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

[multiindex] key extractor as free function - how?
by Markus Werle 23 Jan '07

23 Jan '07

Hi! I read from the docs that we can define key extractors to be one of - identity - member - mem_fun - composite_key Is it possible to use a free function as key extractor in the same way as the ones listed above and can you give an example of this? reagrds, Markus

3 6

Where to get boost 1.32 for solaris c++
by Victor Cheng 23 Jan '07

23 Jan '07

It seems that sun c++ has problems compiling boost 1.33.1 and the most compatible version is 1.32, but I could not find it any where on www.boost.org. It's probably in the cvs repository, but can't find the tag for 1.32. Is there a place where I can download this version?

2 1

[multiindex] combination with boost::tuple
by Markus Werle 23 Jan '07

23 Jan '07

Hi! I'd like to use boost::tuple in a list-replacing multiindex and use the first entry in the tuple as key. The following fails to compile. Could you please provide a correction? #include <boost/tuple/tuple.hpp> #include <boost/multi_index_container.hpp> #include <boost/multi_index/sequenced_index.hpp> #include <boost/multi_index/key_extractors.hpp> namespace bm = ::boost::multi_index; class C { typedef boost::tuple<std::string, std::string> data_t; template <typename T> struct mutable_value { mutable_value(const T& t):t(t){} operator T&()const{return t;} private: mutable T t; }; typedef bm::multi_index_container <mutable_value<data_t>, bm::indexed_by <bm::sequenced<>, bm::ordered_unique <bm::mem_fun<data_t, std::string, &data_t::get<0> > > > > storage_t; storage_t storage_; }; int main() { C c; }

2 4

compiling boost using codewarrior for windows
by Jean-Marc 23 Jan '07

23 Jan '07

when compiling boost ( the cvs version ) with metrowerks codewarrior for windows, I get a huge amount of errors. Is this (great !) compiler supported by boost ? Many thanks for your help Regards JM

1 0

[pool] fixed-size allocation pool
by Michael Marcin 23 Jan '07

23 Jan '07

I was wondering if I could use boost::pool for a fixed-sized object pool and return NULL if it tries to construct more than the maximum. Reading http://www.boost.org/libs/pool/doc/implementation/object_pool.html it seems like I could if I set_next_size(0) but it explicitly says I should not do that. Is there a reason to not support this behavior or is it supported in some other way/library? Thanks, Michael Marcin

1 0

Re: [Boost-users] [general question] on threading and doublecheckedlocking pattern
by Sohail Somani 23 Jan '07

23 Jan '07

It definitely is. But if you could find a way to make it better, even with complicated and platform-dependent code, and then wrap it up into something usable, have it tested and peer-reviewed, and placed somewhere we could all use it, it would still be useful. -------------- Sure -------------- P.S. In this particular case, I don't think you can get better than: - a read barrier and if-branch in the already-initted case - that + a write barrier in the init case - all wrapped into a single int. which is slightly, but not much better than a mutex. It is also slightly better than most implementations of boost::once/pthread_once, and much better than some, depending on platform. -------------- What I've found that pthread_once implementations end up using DBCL anyway but the benefit of using it there is obviously that you don't have to debug it since they put in the necessary barriers. As far as I'm concerned the problem is solved. For the 0.0001% of C++ programmers that actually need the performance and are willing to shell out the debugging hours, they are more than welcome. I (personally) think a boost singleton based on boost once would be good enough for everyone else until there are solutions per os/hw combination. Do you think people would like one? Not to take away from your valid points of course. Sohail

1 0

Re: [Boost-users] [general question] on threading and double checkedlocking pattern
by Sohail Somani 22 Jan '07

22 Jan '07

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of > Ovanes Markarian > > I might be wrong here, but I would like to hear some thoughts > of C++ professionals. > > I read in lots of sources about DCLP and how it is broken in > conjunction with Singleton pattern > and my question is of a very simple nature. Is it possible to > solve this problem by introducing > function calls via function pointers? Or is there still a > way, that these calls might be inlined > and if yes then how could this look like? This example is not > about freeing the singleton, it is > just about creating it. Hi, Perhaps the path-of-least-debugging is to use boost::once? Sohail

2 1

[general question] on threading and double checked locking pattern
by Ovanes Markarian 22 Jan '07

22 Jan '07

I might be wrong here, but I would like to hear some thoughts of C++ professionals. I read in lots of sources about DCLP and how it is broken in conjunction with Singleton pattern and my question is of a very simple nature. Is it possible to solve this problem by introducing function calls via function pointers? Or is there still a way, that these calls might be inlined and if yes then how could this look like? This example is not about freeing the singleton, it is just about creating it. class Singleton; typedef Singleton* (*singleton_getter)(); class Singleton : boost::noncopyable { static Singleton* pInstance; static singleton_getter getter; //other internal data... Singleton(); static Singleton* creating_singleton_getter(); static Singleton* non_creating_singleton_getter(); public: static Singleton* instance(); }; //implementation Singleton* Singleton::pInstance = NULL; singleton_getter Singleton::getter = &creating_singleton_getter; Singleton::Singleton() {} Singleton* Singleton::creating_singleton_getter() { //aquire mutex here if(Singleton::pInstance==NULL) Singleton::pInstance = new Singleton; Singleton::getter = &non_creating_singleton_getter; return Singleton::pInstance; } //mutex unlocked Singleton* Singleton::non_creating_singleton_getter() { //no mutex anymore return Singleton::pInstance; } Singleton* Singleton::instance() { return (Singleton::getter)(); } With Kind Regards, Ovanes Markarian

2 1

[date_time] Misparses POSIX timezone specs
by Stephen Turner 22 Jan '07

22 Jan '07

If I'm being stupid, tell me; but I believe the date_time library misreads all POSIX timezone specs by swapping + and -. It uses negative for west of Greenwich, but POSIX (admittedly unintuitively) uses negative for east of Greenwich. Compare this Linux box: [chiark:~] (1)% date Fri Jan 19 16:05:53 GMT 2007 [chiark:~] (2)% setenv TZ PST8 [chiark:~] (3)% date Fri Jan 19 08:06:04 PST 2007 with this code which wraps the date_time library (the internals aren't important): DWORD dw1 = 1149163201; // 1 June 2006 12:00:01 GMT CMyTime::set_timezone_posix("PST8"); CMyTime t = dw1; h = t.GetHour(); // h = 20 -- Stephen Turner, Cambridge, UK http://homepage.ntlworld.com/adelie/stephen/ "The internet is a reflection of our society. If we do not like what we see, the problem is not to fix the mirror, we have to fix society." (Vint Cerf)

2 1

[bind] Problem compiling when passing instantiation of class
by Paul Giaccone 22 Jan '07

22 Jan '07

I am using Boost bind in conjunction with Boost unit test, as I have many times before, but I can't get my code to compile. I've looked at http://boost.org/libs/bind/bind.html#Troubleshooting but don't see anything that covers my case. Here is the line that is failing (with disguised variable names): test_suite -> add(BOOST_TEST_CASE(boost::bind(my_test, obj1, obj2, obj3, 1.0, obj4, 10, 1E-9, results))); and here is the error that Microsoft Visual C++ .NET is generating: main.cpp(152): error C2665: 'boost::bind' : none of the 2 overloads can convert parameter 4 from type 'Obj3Type' The variables obj3 and obj4 are of type Obj3Type and Obj4Type respectively (see definitions below): The function my_test is defined as: void my_test(boost::shared_array<Vector>& a, boost::shared_array<Vector>& b, Obj3Type& c, const double d, const Obj4Type& e, const unsigned int f, const double g, boost::shared_array<std::pair<double, std::valarray<double> > >& h) where Obj3Type and Obj4Type are defined as: class Obj3Type { bool var1; unsigned int var2; unsigned int var3; double var4; double var5; MyType var6; boost::scoped_array<int> var7; boost::scoped_array<boost::scoped_array<double> > var8; }; struct Obj4Type { MyType var1; double var2; double var3; double var4; double var5; double var6; Vector var7; Vector var8; double var9; bool var10; } where MyType is an enumerated type and Vector is another struct of mine. (Where I have used structs rather than classes, this is for convenience.) Note that obj4 does not generate a compiler error, even when I move it to appear before obj3 in the parameter list. I wondered if it was necessary to overload operator= for Obj3Type as, presumably, the compiler cannot generate it for the scoped arrays in that struct but can for everything in Obj4Type, including Vector, for which I have defined operator=. I wrote an operator= for Obj3Type, but it made no difference to the compiler error. This was to be expected really, as obj3 is passed by reference rather than by value. Does it have anything to do with using structs instead of classes? I can't see how that could be the case. So what am I missing here? Thanks for any help.

2 2