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

RE: [Boost-users] Re: Q: boost::format : can formatting variablea rguments lists be done ?
by Foster, Gareth 07 Mar '05

07 Mar '05

Could you do Logger<DBG> MyLogger; MyLogger.print(); ? -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Dan Dimerman Sent: 22 December 2004 13:03 To: boost-users(a)lists.boost.org Subject: [Boost-users] Re: Q: boost::format : can formatting variablearguments lists be done ? Richard Howells wrote: > Vladimir is right. > > You've seen the problems and crashes that leaving type checking to the > programmer causes. You need the compiler to do the checking for you. > > I suggest using the same kind of scheme as the iostreams library. You might /.../ > ...it will be (mostly) type safe - the type dangerous stuff is isolated to > one place per output type - and you can extend it as you go. You'll likely > need extra formatting stuff and that's where boost::format may help you. > I was thinking about something like: MyLogger::print( const boost::format& f ) and since all format's operator %() return a boost::format, I could with this single function reap all the types handled by it at once. I use a function rather than the operator << because they are actually template functions, with the parameter being the severity. So for _DEBUG versions, the MyLogger::print< DBG >() function instantiates to a proper call to print, while in _RELEASE versions it instantiates to an empty function. I just didn't see a clean way to do this with operators. Dan _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

3 2

Re: Logo Contest: Send In Your votes!
by Joel de Guzman 07 Mar '05

07 Mar '05

Hi, I have to apologize to Mike Mangone. For some reason, I missed his submission dated Feb 16. For those who voted already, please check out his entry (no. 99). If you find his entry compelling, I am willing to make corrections to my tally if any of you wish to resend your votes. Again, pardon me for this slip. http://boost.org/more/logo_contest.htm Send your votes to: boost.logo.contest(a)gmail.com Kind Regards, -- Joel de Guzman http://www.boost-consulting.com http://spirit.sf.net

1 0

Serialization library and memory issues
by Mike Lapshin 06 Mar '05

06 Mar '05

Hello Boost-users, I'm trying to save a simple class to xml_woarchive: class StringContainer { std::map<std::string, std::wstring> mStrings; friend class boost::serialization::access; template<class TArchive> void serialize(TArchive& ar, const unsigned int version) { ar & BOOST_SERIALIZATION_NVP(mStrings); } public: typedef std::map<std::string, std::wstring> StringMap; StringContainer() {} ~StringContainer() {} void addString (const std::string& id, const std::wstring& str); bool isStringExists (const std::string& id) const; const std::wstring& getString(const std::string& id) const; void removeString (const std::string& id); }; // somewhere in main code: StringContainer sc; sc.addString("fps.profilerfl", L"FPS: "); sc.addString("triangles.profilerfl", L"Triangles: "); std::wofstream outfs("strings.xml"); assert(outfs.good()); boost::archive::xml_woarchive oa(outfs); oa << BOOST_SERIALIZATION_NVP(sc); But somewhere after it (i think it is in xml_woarchive destructor) my program falls with "Debug Assertion Failed" (I'm using VC7.1) message: HEAP[test.exe]: Invalid Address specified to RtlValidateHeap( 011B0000, 011F29E8 ) Unhandled exception at 0x77f767cd in test.exe: User breakpoint. By the way, xml file writes as expected: <?xml version="1.0" encoding="UTF-8" standalone="yes" ?> <!DOCTYPE boost_serialization> <boost_serialization signature="serialization::archive" version="3"> <sc class_id="0" tracking_level="0" version="0"> <mStrings class_id="1" tracking_level="0" version="0"> <count>2</count> <item class_id="2" tracking_level="0" version="0"> <first>fps.profilerfl</first> <second>FPS: </second> </item> <item> <first>triangles.profilerfl</first> <second>Triangles: </second> </item> </mStrings> </sc> </boost_serialization> What i'm doing incorrectly? -- Mike Lapshin

2 2

boost::threads Am I on the right track?
by The Grumpiest Man You Know 06 Mar '05

06 Mar '05

I've had a minor epiphany, or at least I think so, in use of boost::threads. I would be very grateful if someone who actually knew something about C++ and boost's threads support would glance at the attached code. Any comment would be appreciated especially if there's sign that I'm horribly misusing the scraps I've picked up. I have to say that once I stopped trying to relate the pthreads C API to the bits of the docs I could understand my reaction was, "Sweet!" I think the way the threading and locking "disappear" is ... really elegant. Thanks very much for your time and again if I'm OT please tell me where to go, even if it is to hell. :) ------------ #include <iostream> #include <vector> #include <queue> #include <boost/thread/thread.hpp> #include <boost/function.hpp> using namespace std; using namespace boost; mutex::mutex io; struct Throttle { Throttle(void (*fn)(), int svcQLen = 5) : population(1), load(0), acceptable(svcQLen), service(fn) { }; ~Throttle() { cout << "load is " << load << endl; cout << "Population " << population << endl; } void itsBusy() { mutex::scoped_lock myPermission(use); if (++load > acceptable) { workers.create_thread(service); population++; load = 0; }; }; void observe() { workers.join_all(); }; private: mutex use; int population; int load; int acceptable; void (*service)(); thread_group workers; }; template <class T> struct RequestQ : public queue<T> { typedef T value_type; typedef queue<T> cBase; bool empty() { mutex::scoped_lock permit(use); return cBase::empty(); } value_type pop() { mutex::scoped_lock permit(use); value_type it = cBase::front(); cBase::pop(); mutex::scoped_lock doing(io); cout << "Q -> " << it << endl; return it; } void extend(unsigned int by) { mutex::scoped_lock permit(use); value_type next; if (cBase::empty()) { next = 0; // Yeah, assumes integral type conversion } else { next = cBase::back(); } while(by--) { push(*new value_type(++next)); } } private: void push(const value_type& it) { mutex::scoped_lock doing(io); cout << "Q <- " << it << endl; cBase::push(it); } mutex use; }; /* * We have to initialise this before creating any threads but that's * OK 'cos this IS the way we'll create threads. */ Throttle *manager; RequestQ<int> jobs; void go() { for(int range = 0; !jobs.empty(); range = 0) { int mine = jobs.pop(); if (!jobs.empty()) { manager->itsBusy(); } { mutex::scoped_lock doing(io); bool term = false; if (!(mine % 3)) { cout << "fizz "; range=1; term = true; } if (!(mine % 5)) { cout << "buzz "; range += 1; term = true; } if (!(mine % 7)) { cout << "pop! "; range = 0; term = true; } if (!(mine % 11)) { cout << "puff"; range = 2 * (range + 1); term = true; } if (term) { cout <<endl; } } if (range) { jobs.extend(range); sleep(range); } } } int main(int argc, char **argv) { manager = new Throttle(&go); jobs.extend(argc * 10); go(); manager->observe(); delete manager; } -- Blue Skies

1 0

boost serialization linking problems with certain binutils
by Daniel Goller 05 Mar '05

05 Mar '05

hi, im looking for feedback as well as a possible fix for boost serialization with binutils >2.15.90.0.* with binutils 2.15.90.0.* the boost serialization demo links properly with newer (stable on gentoo) binutils we see a myriad of gnu linkonce errors from the linker so i would like to know if anyone here can link the serialization demo.cpp with 2.15.92.* binutils if yes on which distro and the exact version of the binutils better yet, i would like to know how i can build boost serialization also as a shared lib, as none of the other libs which exist in shared and static version have problems with linking with 2.15.92.* bintuils thank you in advance for your input and help Daniel

2 4

Tricky lambda usage?
by Jason House 05 Mar '05

05 Mar '05

"In a nutshell" one can do the following: for_each(a.begin(), a.end(), std::cout << _1 << ' '); (where a is something like std::list<int>) Is there any trick to print out something like std::map<std::string, std::list<int> > ?

3 2

[Foreach]VC70 problem
by Martin 04 Mar '05

04 Mar '05

Just wanted to say that VC70 got problems with BOOST_FOREACH. Apparantly an internal compiler bug sometimes generates the wrong code and causes a "Run- Time check failure #2 - Stack around ... was corrupted". It has happened to me twice now in completly different contexts. In both cases there is no problem if I compile with VC71 or if I replace BOOST_FOREACH with a normal for loop. Unfortunatly I can't find any small example which reproduces the problem. In both cases I can see in the debugger that code uses the wrong stack offset for a local variable.

2 1

make_variant_over an mpl::vector: Compilation fails on gcc
by Johannes Dahlström 04 Mar '05

04 Mar '05

I ran into a bizarre problem trying to build a variant from an MPL vector. The compilation of make_variant_over seems to fail when given any vector which a) uses all its template arguments, and b) contains *only* user-defined types; ones with at least one built-in type work fine! The problem occurs at least with gcc 3.3.5 and gcc 3.4.4 20050203 prerelease, on Debian unstable/x86. The following code demonstrates the problem: //////////////////////////////////////////////////////////////////////// #include <boost/variant.hpp> #include <boost/mpl/vector.hpp> #include <cstdio> // a template for generating dummy types template <int i> struct d { }; typedef boost::mpl::vector < d<0>, d<1>, d<2>, d<3>, d<4>, d<5>, d<6>, d<7>, d<8>, d<9> > vector; boost::make_variant_over<vector>::type v; //////////////////////////////////////////////////////////////////////// The same thing happens with vector10 and others, but lists do seem to work correctly, providing a work-around. Swapping any built-in type for any of the d<>'s makes the code above to compile cleanly. I can post the complete compiler error output if you like, but there's rather much of it :p -Johannes

1 0

Any constructor evaluation in compile time?
by Yutaka Leon Suematsu 04 Mar '05

04 Mar '05

Hi everyone, I am new at the excellent Boost library. It is an incredible source of smart techniques and ticks. I would like to ask if there is a way to evaluate, in compile time, the existence of a specific constructor, except from the constructor default and copy constructor (I know the functions has_trivial_constructor and has_trivial_copy from the type_traits library). For example in the following code, I would like to evaluate the existence of both "A(const Config& config)" and " template <class Random> A(const Config& config, Random& random) " constructors. Class A { A(){.} A(const Config& config){.} template <class Random> A(const Config& config, Random random){.} . } Also, I would like to know why and how the has_trivial_constructor works. Thank you in advance for your kindly help. All the best, Yutaka Leon Suematsu.

6 9

uBlas matrix superclass?
by V. C. 04 Mar '05

04 Mar '05

I am considering using the uBlas library. After a couple of days of researching all available matrix packages for C/C++, I have come to the conclusion that it is the best one in terms of support, ease-of-use, performance(relatively speaking) and documentation. >From what I have been able to gather, I can't figure out whether there is any superclass to all the matrices.. I want to be able to have a list of matrices, some of which may be sparse and some dense... What type should that list be? Or is that impossible (at first glance it appears that inheritance is not really supported for efficiency reasons, hence the widespread templatization)... Thanks in advance for any help.

1 0