June 2003 - Boost-users - lists.preview.boost.org

F6U2 anachronisms warnings
by ljohrixxx 16 Jun '03

16 Jun '03

Hello, I get the Anacronism warning from the file is_convertible.hpp line: 293 BOOST_TT_AUX_BOOL_TRAIT_DEF2(is_convertible,From,To, (::boost::detail::is_convertible_impl<From,To>::value)) The warning is: /usr/vobs/ThirdParty/boost/boost_1_30_0/boost/type_traits/is_convertib le.hpp", line 239: Warning (Anachronism): Type names qualified by template parameters require "typename". "/usr/vobs/ThirdParty/boost/boost_1_30_0/boost/type_traits/is_converti ble.hpp", line 239: Warning (Anachronism): Type names qualified by template parameters require "typename". I am using Forte6 U2 compiler on Solaris. Can someone please shed light on what does it mean? I get a whole bunch of similar warnings from many other boost headers. -LJ

2 1

Solution for (some) Compilation Problems with Sun Forte 6.2
by Adalbert Perbandt 16 Jun '03

16 Jun '03

Hi there, a colleague of mine tried to build the Boost::filesystem library (from Boost 1.30.0) under Sun SPARC Solaris7 with the Forte6.2 aka Sun Workshop5.3 compiler and got this failure: >> Assertion: (../links/init.cc, line 972) while processing /opt/boost_1.30.0/boost/type_traits/alignment_of.hpp at line 45. He investigated this problem and found out that defining the preprocessor macro BOOST_NO_INCLASS_MEMBER_INITIALIZATION would solve this. So we changed the boost/config/compiler/sunpro_cc.hpp at line 33 as follows: # if (__SUNPRO_CC <= 0x530) || !defined(BOOST_STRICT_CONFIG) # define BOOST_NO_INCLASS_MEMBER_INITIALIZATION Now we can compile the filesystem library successfully. Could please one of the experts comment on this change or maybe include this in the next release? Regards, Adalbert Perbandt.

3 3

Newbee: Build/Installation Questions
by Adalbert Perbandt 16 Jun '03

16 Jun '03

Hi there, I'm still uncertain how to properly build and install boost libraries under Solaris 7. I unpacked the boost_1_30_0.tar.gz file and ran from the boost_1_30_0/libs/test/build directory: $ jam -sTOOLS=sunpro -sBUILD="debug release" \ -sALL_LOCATE_TARGET=/opt/boost_1.30.0 This created some of the lib files below /opt/boost_1.30.0/libs/test/build/bin/... Not all of them because in release mode compiling the file boost_1_30_0/libs/test/src/unit_test_log.cpp failed several times: [...] sunpro-C++-action /opt/boost_1.30.0/libs/test/build/bin/libboost_test_exec_monitor.a/sunpro/release/runtime-link-dynamic/unit_test_log.o Assertion failure in prep_dot >> Assertion: (../links/prepexpression.cc, line 560) while processing ../src/unit_test_log.cpp at line 301. CC -c -DNDEBUG -fast -O4 -xtarget=generic -features=rtti -features=except -I"/opt/boost_1.30.0/libs/test/build" -I"/home/hwadm/Boost/boost_1_30_0" -o "/opt/boost_1.30.0/libs/test/build/bin/libboost_test_exec_monitor.a/sunpro/release/runtime-link-dynamic/unit_test_log.o" "../src/unit_test_log.cpp" ...failed sunpro-C++-action /opt/boost_1.30.0/libs/test/build/bin/libboost_test_exec_monitor.a/sunpro/release/runtime-link-dynamic/unit_test_log.o... [...] Line 301 and around in this file look so: //____________________________________________________________________________// unit_test_log& unit_test_log::operator<<( checkpoint const& cp ) { if( m_pimpl->m_entry_in_progress ) m_pimpl->set_checkpoint( cp ); // LINE 301 return *this; } //____________________________________________________________________________// I cannot see any complexity here that could lead to an assertion. As I learned from the Boost Users Mailinglist archive such assertions seem to be common problems with Sun's forte 6.2 and 7 compilers. O.k. When the build above was finished I got at least the libs. built in debug mode obvisously completely installed below /opt/boost_1.30.0/libs/test/build/bin/<libs>/sunpro/... To get the header files installed I ran $ bjam -sTOOLS=sunpro -sBUILD=install But this command did not install the header files which would be required to use the test library. Instead a complex directory structure was created below ../sunpro/install. Is there no way to automatically get all the header files that belong to the test lib plus the other boost header files used by this lib installed automatically? Many thanks in advance for any hint. Regards, Adalbert Perbandt.

2 5

Forward lookahead - abysmal failure
by noamxnoam 15 Jun '03

15 Jun '03

Hello all, Tried to use the forward lookahead asserts, both positive and negative and could not make a match. Is there some sample code anywhere or a tip to explain where I went wrong? Thanks, Noam

2 3

RE: [Boost-Users] Agree totally.
by Jim.Hyslop 15 Jun '03

15 Jun '03

Dan Dimerman [mailto:yg-boost-users@gmane.org] wrote: > Can anyone who successfuly wrote some code with BGL post it > somewhere ? I presume you're asking because you are having problems writing code with the BGL. Rather than ask for someone to post code, why not tell us the problem you are having? Oh, by the way - the subject line really should have something to do with the body of the message. At least, that's the traditional way of doing things and it works quite well ;=) -- Jim

2 1

RE: [Boost-Users] Regex in a service..
by Koss, Loren 15 Jun '03

15 Jun '03

John, Thanks for the response. How do you deal with breakpoints within a service? I'm compiling then starting the service from the service control manager. Will I still be able to debug it? Loren -----Original Message----- From: John Maddock [mailto:john_maddock@compuserve.com] Sent: Friday, June 13, 2003 3:22 AM To: Boost-Users(a)yahoogroups.com Subject: Re: [Boost-Users] Regex in a service.. > I seem to be having a difficult time using the Regex class in a > service. It works wonderfully in a regular app. Can anyone help me > out.. Basically, instantiating a regex class raises an exception. > Any thoughts? Not unless you can tell me what exception and where - if you're using the vc ide then set it to break on OS and C++ exceptions. John. Yahoo! Groups Sponsor <http://us.adserver.yahoo.com/l?M=247865.3425083.4707139.1261774/D=egroupmai l/S=:HM/A=1482387/rand=220444082> Info: <http://www.boost.org <http://www.boost.org> > Wiki: <http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl <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 the Yahoo! <http://docs.yahoo.com/info/terms/> Terms of Service. [Non-text portions of this message have been removed]

3 2

"Unresolved externals" building MSVC 6.0 boost_1_30_0
by Peter Hug 13 Jun '03

13 Jun '03

When trying to build boost_1_30_0 as follows: vcvars32.bat bjam "-sTOOLS=msvc" I get some link errors: cpp_main.obj : error LNK2001: unresolved external symbol "int __cdecl cpp_main(i nt,char * * const)" (?cpp_main@@YAHHQAPAD@Z) test_main.obj : error LNK2001: unresolved external symbol "int __cdecl test_main (int,char * * const)" (?test_main@@YAHHQAPAD@Z) unit_test_main.obj : error LNK2001: unresolved external symbol "class boost::uni t_test_framework::test_suite * __cdecl init_unit_test_suite(int,char * * const)" (?init_unit_test_suite@@YAPAVtest_suite@unit_test_framework@boost@@HQAPAD@Z) And the tool finishes telling me these not so great news: ...failed updating 12 targets... Well I'm new but after spending two days trying to work out what the heck goes on, I'm none the wiser!

2 3

Re: RegEx Windows vs Linux
by Mark Sizer 13 Jun '03

13 Jun '03

On a whim, I tried changing 'boost::cmatch' to 'boost::smatch'. That fixed my problem. I'd appreciate it if someone would tell me what was going on, though. Thanks, - Mark

2 1

Regex in a service..
by lorenkoss12 13 Jun '03

13 Jun '03

I seem to be having a difficult time using the Regex class in a service. It works wonderfully in a regular app. Can anyone help me out.. Basically, instantiating a regex class raises an exception. Any thoughts? Thanks Loren

2 1

Spirit for C++ style if expressions
by Vincent Finn 12 Jun '03

12 Jun '03

Hi, I am thinking of using spirit for C++ style expressions and am wondering if it is the way to go What I have will be something like ( (A != 192) AND ( (V > 0) AND (V < "89") ) ) What I need is to be able to parse this replace the AND, > etc... with operations, replace string numbers with real numbers and then apply the operation to a large number of A and V. I would have thought this would be quite a common usage but I haven't seem any samples doing it of partictuar importance is the delayed evaluation Anyone have an example of this? Vin

2 1

MSVC 6.0 SP5: INTERNAL COMPILER ERROR
by Peter 12 Jun '03

12 Jun '03

Arghhhh.... c:\Development\Libraries\boost\boost_1_30_0\boost\iterator_adaptors.hpp(872) : fatal error C1001: INTERNAL COMPILER ERROR (compiler file 'msc1.cpp', line 1794) Please choose the Technical Support command on the Visual C++ Help menu, or open the Technical Support help file for more information c:\Development\Libraries\boost\boost_1_30_0\boost\date_time\date_parsing.hpp (27) : see reference to class template instantiation 'boost::iterator_adaptor<class boost::detail::token_iterator_base<char const *>,class boost::detail::tokenizer_policy<class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> >,class boost::char_delimiters_separator<char,struct std::char_traits<char> > >,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> >,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > const &,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > const *,struct std::forward_iterator_tag,int>' being compiled

3 2

boost::condition and lock concept
by db10www 12 Jun '03

12 Jun '03

I have created a mutex class which is a model of the mutex concept. I also have lock classes for this mutex which models the lock concept (e.g. called CMyLock). (The reason for this is to implement a read/write mutex and associated read and write locks) According to the boost documentation I should be able to use this lock with the boost::condition but I get a compile time error:- 'm_mutex' : is not a member of 'CMyLock' This is due to the condition::wait method defined in condition.hpp making a reference to lock.m_mutex Surely boost::condition only needs a lock object to be an instance of a class which is a model of lock. The lock concept makes no reference to a member variable called m_mutex. Is this a bug of boost::condition or am I doing something wrong. Any ideas? Many thanks.

2 1

Struggling with bind
by Paul Thompson 12 Jun '03

12 Jun '03

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.

2 3

Re: [crc] Simple example please
by Daryle Walker 12 Jun '03

12 Jun '03

On Tuesday, June 10, 2003, at 9:08 AM, Vladimir Prus wrote: [SNIP] > 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? I'll consider it. [A few hours pass...] How about something like this: //================================================================ #include <cstdlib> #include <exception> #include <fstream> #include <ios> #include <iostream> #include <ostream> #include <boost/crc.hpp> #ifndef BLOCK_SIZE #define BLOCK_SIZE 1024 #endif int main ( int argc, char const * argv[] ) try { boost::crc_32_type result; // Loop over each file argument for ( int i = 1 ; i < argc ; ++i ) { std::ifstream ifs( argv[i], std::ios_base::binary ); if ( ifs/*std::ifstream ifs(argv[i], std::ios_base::binary)*/ ) { /*char buffer[ BLOCK_SIZE ]; std::streamsize len; while ( 0 < (len = ifs.readsome( buffer, BLOCK_SIZE )) ) { result.process_bytes( buffer, len ); }*/ char c; while ( ifs.get(c) ) { result.process_byte( c ); } } else { std::cerr << "Failed to open file '" << argv[i] << "'." << std::endl; } } std::cout << std::hex << result.checksum() << std::endl; return EXIT_SUCCESS; } catch ( std::exception &e ) { std::cerr << "Found an exception with '" << e.what() << "'." << std::endl; return EXIT_FAILURE; } catch ( ... ) { std::cerr << "Found an unknown exception." << std::endl; return EXIT_FAILURE; } //================================================================ Note in the "if" statement, I had the construction in it. However, neither one of my compilers took it! Moving the constructor call worked, but I thought my original method should be accepted. One compiler finished after the above 'fix'. The other (a variant of GCC 3.1) complained about missing symbols at link-time; it could not create the global objects hidden in Boost.CRC. (I haven't worked on the code in a long time; there could be quirks on compilers written after Boost.CRC.) When I ran the program under the compiler that worked, the "readsome" function always returned zero! There were no suitable substitutes; the "read" function doesn't tell you how many characters it read (bad if the last block isn't full) and the multi-character "get" functions use delimiters. So this program is stuck at single-character reads at the moment. Maybe I should switch to the stream-buffer method you used. Can you try this out to make sure the problems aren't just mine? Daryle

2 2

RegEx Windows vs Linux
by Mark Sizer 12 Jun '03

12 Jun '03

I compiling the same code on both platforms (same my code, same boost code). Under Windows (MSVC++ 6 w/SP something), this compiles: // these are constructed w/values, it's just ugly here std::string sRegEx(); std::string sMe(); boost::regex regexpPattern( sRegEx ); boost::cmatch mrFound; bool bMatch = boost::regex_search(sMe, mrFound, regexpPattern); Under Linux (gcc 3.3) I get an error message (I added the whitespace): error: could not convert `mrFound' to `boost::match_results< std::__normal_iterator< const char*, std::basic_string< char, std::char_traits<char>, std::allocator<char> > >, std::allocator<char> >&' Digging around a little bit, I found this: template <class iterator, class Allocator> class match_results : yadda yadda and typedef match_results<const char*> cmatch; This certainly SEEMS much simpler than what the error message claims to be looking for. Can anyone help, please? Thanks, - Mark

1 0

Re: [crc] Simple example please
by Daryle Walker 12 Jun '03

12 Jun '03

On Monday, June 9, 2003, at 5:48 PM, Alex Henderson wrote: > 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. > > I've looked at the crc library, but can't get my head round the > example. It seems to test the library rather than show how its used. Yes, the sample program is meant as a test of the library. You can still get an idea of usage from it. There are also short examples in the HTML documentation. > I'm after something like > > std::vector<char> MyStream1; // the whole file (about 500kb) > > boost::SomeCRCTypeThatIsOptimisedForSafetyAndLargeishBlocks<???> x; > > long = x.checkme(MyStream1); > > Could someone please point me to some code where a simple example is > detailed? What's in the documentation is it. Do you have a specific CRC algorithm in mind? If so, you just pick the appropriate template parameters. If not, you can choose one of the samples I provided, like boost::crc_32_type. The optimized CRC function objects can work with a single byte at a time or an array-segment of bytes. By putting your file's data in a std::vector as you have in your example, you limit yourself to just the single-byte computing call, as the elements of a vector may not be contiguous. Just allocate a simple char array and use one of the CRC object's array computing calls or use one of the CRC functions. Based on what you said, I don't think you need the object-wrapping that std::vector gives. Daryle

2 4

Boost::Test v. CppUnit
by Jon Stewart 11 Jun '03

11 Jun '03

Howdy - I've been using CppUnit and various parts of Boost for a little over a year. The 1.3.0 rlease includes support for unit testing, a la CppUnit, JUnit, etc. At first glance, I can't find any fundamental differences between CppUnit and Boost's unit testing framework. I'm sure that CppUnit could benefit from a code review or two by Boost's luminaries, but I'm still surprised that a new framework was created instead of trying to improve CppUnit. Can anyone state a few areas where CppUnit and Boost::test differ? In general, I tend to prefer Boost's solutions over others', but I don't see a clear cut case to be made for Boost over CppUnit, vis a vis unit testing. I'd be happy to change my mind, however. :-) Jon -- Jon Stewart stew1(a)ellipsis.cx

7 8

RE: [Boost-Users] first time boost
by Jim.Hyslop 11 Jun '03

11 Jun '03

G Keshav wrote: > I want to use boost in my MS VIsual Studio project. > Can you please advise stepwise about the procedure. 1) Decide which part of Boost you want to use. 2) download and extract the source from boost.org (follow the instructions there) 3) include the proper headers 4) write your code -- Jim

1 0

first time boost
by G Keshav 11 Jun '03

11 Jun '03

Hi there, I want to use boost in my MS VIsual Studio project. Can you please advise stepwise about the procedure. Thanks and regards, Gagnesh --------------------------------- Do you Yahoo!? Free online calendar with sync to Outlook(TM). [Non-text portions of this message have been removed]

1 0

using shared pointer in a linked data structure
by Zumichu 11 Jun '03

11 Jun '03

Basically I'm having trouble understanding exactly how to use shared_ptr . If I have a class: GraphNode { public: AddChild(shared_ptr<GraphNode> n) { m_Children.push_back(n); shared_ptr<GraphNode> thisptr(this); n->m_Parents.push_back(thisptr); } DeleteChild(shared_ptr<GraphNode> n); private: vector<shared_ptr<GraphNode> > m_Children; vector<shared_ptr<GraphNode> > m_Parents; }; main() { shared_ptr<GraphNode> n1(new GraphNode()); shared_ptr<GraphNode> n2(new GraphNode()); n1->AddChild(n2); } doing it this way gives me a run-time error that seems to have to do with updating the child node's parent vector with the parent node's this pointer. Am I not using the shared_ptr right?

2 1

Re: [crc] Simple example please
by Vladimir Prus 11 Jun '03

11 Jun '03

Daryle, Daryle Walker wrote: > One compiler finished after the above 'fix'. The other (a variant of > GCC 3.1) complained about missing symbols at link-time; it could not > create the global objects hidden in Boost.CRC. (I haven't worked on > the code in a long time; there could be quirks on compilers written > after Boost.CRC.) > > When I ran the program under the compiler that worked, the "readsome" > function always returned zero! There were no suitable substitutes; the > "read" function doesn't tell you how many characters it read (bad if > the last block isn't full) and the multi-character "get" functions use > delimiters. So this program is stuck at single-character reads at the > moment. Maybe I should switch to the stream-buffer method you used. > > Can you try this out to make sure the problems aren't just mine? With gcc 3.3 on Linux, 'readsome' variant, 'get' variant and 'sbumpc' variant all produce the same output. However, I'm not sure that 'readsome' variant should work. Here's what "man std::iostream" tells about 'readsome'. if A > 0, extracts min(A,n) [ A = rdbuf()->in_avail() ] ... The goal is to empty the current buffer, and to not request any more from the external input sequence controlled by the streambuf. Concerning 'in_avail()': If a read position is available, returns the number of characters available for reading before the buffer must be refilled. Otherwise returns the derived showmanyc(). Lastly, about 'showmanyc()': Base class version does nothing, returns zero. So, this all leads to conclusion that 'readsome' is required to read all characted currently available in stream buffer, but may ignore anything else. I believe that the right way is to use 'read' function, together with 'gcount()' with tells how many characters were read by the last unformatted input operation. The main loop will become: do { ifs.read(buffer, BLOCK_SIZE); result.process_bytes(buffer, ifs.gcount()); } while(ifs); Of course, this variant produces the same results for me. Can you try with your compiler? - Volodya P.S. In fact, I still don't understand what 'in_avail()' is good for...

1 0

Re: [crc] Simple example please
by Vladimir Prus 11 Jun '03

11 Jun '03

Daryle, Daryle Walker wrote: > One compiler finished after the above 'fix'. The other (a variant of > GCC 3.1) complained about missing symbols at link-time; it could not > create the global objects hidden in Boost.CRC. (I haven't worked on > the code in a long time; there could be quirks on compilers written > after Boost.CRC.) > > When I ran the program under the compiler that worked, the "readsome" > function always returned zero! There were no suitable substitutes; the > "read" function doesn't tell you how many characters it read (bad if > the last block isn't full) and the multi-character "get" functions use > delimiters. So this program is stuck at single-character reads at the > moment. Maybe I should switch to the stream-buffer method you used. > > Can you try this out to make sure the problems aren't just mine? With gcc 3.3 on Linux, 'readsome' variant, 'get' variant and 'sbumpc' variant all produce the same output. However, I'm not sure that 'readsome' variant should work. Here's what "man std::iostream" tells about 'readsome'. if A > 0, extracts min(A,n) [ A = rdbuf()->in_avail() ] ... The goal is to empty the current buffer, and to not request any more from the external input sequence controlled by the streambuf. Concerning 'in_avail()': If a read position is available, returns the number of characters available for reading before the buffer must be refilled. Otherwise returns the derived showmanyc(). Lastly, about 'showmanyc()': Base class version does nothing, returns zero. So, this all leads to conclusion that 'readsome' is required to read all characted currently available in stream buffer, but may ignore anything else. I believe that the right way is to use 'read' function, together with 'gcount()' with tells how many characters were read by the last unformatted input operation. The main loop will become: do { ifs.read(buffer, BLOCK_SIZE); result.process_bytes(buffer, ifs.gcount()); } while(ifs); Of course, this variant produces the same results for me. Can you try with your compiler? - Volodya P.S. In fact, I still don't understand what 'in_avail()' is good for...

1 0

RE: [Boost-Users] to_simple_string and Windows?
by Tom Matelich 11 Jun '03

11 Jun '03

UNICODE is enabled by the _UNICODE define. Assuming you're still using MSVC 6 (remember to rebuild all often since your dependencies may not getting generated correctly), just go to your Project/Settings/C/C++ tab and in Preprocessor definitions, remove the UNICODE symbol from the list. Otherwise, remove it from your Makefile. > -----Original Message----- > From: adutoit [mailto:adutoit@stonethree.com] > Sent: Tuesday, June 10, 2003 7:01 AM > To: Boost-Users(a)yahoogroups.com > Subject: Re: [Boost-Users] to_simple_string and Windows? > > > It is not a compile-time error, but rather a run-time error. > Ie. if I compile any program which utilises > boost::to_simple_string with any > optimisation whatsoever; it simply aborts at the point where I do my > function call to "to_simple_string". That is irrespective of > my using a > type date or ptime as the function argument... > > I must stress again that "to_simple_string" works fine when I > compile it in > Linux. > > O, another question I have. Where do I disable Unicode > support (wchar_t > usage in fact!) alltogether for Boost? > > Previously I had great fun with lexical_cast.hpp which > utilised wchar_t > versions of some of the streaming operators because I had a > silly UNICODE > define as part of my project options. The solution was to > use /Zc:wchar_t > project option in M$VC6. > > But UNICODE defined caused problems in network browsing code > I used which > `normally' returned a LPSTR instead of the LPWSTR if UNICODE > is defined. > Long story. I did a temp hack in lexical_cast.hpp by > # define DISABLE_WIDE_CHAR_SUPPORT > #undef BOOST_NO_INTRINSIC_WCHAR_T > and getting rid of the redefinition of BOOST_NO_INTRINSIC_WCHAR_T. > > Now is there a way to do this cleanly, compile-time so that I > do not need to > keep on hacking the source code? > > Greetings, > Andre. > > > > ----- Original Message ----- > From: "Jeff Garland" <jeff(a)crystalclearsoftware.com> > To: <Boost-Users(a)yahoogroups.com> > Sent: Tuesday, June 10, 2003 2:51 PM > Subject: RE: [Boost-Users] to_simple_string and Windows? > > > > > Darn; here I go replying to my own post again... > > > Anyway; so far I have narrowed the problem down to one of > optimisation. > I > > > did a debug compilation and had no problem. A moment of > reflection led > me > > > to the conclusion that optimisation might have something > to do with it. > For > > > the record; I am using Boost in Windows, with Visual > Studio as IDE only, > and > > > the Intel 7.0 compiler. Oh; and STLport instead of > Windows' crappy STL > > > implementation. > > > > > > I.e. turning on "Maximise speed" optimisation causes > "to_simple_string" > to > > > fail on my build system. > > > > > > I'll try to setup the simplest proggie I can to > illustrate this problem. > > > > > > Any thoughts? > > > > I'm not aware of anyone else having a problem. Perhaps you can post > > the details of the compilation failure? > > > > Jeff > > > > > > 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/ > > > > > ------------------------ Yahoo! Groups Sponsor > ---------------------~--> > Get A Free Psychic Reading! Your Online Answer To Life's > Important Questions. > http://us.click.yahoo.com/Lj3uPC/Me7FAA/ySSFAA/EbFolB/TM > -------------------------------------------------------------- > -------~-> > > 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/ > > ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. -----------------------------------------------------------------------

2 1

newbie boost thread build under Linux help
by Mark Sizer 11 Jun '03

11 Jun '03

I had such success with Boost under Windows, that I'm now trying to get it up and running under Linux. This is not a happy experience. I got it to work, but this can't be the way it's supposed to be done: I'm using boost_1_30_0. I installed the bjam RPM (boost_jam_3.1.4-1.i386.rpm). bjam "-sTOOLS=gcc" from a bash command line. I had all sorts of very strange error messages, so I set about upgrading my Linux environment. I boostrapped the gcc 3.3 compiler (I had 2.6 something). Same error messages. I upgraded libstdc++ to libstdc++-v3.0.97-gcc-3.1.x.tar.gz (I'm seeing the gcc version mismatch as I type this, but it is the latest libstdc++). The "merging into an existing compiler" instructions seemed really complicated so I installed it and re-bootstrapped the compiler. Got a more sensible error message: Unknown compiler version please run the configure tests and report the results. (from gnu.hpp) I hacked gnu.hpp to allow the 3.3 compiler. The error message changed to being unable to find the std:: headers (first one was <utility> from select_stdlib.hpp). I seem to have at least four versions: /usr/include/g++-2/utility /usr/include/g++-3/utility /usr/include/g++-v3/utility (judging by the dates, this seems the correct one) /usr/local/include/c++/utility Obviously, none of these are in the compiler's include path. bjam "-sTOOLS=gcc" "-sGCC_INCLUDE_DIRECTORY=/usr/include/g++-v3" Wasn't helpful: "bits/c++config.h" can't be found. That's in /usr/include/g++-v3/i686-pc-linux-gnu. However, there doesn't seem to be a way to specify more than one include directory (neither multiple "-s" nor delimited paths in a single "-s" worked). So I copied everything from i686-pc-linux-gnu/bits to ../../bits. It built. This can't be normal. How is this process supposed to work? Thanks, - Mark

2 1

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