- Boost-users - lists.boost.org

Re: [Boost-users] linker problem with regex library
by Oliver.Kowalke＠infineon.com 07 Nov '05

07 Nov '05

Sorry - I forget: os. HP/UX 11.11, gcc 3.4.2 >Hello, >I get the linker error (see below) if I try to compile my test app. >What is the problem? > >/usr/ccs/bin/ld: Invalid loader fixup in text space needed in output >file for symbol "__gxx_personality_sj0" in input file >"/opt/boost/lib/libboost_regex-gcc. > >Best regards, >Oliver > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

Can't Compile store() in Program Options
by Merrill Cornish 07 Nov '05

07 Nov '05

I am trying to use the Boost 1.32.0 program_options library for the first time with MGWin. Since what I need is identical to what is in the Getting Started section of the program_options Tutorial, I copied that. Here is the code from the tutorial: po::options_description desc("Allowed options"); desc.add_options() ("help", "produce help message") ("compression", po::value<int>(), "set compression level") ; po::variables_map vm; po::store(po::parse_command_line(ac, av, desc), vm); po::notify(vm); The problem occurs in the store() function. The compiler says it doesn't have a store() to match that call and suggests two possibilities. Examining the .hpp files, it appears that the two argument version of store() defaults to using wchar_t, which doesn't seem right. So, I tried po::store(po::parse_command_line(ac, av, desc), vm, true); to get UTF8 (simple char). The next complaint is apparently about the parse_command_line() call as the first argument. If I pull parse_command_line() out into a separate assignment statement to an intermediate variable, then put its results into store(), that could/would work except there now seems to be a conflict between the template types of char versus const char. I can get the assignment of parse_command_line() to an intermediate result to compile, but store() fails because of mismatched type problems. Or, I can get store() to compile, but the assignment fails for mismatched type problems. I apologize for not having the specific error messages here since the machine I'm compiling on is not available right now. Merrill P.S. About the Documentation: The program_options library document looks slick and very professional. Furthermore, it obviously took some time to create with all of the examples. However, I've never seen programming library documentation that didn't actually list the methods and functions in the library. As it is, if you are lucky, the method you need to mentioned, in passing, in an example. However, looking in the hpp and cpp files shows there is a lot more there.

2 1

Problem with program_options store()
by Merrill Cornish 07 Nov '05

07 Nov '05

Given this program fragment (based on the Getting Started example) using Boost::program_options compiled with MinGW: options_description desc("Program options"); ... variables_map vm; store(parse_command_line(argc, argv, desc), vm); gave the following error (reformatted for readability): error: no matching function for call to `store(boost::program_options::basic_parsed_options<const char>, boost::program_options::variables_map&)' candidates are: store(const boost::program_options::basic_parsed_options<char>&, boost::program_options::variables_map&, bool) store(const boost::program_options::basic_parsed_options<wchar_t>&, boost::program_options::variables_map&) I can add a "true" as a third argument to store(), but the mismatch still remains in the first argument: <const char> versus <char>&. If I assigned the result of parse_command_line() to an intermediate variable, I can match the &, but the const char versus char still remains. What is the problem way of calling store()? Merrill

2 1

Profiling Boost with gprof
by Dave 07 Nov '05

07 Nov '05

Hello All, I'm trying to profile Boost Graph examples using gprof, but can't get any reasonable results. My system is Mac OS X 10.3.9 and gcc version 3.3. I compile programs with -pg option, but the profile does not seem to contain any useful information. It only shows "no time accumulated", or 100% for main. What am I doing wrong? Thanks for your help! Dave __________________________________ Yahoo! Mail - PC Magazine Editors' Choice 2005 http://mail.yahoo.com

3 2

library planed for expression evaluation?
by Oliver.Kowalke＠infineon.com 07 Nov '05

07 Nov '05

Hello, does some plans for an library for expression evaluation exist? The user should be able to define its own operators (unary and binary with priority/affinity) and operands. The user can then store a sequence of its operators and operands in a vector in infix order. A special function creates from the vector an expression tree and evaluates the tree (calling the implemented action of each operator/operand -> the action can depend of the subsequent/child operators). With regards, Oliver

1 0

Threads - suggestion for producer/consumer pattern - msg_queue (was ThreadPool)
by Oliver.Kowalke＠infineon.com 07 Nov '05

07 Nov '05

Hello, the email with the 'ThreadPool' subject requested an implemention of the synchronization channel between producer and consumer threads. I've a suggestion (adopted from several books) - the only problem is how to tell the consumer threads, that all producer have finished their jobs. You can signal it from outside (deactive member function) or each producer thread has to register on the message_queue (and unregister if it has finished its job). With best regards, Oliver <<test.cpp>> <<message_queue.h>>

1 0

When will boost::socket library issue?
by huangruiguang＠kingnet.biz 07 Nov '05

07 Nov '05

1 0

Translating an absolute path to a path relative to the current directory
by Hermann Rodrigues 06 Nov '05

06 Nov '05

Hi, I am running into the following problem these days: I need to translate an absolute path to a path relative to the current directory on Windows. The translation process need to include all required ".." directory specifiers in the path to make sure it can be accessed from the current directory. The following path samples may clarify the problem. -------------------------------------------------------------------------------- Current Directory: C:\x\y\z Absolute Path: C:\x\y\z\k\w.txt Translated Path: k\w.txt (C:\x\y\z is a common parent) Current Directory: C:\x\y\z Absolute Path: D:\a\b\c\d.txt Translated Path: D:\a\b\c\d.txt (no common parent found) Current Directory: C:\x\y\z Absolute Path: C:\x\k\w.txt Translated Path: ..\..\k\w.txt (C:\x is a common parent) -------------------------------------------------------------------------------- Is there any built-in support in the boost filesystem library to achieve such translation? At first I thought the path::relative_path could solve my problem but it does not seem to take into account the current directory, only the path itself. Any sugestion? Thanks in advance. Hermann

2 2

Re: [Boost-users] [date-time] posix timezone string
by Stering Wolfram 04 Nov '05

04 Nov '05

[I am receiving this list in digested form, so this reply might not be correctly "linked" to its parent, I am afraid.] On Thu, 03 Nov 2005 13:25:07 +0100, Stering Wolfram wrote >> I am implementing a year_based_date-derived date generator to >> calculate the easter sunday for a given year. Now, this is the easy >> part, and it is done already. > Great! >> But I wonder how to generate a meaningful posix timezone string that >> is expected to result from calling the generator's to_string() member. >> >> The documentation at the (not well formatted, b.t.w.) "Description" >> section of class posix_time_zone didn't give any hints for Easter >> Sunday either. >> >> Any ideas? > I'm afraid I need a bit more context -- are you trying to generate a > local_time? In general I wouldn't think you would care about local_time for > Easter? Perhaps a bit of code to show what you want to do? No local_time, just a boost::gregorian::date should be generated, like partial_date for example. So, here we go: What I want to achieve is something like the following: - ----------------------------------------------- // Use the Gauß Formula easter_sunday<date, GregorianEasterSundayGauss> es_gauss; // Use Oudin's Formula easter_sunday<date, GregorianEasterSundayOudin> es_oudin; partial_date pd( 16, Apr ); date d1 = es_gauss.get_date( 2006 ); // 2006-Apr-16 date d2 = es_oudin.get_date( 2006 ); // 2006-Apr-16 date d3 = pd.get_date( 2006 ); // 2006-Apr-16 assert( d1 == d2 ); assert( d1 == d3 ); // But, of course: date d4 = es_gauss.get_date( 2007 ); // 2007-Apr-08 date d5 = pd.get_date( 2007 ); // 2007-Apr-16 assert( d4 != d5 ); - ----------------------------------------------- with easter_sunday<> being defined as follows, modeled after other year based generators (from date_generators.hpp) with the policy defining the algorithm used to actually calculate the easter sunday date for a given year, which differs for the gregorian calendar (catholic, protestants) and the julian calendar (orthodox). The policy could be expected to provide an operator()( year ) which returns an easter_sunday::date_type, for example. - ----------------------------------------------- template <typename date_type, typename policy> easter_sunday : public year_based_generator<date_type> { public: typedef date_type::calendar_type calendar_type; typedef date_type::year_type year_type; easter_sunday() {} // does not require any arguments. virtual ~easter_sunday() {} virtual date_type get_date( year_type y ) const { return _generator_alg( y ); } virtual std::string to_string() const { return "easter_sunday"; /* or what? */ } private: policy _generator_alg; }; - ----------------------------------------------- One remaining glitch is the typedef, as exists for the other generators to pull in the name into the gregorian namespace (gregorian_types.hpp), which is not possible for the policy-based easter_sunday, as a partial template specialization typedef would be required... It would be nice to just say easter_sunday<GregorianEasterSundayGauss> for example, being able to leave out the boost::gregorian::date. A possible modification could be to change the (compile time) policy-based implementation to use a function pointer in the constructor to set the algorithm to be used (kind of Strategy pattern) at run time. Ah, yes, my original question actually was the to_string() which, according to the documentation of year_based_generator, is expected to return a string for use in a POSIX time_zone string. regards, -wolfi -- DI Wolfram Stering (Entwicklung) HALE electronic GmbH., Salzburg, Austria Tel: +43 (662) 439011 550 Fax: +43 (662) 439011 9 Email: wolfram.stering(a)hale.at

2 1

Re: [Boost-users] Threads (Not Mutex) Docs?
by Phil Endecott 04 Nov '05

04 Nov '05

> 1. Start at http://www.boost.org/doc/html > 2. Click on 12. Boost.Threads > 3. Click on Header <boost/thread/thread.hpp> in the contents > 4. Click on the "class thread;" link > > This seems pretty intuitive and perfectly acceptable to me. I replied to Merrill's question because I recognised the difficulty of finding this documentation as something that I had encountered, and continue to encounter from time to time. I suspect that those of you who have been using it for longer have got used to the way that it is organised and do consider it intuitive, but that it not how it has seemed to at least a couple of us "outsiders". Once located the Boost documentation is always good and very often excellent. This makes it a particular shame if it remains hidden. The particular problems that I've encountered are: - The "documentation" link on the front page at boost.org is in a smaller font even than the body text, and is not obviously a link. - The "automatic redirection failed" messages that appear all over the place are disconcerting. - The links to the actual class documentation, e.g. the "class thread;" link in (4) above, are not obviously links. - Links within the documentation often seem to be missing. For example, on the "Chapter 12 Boost.Threads" page, after the "Introduction" heading it looks as if a link is missing before the word "allows". Similarly after the heading "Runtime libraries" something seems to be missing between "using" and "must". Cheers, --Phil.

1 0