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

Mutex scoped_lock compile errors
by Darren Vincent Hart 10 Jun '03

10 Jun '03

Esteemed Boost Gurus, I am trying to compile a method with a scoped_lock call. If I use a mutex object declared in the class implementation file (logger.cpp) it builds, if I use a mutex of the same type declared as a private member of the logger class (in logger.h), I get the following compiler error: logger.cpp: In member function `void pel::logger::info(const std::string&, const std::string&, int, const std::string&) const': logger.cpp:36: error: no matching function for call to ` boost::detail::thread::scoped_lock<boost::mutex>::scoped_lock(const boost::mutex&)' /usr/include/boost/thread/detail/lock.hpp:62: error: candidates are: boost::detail::thread::scoped_lock<boost::mutex>::scoped_lock(const boost::detail::thread::scoped_lock<boost::mutex>&) /usr/include/boost/thread/detail/lock.hpp:67: error: boost::detail::thread::scoped_lock<Mutex>::scoped_lock(Mutex&, bool = true) [with Mutex = boost::mutex] make: *** [logger.o] Error 1 The relevant method from the implementation file is as follows: void logger::info(const std::string& msg, const std::string& file, int line, const std::string& function) const { // io_mutex is declared above in the pel namespace boost::mutex::scoped_lock scoped_lock(io_mutex); // no problem // mutex is a private member of the logger class boost::mutex::scoped_lock scoped_lock2(mutex); // line 36, fails std::cout << "...blah..." << std::endl; } Can someone explain why these are treated differently by the compiler ? Thanks, Darren Hart dvhart(a)byu.edu

3 4

RE: [Boost-Users] Struggling with bind
by Jean-Marc Prud'homme 10 Jun '03

10 Jun '03

Hi Paul, _1 is a placeholder argument. When you write bind(&CDocumentAttachments::Execute, _1, _T("print")) you are creating a functor that can be called with a single argument of type CDocumentAttachments. This is what is required by the for_each algorithm. It is equivalent to bind<HINSTANCE>(mem_fn(&CDocumentAttachments::Execute), _1, _T("print")) Look at the mem_fn doc. You still need this kind of functor if you add more arguments to CDocumentAttachments::Execute. This should create the functor you need bind(&CDocumentAttachments::Execute, _1, _T("print"), false) Hope this helps, Jean-Marc -----Original Message----- From: Paul Thompson [mailto:friedegg@ntlworld.com] Sent: Tuesday, June 10, 2003 7:54 AM To: Boost-Users(a)yahoogroups.com Subject: [Boost-Users] Struggling with bind Hi all. I've been using boost::bind to work around deficiencies with std::mem_fun_ref and std::bind2nd under MSVC6. I have a class called CDocumentAttachments which had a member function thus: HINSTANCE Execute(LPCTSTR verb) const; I was using bind with this as follows: for_each(filenames.begin(), filenames.end(), bind(&CDocumentAttachments::Execute, _1, _T("print"))); "filenames" is a const vector of CDocumentAttachments objects. This worked fine. A little later on in development, I changed the function signature for CDocumentAttachments::Execute to: HINSTANCE Execute(LPCTSTR verb, bool visible) const; How will I pass the extra bool to bind? I can't find any examples of passing more than one parameter to a member function. I thought about (among many others) for_each(filenames.begin(), filenames.end(), bind(&CDocumentAttachments::Execute, _2, _T("print"), false)); but the compiler (VC6) complains with two of the following: k:\vc\include\boost\bind.hpp(760) : error C2679: binary '[' : no operator defined which takes a right-hand operand of type 'const class boost::arg<2>' (or there is no acceptable conversion) k:\vc\include\boost\bind\bind_template.hpp(33) : see reference to function template instantiation 'struct HINSTANCE__ *__cdecl boost::_bi::evaluator3<struct HINSTANCE__ *>::eval(const class boost::_bi::list3<class boost::arg<2>,class boost:: _bi::value<char *>,class boost::_bi::value<bool> > &,struct boost::_mfi::cmf2<struct HINSTANCE__ *,class CDocumentAttachments,char const *,bool>,class boost::_bi::list1<class CDocumentAttachments const &> &)' being compiled I know I'm being really thick here, and I'm sure there's something really obvious that I'm missing, but reading through the documentation hasn't helped me. Thank you very much. Paul -- Paul Thompson Programmer McLellan Software Design Ltd. Info: <http://www.boost.org> Wiki: <http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl> Unsubscribe: <mailto:boost-users-unsubscribe@yahoogroups.com> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

1 0

to_simple_string and Windows?
by adutoit 10 Jun '03

10 Jun '03

Hi. I have used the boost::date_time library in Linux with no problems, but the same can not be said of my experience with it in Windows... My current pain in Windows is with the "to_simple_string" method used to convert a boost::date_time value to guess what, a string. This works fine in Linux, but not in stupid Windows. The following line just borks: std::cout << to_simple_string(date(2003,4,30)) << std::endl; Any little unknown configuration tricks etc. I need to be aware of here? Has anyone used the "to_simple_string" method without problems in Windows? I'll appreciate any help in this regard. Greetings, A. [Non-text portions of this message have been removed]

2 3

New file uploaded to Boost-Users
by Boost-Users＠yahoogroups.com 10 Jun '03

10 Jun '03

Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the Boost-Users group. File : /uBLA Performance 2.pdf Uploaded by : sumohurd <matt(a)finray.net> Description : Simplistic uBLA Performance chart, second version You can access this file at the URL http://groups.yahoo.com/group/Boost-Users/files/uBLA%20Performance%202.pdf To learn more about file sharing for your group, please visit http://help.yahoo.com/help/us/groups/files Regards, sumohurd <matt(a)finray.net>

1 0

[crc] Simple example please
by Vladimir Prus 10 Jun '03

10 Jun '03

Hi Alex, > I'd like to calculate a value for a cyclic redundancy check (crc) for a > number of files. > > This isn't going to be transmitted anywhere, I simply want to compare the > values to see whether the files are exactly the same. I know I can simply > check byte for byte, but I've got a few thousand files and it's easier to > calculate the crc for each and see whether any are the same first. Here's the function that I use for the same purpose. Might not be optimal, but works: unsigned file_crc(const string& name) { #if defined(__GNUC__) && __GNUC__ < 3 ifstream ifs(name.c_str(), ios::binary); #else ifstream ifs(name.c_str(), ios_base::binary); #endif if (!ifs) return 0; else { using namespace boost; crc_32_type crc32; int c; while( (c = ifs.rdbuf()->sbumpc()) != -1) crc32.process_byte(char(c)); return crc32.checksum(); } } I believe some example of this kind should be included... Daryle, what do you think? HTH, Volodya

3 2

Lapack for uBlas
by Achim Domma 10 Jun '03

10 Jun '03

Hi, does somebody know a Lapack implementation which can be used with uBlas? I need a solver for an linear equation system and would prefer to not rewrite it by hand. Achim

2 3

vector ranges
by cerenoc 10 Jun '03

10 Jun '03

Hello, I just wanted to clarify if the following behavior is correct. As far as I understood, in matrix and vector range function, the 1st argument defines the starting point and the second defines the length of the range. The following code I ran on gcc 2.96 shows that the behavior is different when starting at 0 and starting somewhere else. Is this a bug, or am I misunderstanding something? Thank you, Julian // define some vector r(3) std::cout << r << "\n\n"; boost::numeric::ublas::vector<double> rr = r(range(0, 0)); std::cout << rr << "\n\n"; rr = r(range(0, 1)); std::cout << rr << "\n\n"; rr = r(range(0, 2)); std::cout << rr << "\n\n"; rr = r(range(0, 3)); std::cout << rr << "\n\n"; //rr = r(range(1, 0)); // THIS RESULTS IN ASSERTION ERROR //std::cout << rr << "\n\n"; rr = r(range(1, 1)); std::cout << rr << "\n\n"; rr = r(range(1, 2)); std::cout << rr << "\n\n"; rr = r(range(1, 3)); std::cout << rr << "\n\n"; OUTPUT: [3](0,1,2) [0]() [1](0) [2](0,1) [3](0,1,2) [0]() [1](1) [2](1,2)

3 2

confused by matrix assignment
by cerenoc 10 Jun '03

10 Jun '03

Hi, I'm sorry I am new to this and also not up on the math notation you guys are using. I am confused by the definitions of slices and strides. I am trying to perform a reasonably simple operation of e.g. assigning a vector v(2) to a (sub)matrix m(3,3). In matlab notation, for instance: m = 0 1 2 3 4 5 6 7 8 v = 15 16 m(2, 1:2) = v Resulting in m = 0 1 2 15 16 5 6 7 8 What is the best way to do this with uBLAS? Thank you very much Julian

2 2

RE: [Boost-Users] Re: Tree Object Needed!
by Jim.Hyslop 09 Jun '03

09 Jun '03

Hossein Haeri wrote: > Does the method you mentioned work for my problem in > which there is no ordering? If yes, please explain it > in more detail (with an example). Trees are generally used with ordered sequences, aren't they? If your data are not ordered, then how do you propose to insert them into the tree? What, well, order would they appear in the tree? If you provide a little more information about the problem domain, we can help you figure out the best container to use. -- Jim

4 8

RE: [Boost-Users] Re: Tree Object Needed!
by Mark Storer 09 Jun '03

09 Jun '03

I don't mean to come off like a hard case, and I may well be wrong but... This whole line of questioning has that "homework assignment" smell to it. The initial question was _highly_ school-related: > I need to have some template doing the jobs of an > ordinary BST (Binary Search Tree). Nothing that fancy, > but something at the level we all have been educated > in CS1/FCS1, or similar courses. Does boosts > repository contain such an object? The utter lack of background detail leads me to believe that this is just a toy problem... a homework assignment. And then there's the yahoo email address... most proffesional programmers have @mybusiness addresses, though that doesn't mean everyone here is a proffesional. Oh, and the final bit of oddness: > Why don't I write it myself? Simply, because I want to > optimise the time! How is waiting a week or so for some mailing list to hand you an answer "optimizing the time"? This has just been bothering me, I felt the need to speak out. If I'm wrong, then I offer my heartfelt appoligies. This may be what Mr. Hyslop was trying to dig at, I'm just not a patient or subtle person. And I don't mind appologizing for my mistakes (lots of practice, as you might imaging with this sort of behavior). --Mark Storer Software Engineer Cardiff Software #include <disclaimer> typdef std::disclaimer<Cardiff> Discard;

2 1