March 2006 - Boost-users - lists.preview.boost.org

How can I package boost library in a rpm?
by yinglcs2＠yahoo.com 22 Mar '06

22 Mar '06

Hi, If I need to build the boost library myself, how can I package it as a RPM? Thank you. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

Re: [Boost-users] Boost.Regex - why does it core dump?
by Smith Matt - matsmi 22 Mar '06

22 Mar '06

> Oh !*^% it's an outstanding shared_ptr bug: if you define > BOOST_SP_USE_PTHREADS both when building regex and again in your application > this problem will go away. I'll go off and nag the people responsible ;-> Well, surprisingly enough (because, as I understand it, gcc on HP-UX does not support multi-threading), that worked! Thanks for your help, John! Matt ************************************************************************* The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank you. *************************************************************************

1 0

Re: [Boost-users] [date_time] How to create a date from stringwithspecific format?
by Andrew Holden 22 Mar '06

22 Mar '06

> -----Original Message----- > From: Christian Henning [mailto:chhenning@gmail.com] > Sent: Wednesday, March 22, 2006 9:22 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] [date_time] How to create a date from > stringwithspecific format? > > Andrew, thanks for pointing out the submatch feature. Unfortunately no > match can be found when I try your code. Is it working on your > machine? Here, the dateMatch container size just 1. > > Any ideas why? > > Christian The code seems to work correctly on my machine (MSVC8 with Boost 1.33.1). It throws an out of range error because strDate specifies November 33. If I change strDate to 11-30-1997, then I can "cout << d" and get the date.

3 2

Re: [Boost-users] [date_time] How to create a date fromstringwithspecific format?
by Andrew Holden 22 Mar '06

22 Mar '06

> -----Original Message----- > From: Christian Henning > > int _tmain(int argc, _TCHAR* argv[]) > { > //second approach > boost::regex oDateReg2( "\\d{2}-\\d{2}-\\d{4}" ); The parentheses in my version of the example were important. They specify the size of the match object, as well as the exact portions of the string to capture. This line should read as follows: boost::regex oDateReg2( "(\\d{2})-(\\d{2})-(\\d{4})" ); or if you wish to allow single-digit months and days, boost::regex oDateReg2( "(\\d{1,2})-(\\d{1,2})-(\\d{4})" ); FYI, the order of the match object is the same as the CLOSING parentheses. This is good to know if you ever start nesting them. > > When I run the captures_examples of the regex lib the smatch stuff > works. But not here. Very strange. > > Thanks, > Christian > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

boost, COM and .NET [maybe OT]
by Thomas Laguzzi 22 Mar '06

22 Mar '06

Hello there, I'd like to thank Alan for the reply to my previous post. I'm sorry if this message is OT. What I ask is not a technical detail, but an opinion from you, c++ gurus :) Finally I've setup a simply class hierarchy that derives from a common object. Each class has: 1) Smart Pointer (boost::intrusiveptr),typedef intrusiveptr<ClassName> ClassName_Ptr; 2) Array of smart pointers ( typedef array <ClassName_Ptr> ClassName_Array) Everything works fine , and I'm happy :) the framework has no depends , it's self-contained and performant :) It is composed by a bunch of classes, but with MANY members. Now, I've got to make this small framework usable from: 1) .COM for compatibility 2) .NET for future applications and of course 3) c++ and boost :) All the three ways use the "allocate and forget about it" paradigm for memory allocation. I only use c++ types and I don't have ANY pointer used directly in the code. Changing to .NET is not hard and doesn't require wrapper classes (since the lib is self-contained), just some preprocessor programming and recompiling. Option 3) require no change :) What I worry about is .COM interop. Since classes are not derived from interfaces, I would require a lot of wrappers to native classes (because i don't want to use directly COM rules and types). So I see two solutions: 1) ok, write these wrappers as "external" code (I have lot's of properties and methods to wrap), maybe by using GCCXML and an automatic generator. 2) write base Interfaces and derive my framework classes from these, eventually adapting intrusive_ptr to use COM AddRef's 3) use .NET to COM interop and resolve problem 1 and 2 together, but distribuite 200MB of dll's Anyone has experience with intrusive_ptr (or shared_ptr) and COM pointers? Any idea about the general problem? Thanks & Sorry for my English! Thomas

2 1

Boost.Regex - why does it core dump?
by Smith Matt - matsmi 22 Mar '06

22 Mar '06

I have built Boost 1.33.1 on a HP-UX 11.22 OS with GCC 4.02. I have created a little test program, and when it gets to the following line of code it core dumps. boost::regex Exp( "[0-9]" ); This happens when I link both statically or dynamically. Any insights into what might be the problem would be greatly appreciated. Matt Smith Acxiom Corporation Data Compilation CWY0101-80 PO Box 2000 301 Industrial Blvd. Conway, AR 72033-2000 (501) 342-2434 Matt.Smith(a)acxiom.com *************************************************************************** The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank You. ****************************************************************************

3 7

Re: [Boost-users] Boost.Regex - why does it core dump?
by Smith Matt - matsmi 22 Mar '06

22 Mar '06

> 1) add a try catch block and verify that the dump isn't a result of an uncaught exception. Done. Same result. > 2) provide a backtrace? Okay; here it is: Program terminated with signal 11, Segmentation fault. #0 0x407c851 in boost::detail::atomic_increment (pw=0x400087a4) at sp_counted_base_gcc_ia64.hpp:38 38 "r"(pw)); (gdb) backtrace #0 0x407c851 in boost::detail::atomic_increment (pw=0x400087a4) at sp_counted_base_gcc_ia64.hpp:38 #1 0x407cc60 in boost::detail::sp_counted_base::add_ref_copy (this=0x400087a0) at sp_counted_base_gcc_ia64.hpp:117 #2 0x407cdb0 in boost::detail::shared_count::shared_count (this=0x7ffff0f8, r=@0x7ffff128) at shared_count.hpp:170 #3 0x4087700 in boost::shared_ptr<boost::re_detail::cpp_regex_traits_implementation<char > >::boost::shared_ptr<boost::re_detail::cpp_regex_traits_implementation<c har> > (this=0x7ffff0f4) at shared_ptr.hpp:106 #4 0x40af160 in std::pair<boost::shared_ptr<boost::re_detail::cpp_regex_traits_implement ation<char> >,boost::re_detail::cpp_regex_traits_base<char> const* >::std::pair<boost::shared_ptr<boost::re_detail::cpp_regex_traits_implem entation<char> >,boost::re_detail::cpp_regex_traits_base<char> const* > (this=0x7ffff0f4, __a=@0x7ffff124, __b=@0x7ffff0f0) at stl_pair.h:85 #5 0x4119f40 in boost::object_cache<boost::re_detail::cpp_regex_traits_base<char>,boost: :re_detail::cpp_regex_traits_implementation<char> >::do_get ( k=@0x7ffff1b0, max_cache_size=5) at object_cache.hpp:122 #6 0x411b190 in boost::object_cache<boost::re_detail::cpp_regex_traits_base<char>,boost: :re_detail::cpp_regex_traits_implementation<char> >::get ( k=@0x7ffff1b0, max_cache_size=5) at object_cache.hpp:78 #7 0x411b250 in boost::shared_ptr<boost::re_detail::cpp_regex_traits_implementation<char > > boost::re_detail::create_cpp_regex_traits<char> (l=@0x7ffff1e0) at cpp_regex_traits.hpp:822 #8 0x411b380 in boost::cpp_regex_traits<char>::boost::cpp_regex_traits<char> ( this=0x40008760) at cpp_regex_traits.hpp:842 #9 0x411b480 in boost::regex_traits<char,boost::cpp_regex_traits<char> >::boost::regex_traits<char,boost::cpp_regex_traits<char> > (this=0x40008760) at regex_traits.hpp:68 #10 0x411b500 in boost::regex_traits_wrapper<boost::regex_traits<char,boost::cpp_regex_tr aits<char> > >::boost::regex_traits_wrapper<boost::regex_traits<char,boost::cpp_regex _traits<char> > > (this=0x40008760) at regex_traits.hpp:162 #11 0x411b5b0 in boost::re_detail::regex_data<char,boost::regex_traits<char,boost::cpp_re gex_traits<char> > >::boost::re_detail::regex_data<char,boost::regex_traits<char,boost::cpp _regex_traits<char> > > (this=0x4000f410) at basic_regex.hpp:54 #12 0x411b740 in boost::re_detail::basic_regex_implementation<char,boost::regex_traits<ch ar,boost::cpp_regex_traits<char> > >::boost::re_detail::basic_regex_implementation<char,boost::regex_traits <char,boost::cpp_regex_traits<char> > > ( this=0x4000f410) at basic_regex.hpp:85 #13 0x411b880 in boost::basic_regex<char,boost::regex_traits<char,boost::cpp_regex_traits <char> > >::do_assign (this=0x7ffff2d8, p1=0x404d950 ".", p2=0x404d951 "", f=0) at basic_regex.hpp:517 #14 0x4079f10 in boost::basic_regex<char,boost::regex_traits<char,boost::cpp_regex_traits <char> > >::assign+0x90 () #15 0x407a000 in boost::basic_regex<char,boost::regex_traits<char,boost::cpp_regex_traits <char> > >::assign+0xc0 () #16 0x407a100 in boost::basic_regex<char,boost::regex_traits<char,boost::cpp_regex_traits <char> > >::boost::basic_regex<char,boost::regex_traits<char,boost::cpp_regex_tra its<char> > >+0xd0 () #17 0x4067ee0:0 in ValidateArgs+0x430 () #18 0x4068800:0 in main+0xc0 () Current language: auto; currently c++ ************************************************************************* The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system. Thank you. *************************************************************************

2 1

Re: [Boost-users] [regex] Working with wchar_t on olderUNIXplatforms
by Andrei Tarassov 22 Mar '06

22 Mar '06

As I understand using this feature requires ICU. Unfortunately, this is not an option for us :-( Andrei -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of John Maddock Sent: Tuesday, March 21, 2006 20:54 To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] [regex] Working with wchar_t on olderUNIXplatforms > I am intrigued with what you said about converting data from UTF-8 to > UTF-32 on the fly. It is absolutely not a problem to convert my > Unicode strings to UTF-8 encoded strings. Where could I read about > those on the fly conversions and what limitations do they have (e.g. > how locale settings are handled)? What locale settings? UTF-8 is mostly locale-independent (as an encoding), the only locale specific code is in the traits class to handle collation: and it only sees UTF-32 code points. The on-the-fly conversions are performed by iterator adapters in boost/regex/pending/unicode_iterator.hpp and the docs for the Unicode aware code is here: http://www.boost.org/libs/regex/doc/icu_strings.html John. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 1

file-system "system_complete()" and C:/
by Matt England 22 Mar '06

22 Mar '06

The following is a note of my developers that was attached to a bug close report. Is this special check for to convert "x:" to "x:/" a good practice for the Boost file-system library usage? Is there a better way to do this and/or are we missing something? -Matt The boost system_complete() function is used to convert relative paths into absolute paths. For some reason, boost interprets 'c:' as an incomplete path, and the windows boost behavior is to initially complete the path to the working directory. In other words 'c:' was becoming (in my case) 'c:/mydata/'. I added a check in Client.cpp to see if the --get parameter is equal to 'x:' and if it is, convert it to 'x:/'.

2 1

Re: [Boost-users] [date_time] How to create a date from string with specific format?
by Dean Sturtevant 21 Mar '06

21 Mar '06

Good question -- this is the kind of thing I've been struggling with lately (I tried to post a question about it to the group, but I worded it poorly and the moderator rejected it). Anyway, here's the solution I came up with. I am sure there are better ones! - Dean //! Utility function to split apart string. Returns # of components. inline int split_into_two_or_less(const std::string& s, char sep, std::string& first, std::string& second) { int sep_pos = static_cast<int>(s.find(sep)); if (sep_pos == -1) { first = s; return 1; } else { first = s.substr(0,sep_pos); second = s.substr(sep_pos+1); return 2; } } class expected_character : public boost::bad_lexical_cast { char _messagebuf[81]; public: expected_character(const char* who, char ch, const char* str, int pos) { snprintf(_messagebuf, sizeof(_messagebuf), "%s: Expected '%c' at position %d in \"%s\"", who, ch, pos, str); _messagebuf[sizeof(_messagebuf)-1] = '\0'; } const char* what() const throw() { return _messagebuf; } }; //! Parse date string of form YYYY-MM-DD // Code adapted from boost/date_time/date_parsing.hpp's parse_undelimited_date() template <class date_type> inline date_type parse_delimited_iso_8601_date(const std::string& s) { int offsets[] = {4,1,2,1,2}; int pos = 0; typedef typename date_type::year_type year_type; typename date_type::ymd_type ymd((year_type::min)(),1,1); // TODO: Is there a boost macro for sizeof(a)/sizeof(type)? or for the end of an array? boost::offset_separator osf(offsets, offsets+sizeof(offsets)/sizeof(offsets[0])); boost::tokenizer<boost::offset_separator> tok(s, osf); for(boost::tokenizer<boost::offset_separator>::iterator ti=tok.begin(); ti!=tok.end();++ti) { switch(pos) { case 0: ymd.year = boost::lexical_cast<unsigned short>(*ti); break; case 1: if (*ti != "-") { throw expected_character("parse_delimited_iso_8601_date", '-', s.c_str(), 5); } break; case 2: ymd.month = boost::lexical_cast<unsigned short>(*ti); break; case 3: if (*ti != "-") { throw expected_character("parse_delimited_iso_8601_date", '-', s.c_str(), 7); } break; case 4: ymd.day = boost::lexical_cast<unsigned short>(*ti); break; } pos++; } return date_type(ymd); }

2 1