December 2010 - Boost-users - lists.preview.boost.org

[Units] Converting from angular_velocity to degrees_per_second
by Andrew Hundt 01 Dec '10

01 Dec '10

I need to make a conversion from a general angular_velocity to degrees/second. For example: boost::units::quantity< boost::units::si::angular_velocity> m_speed((30.0*boost::units::si::radians_per_second)); uint32_t result = static_cast<uint32_t>(boost::units::quantity<boost::units::si::angular_velocity,uint32_t>(m_speed*boost::units::degree::degrees/boost::units::si::seconds).value()); However it fails to compile in gcc 4.4, reporting: error:incomplete type '[type removed]' used in nested name specifier Could someone explain the proper syntax to achieve the desired type specification and subsequent conversion? Do I need to specify a degrees_per_second unit, and if so, how? Thank you in advance for your help. Cheers! Andrew Hundt

2 3

[interprocess] Named mutex
by Chard 01 Dec '10

01 Dec '10

Ion, anyone, I'm trying to use a named mutex to control access to a file. The example given here: http://www.boost.org/doc/libs/1_45_0/doc/html/interprocess/synchronization_… seems to be a starting point. However, I'm not exactly sure what this example is demonstrating. For example: - it appears to be attempting to control access to a file called "file_name". However, streams are opened on the file so, AFAICT, if interleaving calls from two processes stream out to it, the outcome will be unpredictable (even under controlled lock, due to the file-pointer position) ? - the file is thrown away at process exit; what would this mean if the other process is accessing it? - the 'mutex_remove' object causes removal of the named mutex on construction. I presume this is to demonstrate a clean up of a residual named mutex if a previous execution did not exit cleanly. However, if process 1 starts up and executes until just after the mutex creation, then process 2 starts up and executes the mutex_remove, the mutex appears to be in an invalid state, and the scoped_lock calls have no effect - this is certainly the case on Windows. This would defeat the whole point. Any clarification appreciated...

1 0

[spirit] This parser hang
by Kamil Zubair 01 Dec '10

01 Dec '10

This code: std::string s("blablabla"); std::string::iterator start = s.begin(); std::string::iterator finish = s.end(); bool result = phrase_parse(start, finish, +(!char_(':')), space); It's just a simple parser that match string which doesn't have ':' character in it right ? But when I run it, function phrase_parse is hang, it never return anything. I'm using boost 1.44 with VC 2008

3 4

[range] Tuple initialization?
by Marcin Zalewski 01 Dec '10

01 Dec '10

I have a following fragment of code: local = boost::begin(f(*current)); local_end = boost::end(f(*current)); local and local_end are two iterators, f is a function that returns a range, and current is some iterator I have around. I would like to initialize two iterators from the range returned by f, but it seems that I only have two choices: create a local range variable or call f twice and have two temporary ranges. Am I missing something, or are these indeed the only choices? Would it be reasonable to have a begin_end (or such) function that returns a tuple of iterators from a range for tuple initialization of two iterators? Currently, the range concepts do not require anything of the sort. Thanks, Marcin

1 0

Valgrind false positive in make_shared
by Dave Steffen 01 Dec '10

01 Dec '10

> Date: Thu, 18 Nov 2010 19:51:55 +0200 > From: "Peter Dimov" <pdimov(a)pdimov.com> > > Dave Steffen wrote: > > I got this bad behavior with GCC 4.4.4. It goes away with GCC 4.2.4. > > What is the definition of std::forward in GCC 4.4.4? What happens if you > > change boost::detail::sp_forward in make_shared.hpp:150: > ::new( pv ) T( boost::detail::sp_forward<Arg1>( arg1 ), > > boost::detail::sp_forward<Args>( args )... ); > > to std::forward, do you get a compile warning on this line and does > Valgrind still warn? Hi Coming back to this issue after a week or two... trying the above made no difference: no compiler warnings, and no change in Valgrind's error messages. Any thoughts? Thanks! -- Dave Steffen, Ph.D. - Software Engineer Numerica Corporation <http://www.numerica.us> 4850 Hahns Peak Drive, Suite 200 Loveland, Colorado 80538 main (970) 461-2000 x 227 direct (970) 612-2327 Email: dave.steffen(a)numerica.us fax (970) 461-2004 This message and any attachments are intended only for the individual or entity to which the message is addressed. This is a private message and may contain privileged information. If you are neither the intended recipient nor the agent responsible for delivering the message to the intended recipient, you are hereby notified that any review, retransmission, dissemination, or taking of any action in reliance upon, the information in this communication is strictly prohibited, and may be unlawful. If you feel you have received this communication in error, please notify me immediately by returning this email to me and deleting it from your computer.

2 4

[BGL] Searching for a convenient way to combine predicates for filtered_graph
by Cedric Laczny 01 Dec '10

01 Dec '10

Hi, I was wondering if there is a convenient (and to some extent intuitive) way to combine several predicates used in filtering a graph? The idea is to have some predicates defined and arbitrarily combine them so that the filtered_graph will check for compliance of each individual predicate and either make this vertex/edge visible or not. Of course, the predicates should be specific to vertices or edges, respectively, when being combined. When using std::vector or such, they would need to be all of the same type which does not seem very nice/feasible IMHO. Also defining a "big" predicate having a multitude of "smaller" predicates as members is not really an option as this is very restricted. Something like big_predicate = predicate1 || predicate2 || predicate3 (syntax should just illustrate the idea) maybe? Best, Cedric

3 9

[ serialisation ] 1.45 portable binary archive fails for shared_ptr on AIX/v11.1
by Avi Bahra 01 Dec '10

01 Dec '10

// This file is used to show problem with the portable binary archive(1.45) // when using the AIX/v11.1 compiler // // This test passes on Linux:gcc-4.2.1/HPUX:acc but crashes on AIX. // #include <assert.h> #include <iostream> #include <fstream> #include "portable_binary_oarchive.hpp" #include "portable_binary_iarchive.hpp" #include <boost/serialization/string.hpp> #include <boost/serialization/shared_ptr.hpp> #include <boost/serialization/weak_ptr.hpp> class CLimit { public: CLimit(const std::string& name) : name_(name) {} CLimit() {} const std::string& name() const { return name_;} private: std::string name_; friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int) { ar & name_; } }; class InLimit { public: InLimit(const std::string& name,boost::shared_ptr<CLimit> l) : name_(name), limit_(l) {} InLimit() {} const std::string& name() const { return name_;} CLimit* limit() const { return limit_.get();} private: std::string name_; boost::shared_ptr<CLimit> limit_; friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int) { ar & name_; ar & limit_; } }; int main() { std::string fileName= "pba.txt"; { const InLimit saved; std::ofstream ofs( fileName.c_str(), std::ios::binary); portable_binary_oarchive oa(ofs); oa << saved; std::cout << "saved OK\n"; } { InLimit restored; std::ifstream ifs( fileName.c_str(), std::ios::binary); portable_binary_iarchive ia( ifs ); ia >> restored; std::cout << "restored OK\n"; } return 0; } // ---------------------------------------------------------------- Best regards, Ta, Avi

1 0

Re: [Boost-users] [Bind] Retrun-type of bind expressions (SOLVED)
by Cedric Laczny 01 Dec '10

01 Dec '10

Hi, I found out how to specify the return type using boost::function and boost::bind, basically thanks to http://stackoverflow.com/questions/1044703/how-do-you-pass-boostbind-object… to-a-function/1044732#1044732 Say we are given two function objects (DefaultConstructible, CopyConstructible and only having "bool operator()"). As a little side note, it might be necessary to include "typedef bool return_type" in the definition of the functions objects (struct, class) (s. http://www.boost.org/doc/libs/1_45_0/libs/bind/bind.html#with_function_obje…) Then the return type can be declared as follows: typedef boost::function< bool( ARG-TYPE ) > CombinedPred; CombinedPred combined_pred = boost::bind(functor1, _1) && boost::bind(functor2, _1) where ARG-TYPE is the input paramter type the operator() was defined for in the functors (e.g. int, if operator() is working on ints). The basic idea is that functors that "return" bool and have the same input argument type will also have the same type(s) when being combined, e.g. by logical operators. Other combinations are of course also possible as boost::bind provides several overloaded operators like "&&", "||", "!" etc. Best, Cedric

1 0

Re: [Boost-users] Understanding Threads: Slow on Win32 with DLL
by Simon Adler 01 Dec '10

01 Dec '10

Hello Nat, Yes - i am sure that every Thread is just computing the reduced number - i tested it more than once. They are not producing log output. I have a timer to measure the time basing on the Frequency Performance Counter. After 5 seconds of runtime I do an output of the average computation time. Will now try to eliminate the dll - have a little luck that the required class has no dependencies, so it will take not long - will show the results shorty. Simon

1 0

[BGL] reverse_graph and filter_graph does not compile on 1.45
by gast128 01 Dec '10

01 Dec '10

Hello all, we got here a situation on VS2008 which did compile on 1.44 with a reverse graph of a filtered graph. I get constantly a f*cking 'You seem to be top-posting. Don't do that.' error from gmane, so I give up to include a reproducable case. It's easy to do it urself by making a construct like 'boost::make_reverse_graph (fgraph), where fgraph is a fitlered graph.' Anyone?

2 4