gdb shows that result is always equal to 1, which of course leads to an infinite loop. Single-stepping through all of the code directly related to the random generators shows the same sequence of instructions for both programs. If I use the uniform distribution, it gets stuck when I try to retrieve the second value. If I use the normal distribution, it gets stuck on the third retrieval. Also worth noting, by default the generators are supplied with a different seed value every time I run either program. Getting back to the programming environment, when I build on the Solaris machines in my school's CS department, the program never gets stuck. For my part, it's exactly the same source, and I still use g++. I have not been able to coerce any warnings or errors from the compiler. So, I'm all out of ideas. I'm open to suggestions for fixing this, but I'm hoping the solution is less extreme than compiling gcc from source.

3 5

Problem in compiling simple boost mpl program
by Missing Rainbow 16 Oct '08

16 Oct '08

Hi All, I am making use of boost 1.32.0 (as given in the MPL book CD). I have written a small program using mpl::vector. I am getting the following error. What could be the problem? (begin-quote) lotus 241 $ CC -I$BOOST_ROOT -c hello.c "/backup/dev_area/anna/mplbookcd/boost_1_32_0/boost/mpl/if.hpp", line 63: Warning (Anachronism): Type names qualified by template parameters require "typename". "/backup/dev_area/anna/mplbookcd/boost_1_32_0/boost/mpl/if.hpp", line 72: Where: While specializing "boost::mpl::if_<boost::mpl::T1, boost::mpl::T2, boost::mpl::T3>". "/backup/dev_area/anna/mplbookcd/boost_1_32_0/boost/mpl/if.hpp", line 72: Where: Specialized in non-template code. "/backup/dev_area/anna/mplbookcd/boost_1_32_0/boost/mpl/aux_/preprocessed/no_ttp/plus.hpp", line 148: Warning (Anachronism): Type names qualified by template parameters require "typename". "/backup/dev_area/anna/mplbookcd/boost_1_32_0/boost/mpl/aux_/preprocessed/no_ttp/minus.hpp", line 148: Warning (Anachronism): Type names qualified by template parameters require "typename". "/backup/dev_area/anna/mplbookcd/boost_1_32_0/boost/mpl/vector/aux_/iterator.hpp", line 92: Error: Partial specialization parameter m_ is not used in the arguments. "hello.c", line 4: Error: mpl is not defined. "hello.c", line 4: Warning (Anachronism): Using vector as a template without a declaration. "hello.c", line 4: Error: vector is not defined. 3 Error(s) and 4 Warning(s) detected. lotus 242 $ cat hello.c #include "boost/mpl/vector.hpp" typedef mpl::vector<int, char*, double& > seq; lotus 243 $ CC -V CC: Sun C++ 5.5 2003/03/12 lotus 244 $ uname -a SunOS lotus 5.10 Generic_125100-09 sun4v sparc SUNW,Sun-Fire-T200 lotus 245 $ (end-quote) Rgds, anna

2 1

MPL: Lazy Evaluation and Nullary Metafunction
by Missing Rainbow 16 Oct '08

16 Oct '08

Hi All, I am reading about the lazy evaluation and nullary metafunction in the MPL book. While discussing the lazy evaluation technique the following code is presented (page 57): struct add_pointer_f { template <class T> struct apply : boost::add_pointer<T> {}; }; typedef mpl::vector<int, char*, double&> seq; typedef mpl::transform<seq, boost::add_pointer<_> > calc_ptr_seq; The point of the above code was to demonstrate that the nullary metafunction calc_ptr_seq represents the sequence "int*, char**, double&*", even though the last element in that sequence is not valid. Thanks to lazy evaluation. As long as we don't access the third element in the sequence represented by calc_ptr_seq, everything should work fine. Is my understanding correct? Also, other than nullary metafunction, is there any other approach to lazy evaluation? Or will lazy evaluation _always_ be implemented with the help of nullary metafunction? Thank you. Rgds, anna

2 1

help downloading Version 1.31.0
by Vishakh Paul Thomas 16 Oct '08

16 Oct '08

Hi I'd like to download boost Version 1.31.0 as that's what my app used earlier. We are porting the app to Linux from solaris and I do not want to try using the latest one as I am unaware of the dependencies. I looked around on boost.org and was unable to find any link to download it. Can you please let me know where I can find it? Thanks! Vishakh

2 1