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

bjam hanging on MacOS X
by Nick Bastin 31 May '04

31 May '04

I downloaded Boost 1.31.0 and built bjam with the build.sh script. Now, when I try to build boost with bjam -sTOOLS=darwin install, it just hangs, with no output to the console at all. Is there a way to set it into a verbose mode to see what's going on, or does anybody have any idea what's wrong? This is MacOS X 10.3.3 (7G33) on a 12" powerbook G4 (1.33Ghz) with the latest developer tools (7K249). -- Nick

2 4

Mingw build Boost.Python
by nagual 31 May '04

31 May '04

Hello, All! I try build Boost.Puthon use MinGW g++ (GCC) 3.3.1 (mingw special 20030804-1) command line: cd C:\Lang\Projects\boost\libs\python\build\ bjam -sTOOLS=mingw -s"PYTHON_ROOT=C:\Lang\Python" -s"PYTHON_VERSION=2.3" output: ...found 1165 targets... ...updating 19 targets... gcc-C++-action ..\..\..\bin\boost\libs\python\build\boost_python.dll\mingw\debug\slice.obj In file included from C:/Lang/Projects/boost/libs/python/src/slice.cpp:1: C:/Lang/Projects/boost/boost/python/slice.hpp:29: warning: `dllexport' attribute ignored C:/Lang/Projects/boost/boost/python/slice.hpp:29: error: declaration does not declare anything C:/Lang/Projects/boost/boost/python/slice.hpp:29: error: syntax error before `) ' token C:/Lang/Projects/boost/libs/python/src/slice.cpp:18: error: prototype for ` boost::python::slice::slice()' does not match any in class ` boost::python::slice' C:/Lang/Projects/boost/boost/python/slice.hpp:20: error: candidates are: boost::python::slice::slice(const boost::python::slice&) C:/Lang/Projects/boost/boost/python/slice.hpp:240: error: boost::python::slice::slice(boost::python::detail::new_non_null_reference_t* ) C:/Lang/Projects/boost/boost/python/slice.hpp:240: error: boost::python::slice::slice(boost::python::detail::new_reference_t*) C:/Lang/Projects/boost/boost/python/slice.hpp:240: error: boost::python::slice::slice(boost::python::detail::borrowed_reference_t*) C:/Lang/Projects/boost/boost/python/slice.hpp:39: error: template<class Integer1, class Integer2, class Integer3> boost::python::slice::slice(Integer1, Integer2, Integer3) C:/Lang/Projects/boost/boost/python/slice.hpp:34: error: template<class Integer1, class Integer2> boost::python::slice::slice(Integer1, Integer2) "g++" -c -Wall -ftemplate-depth-100 -DBOOST_PYTHON_DYNAMIC_LIB -DBOOST_PYTHON_SOURCE -g -O0 -fno-inline -mno-cygwin -I"..\..\..\bin\boost\libs\python\build" -I "C:\Lang\Projects\boost" -I "C:\Lang\ActiveState\Python\include" -o"..\..\..\bin\boost\libs\python\build\boost_python.dll\mingw\debug\slice.o bj" "C:\Lang\Projects\boost\libs\python\build\../src/slice.cpp" [i skip other] After change slice.hpp line19-30: class BOOST_PYTHON_DECL slice : public object { private: // Helper function to work around bugs in MSVC 6 //BOOST_PYTHON_DECL static object new_slice(PyObject*, PyObject*, PyObject*); public: // Equivalent to slice(::) //BOOST_PYTHON_DECL slice(); slice.cpp compile ok. But I recive bjam output: ...found 1165 targets... ...updating 19 targets... gcc-C++-action ..\..\..\bin\boost\libs\python\build\boost_python.dll\mingw\debug\slice.obj gcc-Link-action ..\..\..\bin\boost\libs\python\build\boost_python.dll\mingw\debug\boost_pyth on_debug.dll ..\..\..\bin\boost\libs\python\build\boost_python.dll\mingw\debug\boost_pyth on_debug.lib spawn: Invalid argument I build static version using command: cd C:\Lang\Projects\boost\libs\python\build\ bjam -sTOOLS=mingw -s"PYTHON_ROOT=C:\Lang\ActiveState\Python" -s"PYTHON_VERS ION=2.3" -s"BUILD=debug release" lib Help my build Boost.Python library dynamic version using MinGW With best regards, nagual. E-mail: nagual(a)promsoft.ru

2 4

Regex++ Installation Problems
by kinshuk srivastava 30 May '04

30 May '04

Hi I had written earlier too regarding installation problems with the regex++...sorry for not providing the complete details last time..But I am again sending the errors... The Gcc compiler which m using is gcc3.2 nd the operating system is SuSE 9.6...I think the regex++ version is 331... the following are the errors: make -fgcc.mak mkdir -p gcc mkdir -p gcc/boost_regex g++ -o gcc/boost_regex/c_regex_traits.o -c -O2 -I../../../ ../src/c_regex_traits.cpp g++ -o gcc/boost_regex/c_regex_traits_common.o -c -O2 -I../../../ ../src/c_regex_traits_common.cpp g++ -o gcc/boost_regex/cpp_regex_traits.o -c -O2 -I../../../ ../src/cpp_regex_traits.cpp g++ -o gcc/boost_regex/cregex.o -c -O2 -I../../../ ../src/cregex.cpp ../../../boost/regex/v3/regex_match.hpp: In member function `unsigned int boost::RegEx::GrepFiles(bool (*)(const char*, const boost::RegEx&), const char*, bool, unsigned int)': ../../../boost/regex/v3/regex_match.hpp:1937: Internal compiler error in expand_call, at calls.c:3049 Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions. make: *** [gcc/boost_regex/cregex.o] Fehler 1 Thanx in advance... -Kinshuk ________________________________________________________________________ Yahoo! India Matrimony: Find your partner online. http://yahoo.shaadi.com/india-matrimony/

2 1

bjam build failed (winxp, mingw)
by krejci.lukas＠volny.cz 29 May '04

29 May '04

Hi, today I downloaded boost 1.31.0 and precompiled bjam. then I invoked bjam: cd C:\_projects\_coding\_sdk\boost_1_31_0 bjam -s"TOOLS=mingw" -s"MINGW_ROOT_DIRECTORY=C:\Program Files\Dev-Cpp" --prefix=C:\_projects\_coding\@mingw --with-python-root="C:\Program Files\Python23" --with-pydebug install and this was the output: ...patience... ...found 9330 targets... ...updating 3796 targets... MkDir1 bin spawn: No such file or directory I am new to jam/bjam. Can anyone help me with this problem? Thanks for your help, L.

2 1

Re: Using filenames starting with a dot
by Steven Woolgar 28 May '04

28 May '04

> I wanted to create a hidden files (for configuration etc) by starting > the filename with a dot on a Linux system. > The following line aborts the program when filename is a std::string > starting with a dot: > > > boost::filesystem::path filename_complete = basepath / filename; > > Are there any tricks for that? I think in fact an exception is being thrown. Wrap it in a try block to find out what is getting thrown. Once you know, you will find that the code has traits to control the behavior in regards to non-standard naming conventions. Woolie

1 0

Calling thread::join from another thread?
by Wolfgang Meyer 28 May '04

28 May '04

Given a thread A created by a thread B. Is calling the join-method of A from a thread C a well-defined operation? (With C != B && C != A) It seems to work fine, but I want to make sure that this was not just 'luck'. Assumed, the operation is not defined: Can someone imagine a way to wrap a thread such that is can be joined by an arbitrary other thread? Thanks for clarification, Wolfgang Meyer

2 1

Using filenames starting with a dot
by Uli Türk 28 May '04

28 May '04

Hi everyone! I wanted to create a hidden files (for configuration etc) by starting the filename with a dot on a Linux system. The following line aborts the program when filename is a std::string starting with a dot: > boost::filesystem::path filename_complete = basepath / filename; Are there any tricks for that? Thanks! Uli

3 2

Re: [Boost-users] boost::thread, boost::bind, and member functions
by Mike Feldmeier 27 May '04

27 May '04

Hah! I found the answer in a python mailing list from someone who was having trouble building the python part of boost in conjunction with stlport. Changing my build batch to the following was the key: set stlport-iostream="on" set STLPORT_VERSION="4.6.2" set STLPORT_PATH="C:/" bjam "-sTOOLS=vc7.1-stlport" "-sBUILD=debug release" install Thanks to everyone for their help! Mike _________________________________________________________________ MSN Toolbar provides one-click access to Hotmail from any Web page FREE download! http://toolbar.msn.click-url.com/go/onm00200413ave/direct/01/

1 0

RE: [Boost-users] lexical_cast tests failed
by Rozental, Gennadiy 27 May '04

27 May '04

> I was testing out lexical_cast, and I would expect that for > any value v of type T, it should be true that > > v == lexical_cast<T>(lexical_cast<string>(v)) > > but this does not appear to work for float or double in all > cases. In particular, the following tests failed: > > float value = numeric_values<float>::min(); > BOOST_CHECK_EQUAL(value, > lexical_cast<float>(lexical_cast<string>(value))); > value = numeric_values<float>::max(); > BOOST_CHECK_EQUAL(value, > lexical_cast<float>(lexical_cast<string>(value))); > > double dvalue = numeric_values<double>::min(); > BOOST_CHECK_EQUAL(dvalue, > lexical_cast<double>(lexical_cast<string>(dvalue))); > dvalue = numeric_values<double>::max(); > BOOST_CHECK_EQUAL(dvalue, > lexical_cast<double>(lexical_cast<string>(dvalue))); > > Should I have expected these tests to pass? > > thanks, > > --chris Try to use BOOST_CHECK_CLOSE instead. Don't use BOOST_CHECK_EQUAL to compare floats. Gennadiy.

1 0

lexical_cast tests failed
by Chris Surridge 27 May '04

27 May '04

I was testing out lexical_cast, and I would expect that for any value v of type T, it should be true that v == lexical_cast<T>(lexical_cast<string>(v)) but this does not appear to work for float or double in all cases. In particular, the following tests failed: float value = numeric_values<float>::min(); BOOST_CHECK_EQUAL(value, lexical_cast<float>(lexical_cast<string>(value))); value = numeric_values<float>::max(); BOOST_CHECK_EQUAL(value, lexical_cast<float>(lexical_cast<string>(value))); double dvalue = numeric_values<double>::min(); BOOST_CHECK_EQUAL(dvalue, lexical_cast<double>(lexical_cast<string>(dvalue))); dvalue = numeric_values<double>::max(); BOOST_CHECK_EQUAL(dvalue, lexical_cast<double>(lexical_cast<string>(dvalue))); Should I have expected these tests to pass? thanks, --chris

1 0

[regex] invalid syntax for non-gready repeats
by Thorsten Ottosen 27 May '04

27 May '04

Hi, this inocent code throws an exception: static const regex rpath( "Return-Path:(.*?)Received: ", regex_constants::icase ); according to http://www.boost.org/libs/regex/doc/syntax.html this is how non-gready repeats are specified. Can anybody figure out what is wrong? Thanks Thorsten

2 2

Use of a member pointer to a member function in for_each - is it possible?
by Vesselin Kostadinov 27 May '04

27 May '04

Hi, I am struggling to find the proper combination of ataptors/binders to use a member pointer to a member function in a for_each. Is this possible at all? Any help is appreciated. Here is the code: #include <iostream> #include <algorithm> #include <vector> #include <boost/bind.hpp> #include <boost/ref.hpp> using namespace std; class foo { private: int value; void default_proc(int& x) { cout << "default_proc. Value = " << value << ", Param = " << x << "\n";} void init1_proc(int& x) { cout << "init1_proc. Value = " << value << ", Param = " << x << "\n";} public: void (foo::*func_ptr)(int& x); //yes, I need the reference here foo(int init) : value(init), func_ptr(init == 1 ? &foo::init1_proc : &foo::default_proc) {} }; int main() { vector<foo> foos; foos.push_back(foo(0)); foos.push_back(foo(1)); foos.push_back(foo(2)); int param = 3; //i need this functionality: //prints "default_proc. Value = 0, Param = 3": (foos[0].*foos[0].func_ptr)(param); //prints "init1_proc. Value = 1, Param = 3": (foos[1].*foos[1].func_ptr)(param); //prints "default_proc. Value = 2, Param = 3": (foos[2].*foos[2].func_ptr)(param); //using the right combination of for_each, bind, ref, mem_fn, etc: // std::for_each(foos.begin(), foos.end(), // boost::bind(???????), // _1, boost::ref(param)); } The compiler is Borland C++ 5.6.4 for Win32. Thanks in advance

2 1

Re: [Boost-users] boost::thread, boost::bind, and member functions
by Mike Feldmeier 26 May '04

26 May '04

GAH! No change. My %LIB% is now C:\STLport-4.6.2\lib;C:\Program Files\Microsoft Visual Studio .NET 2003\SDK\v1.1\Lib\ My %INCLUDE% is now C:\STLport-4.6.2\stlport;C:\Program Files\Microsoft Visual Studio .NET 2003\SDK\v1.1\include\ My rebuilds have been starting with copying dlls and libs to /Boost, so I blew away the bin directory and the target /Boost directory and watched it rebuild each. I checked a few of the other boost dlls as well, they basically have the same dependencies. >From: GoochRules! <goochrules(a)hotmail.com> >Reply-To: boost-users(a)lists.boost.org >To: boost-users(a)lists.boost.org >Subject: Re: [Boost-users] boost::thread, boost::bind, and member functions >Date: Tue, 25 May 2004 20:10:26 -0600 > >>Sure enough, running depends on the boost dll shows MSVCP71D.DLL, >>MSVCR71D.DLL and KERNEL32.DLL, but no STLPORT! > >To me it looks like your STLport include/lib directories come after the >Platform SDK/VS.Net2003 directories, which means STLport is not being used >at all, hence it is not listed in the dependencies. > >Try: >set INCLUDE=$STLport_include_path;%INCLUDE% >set LIB=$STLport_lib_path;%LIB% > >--mpeltzer > > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users _________________________________________________________________ Stop worrying about overloading your inbox - get MSN Hotmail Extra Storage! http://join.msn.click-url.com/go/onm00200362ave/direct/01/

1 0

Re: [Boost-users] boost::thread, boost::bind, and member functions
by Mike Feldmeier 26 May '04

26 May '04

>From: GoochRules! <goochrules(a)hotmail.com> >Reply-To: boost-users(a)lists.boost.org >To: boost-users(a)lists.boost.org >Subject: Re: [Boost-users] boost::thread, boost::bind, and member functions >Date: Tue, 25 May 2004 12:37:24 -0600 > >>Unfortunately, this whole roadtrip started with the need for STL >>collections to cross DLL boundaries, which is why I am using STLPort. The >>STL that comes with VC++ 7.1 doesn't (with the exception of string and >>vector). When I built boost, I built it with the vc7.1-stlport toolset, >>and I have the stlport library directory in my "additional library path". >>Along this line, I wonder if I need to be doing something special with the >>STLPort DLL? > >One thing to check, if you have the Platform SDK installed, you should have >a 'depends.exe' application... this shows you the dll an exe or dll depends >upon, and even which functions of those dlls are used. Make sure stlport >is one of 'em. If it isn't, make sure that your additional directories >(include and library) are **before** the VS and PSDK directories. Note >that boost will still depend upon msvcpt71.dll and msvcr71.dll (when >compiled with VS.Net2003, but it will use the stlport dll for stl objects. > Sure enough, running depends on the boost dll shows MSVCP71D.DLL, MSVCR71D.DLL and KERNEL32.DLL, but no STLPORT! I built boost with: set STLPORT_PATH=C:\ (STLPort is in C:\STLport-4.6.2) bjam "-sTOOLS=vc7.1-stlport" "-sBUILD=debug release" install Is this not correct? I don't get any errors indicating that it can't find STLPort. Thanks, Mike _________________________________________________________________ Learn to simplify your finances and your life in Streamline Your Life from MSN Money. http://special.msn.com/money/0405streamline.armx

2 1

RE: [Boost-users] boost::thread, boost::bind, and member functions
by Mike Feldmeier 25 May '04

25 May '04

>From: "Ben Hutchings" <ben.hutchings(a)businesswebsoftware.com> >Reply-To: boost-users(a)lists.boost.org >To: <boost-users(a)lists.boost.org> >Subject: RE: [Boost-users] boost::thread, boost::bind, and member functions >Date: Tue, 25 May 2004 17:32:10 +0100 > >Mike Feldmeier <pooky_bunny(a)hotmail.com> wrote: >[ Ben Hutchings wrote:] > > > Mike Feldmeier <pooky_bunny(a)hotmail.com> wrote: ><snip> > > > > This consistently produces the linking error: > > > > > > > > converter.obj : error LNK2019: unresolved external symbol > > > > "__declspec(dllimport) public: __thiscall boost::thread::thread( > > > > class boost::function0<void,class > > > > _STL::allocator<class boost::function_base> > const &)" > > > > (__imp_??0thread@boost@@QAE@ABV?$function0@XV?$allocator@Vfunction > > > > _base@boost@@@_STL@@@1@@Z) referenced in function "public: void > > > > __thiscall mytest::test::converter::run(void)" > > > > (?run@converter@test@mytest@@QAEXXZ) > > > > > >I think you aren't linking with the thread library. ><snip> > > I'm using Visual C++ 7.1, but using straight C++, no .NET. I've > > defined the addtional library path (C:\boost\lib) and tried linking > > (in the additional dependencies) with > > boost_thread-vc71-mt-gdp-1_31.lib, boost_thread-vc71-mt-gdp.lib, > > boost_thread-vc71-mt-p-1_31.lib, and boost_thread-vc71-mt-p.lib (one > > at a time, of course), but I get the same error message with all of > > them. ><snip> > >I've worked it out now. In your main program you are using STLport - >notice the "_STL" namespace instead of "std" but Boost.Thread is >built with the library that comes with VC++ 7.1, which of course uses >the "std" namespace. Unless STLport has some specific feature that >you need I suggest you use the standard library that comes with VC++ >7.1; it's a very good implementation. If you must use STLport then >you need to build Boost.Thread (and any other libraries with which >you share standard library types) using it. > Unfortunately, this whole roadtrip started with the need for STL collections to cross DLL boundaries, which is why I am using STLPort. The STL that comes with VC++ 7.1 doesn't (with the exception of string and vector). When I built boost, I built it with the vc7.1-stlport toolset, and I have the stlport library directory in my "additional library path". Along this line, I wonder if I need to be doing something special with the STLPort DLL? >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users _________________________________________________________________ MSN Toolbar provides one-click access to Hotmail from any Web page FREE download! http://toolbar.msn.click-url.com/go/onm00200413ave/direct/01/

2 1

RE: [Boost-users] boost::thread, boost::bind, and member functions
by Ben Hutchings 25 May '04

25 May '04

Mike Feldmeier <pooky_bunny(a)hotmail.com> wrote: [ Ben Hutchings wrote:] > > Mike Feldmeier <pooky_bunny(a)hotmail.com> wrote: <snip> > > > This consistently produces the linking error: > > > > > > converter.obj : error LNK2019: unresolved external symbol > > > "__declspec(dllimport) public: __thiscall boost::thread::thread( > > > class boost::function0<void,class > > > _STL::allocator<class boost::function_base> > const &)" > > > (__imp_??0thread@boost@@QAE@ABV?$function0@XV?$allocator@Vfunction > > > _base@boost@@@_STL@@@1@@Z) referenced in function "public: void > > > __thiscall mytest::test::converter::run(void)" > > > (?run@converter@test@mytest@@QAEXXZ) > > > >I think you aren't linking with the thread library. <snip> > I'm using Visual C++ 7.1, but using straight C++, no .NET. I've > defined the addtional library path (C:\boost\lib) and tried linking > (in the additional dependencies) with > boost_thread-vc71-mt-gdp-1_31.lib, boost_thread-vc71-mt-gdp.lib, > boost_thread-vc71-mt-p-1_31.lib, and boost_thread-vc71-mt-p.lib (one > at a time, of course), but I get the same error message with all of > them. <snip> I've worked it out now. In your main program you are using STLport - notice the "_STL" namespace instead of "std" but Boost.Thread is built with the library that comes with VC++ 7.1, which of course uses the "std" namespace. Unless STLport has some specific feature that you need I suggest you use the standard library that comes with VC++ 7.1; it's a very good implementation. If you must use STLport then you need to build Boost.Thread (and any other libraries with which you share standard library types) using it.

1 0

RE: [Boost-users] boost::thread, boost::bind, and member functions
by Mike Feldmeier 25 May '04

25 May '04

>From: "Ben Hutchings" <ben.hutchings(a)businesswebsoftware.com> >Reply-To: boost-users(a)lists.boost.org >To: <boost-users(a)lists.boost.org> >Subject: RE: [Boost-users] boost::thread, boost::bind, and member functions >Date: Tue, 25 May 2004 12:07:32 +0100 > >Mike Feldmeier <pooky_bunny(a)hotmail.com> wrote: > > I've seen variations of this come up in this newsgroup and elsewhere, > > but I have yet to find an answer that works for me. > > > > class threaded_manager > > { > > public: > > void start(void); > > }; > > > > threaded_manager mgr; > > > > boost::thread(boost::bind(&threaded_manager::start, &mgr)); > > > > This consistently produces the linking error: > > > > converter.obj : error LNK2019: unresolved external symbol > > "__declspec(dllimport) public: __thiscall boost::thread::thread(class > > boost::function0<void,class > > _STL::allocator<class boost::function_base> > const &)" > > (__imp_??0thread@boost@@QAE@ABV?$function0@XV?$allocator@Vfunction_bas > > e@boost@@@_STL@@@1@@Z) referenced in function "public: void __thiscall > > mytest::test::converter::run(void)" > > (?run@converter@test@mytest@@QAEXXZ) > >I think you aren't linking with the thread library. Many of the Boost >libraries (such as Boost.Bind) include their implementation entirely in >header files, so they will be compiled into your program as needed. >Boost.Thread doesn't; you need to build it as a library (and on Windows >it must be a DLL) and link it with your program. > I'm using Visual C++ 7.1, but using straight C++, no .NET. I've defined the addtional library path (C:\boost\lib) and tried linking (in the additional dependencies) with boost_thread-vc71-mt-gdp-1_31.lib, boost_thread-vc71-mt-gdp.lib, boost_thread-vc71-mt-p-1_31.lib, and boost_thread-vc71-mt-p.lib (one at a time, of course), but I get the same error message with all of them. > > Separating the two boost functions on to separate lines works fine: > > > > boost::bind(&threaded_manager::start, &mgr); > > boost::thread(); > > > > ..which indicates to me that bind is returning an object that is not > > compatible with thread, but I've seen multiple examples of just this > > type of call. ><snip> > >If that were the case you should see a compiler error. The two lines >above may work because the default constructor for boost::thread is >compiled inline whereas the other constructor must be linked in. >Having said that, though, neither v1.30 nor v1.31 has the default >constructor definition in the header, so I don't entirely understand >what's going on. Did you perhaps mistakenly declare a function >returning a boost::thread? > Thus far, this line (or two lines) are the only boost-related lines in my code. >Ben. > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users Thanks, Mike _________________________________________________________________ Get 200+ ad-free, high-fidelity stations and LIVE Major League Baseball Gameday Audio! http://radio.msn.click-url.com/go/onm00200491ave/direct/01/

1 0

RE: [Boost-users] Regex++ Installation Problems
by Ben Hutchings 25 May '04

25 May '04

kinshuk srivastava <srivastavakinshuk(a)yahoo.co.in> wrote: > I had written earlier too regarding installation > problems with the regex++...sorry for not providing > the complete details last time..But I am again sending > the errors... > The Gcc compiler which m using is gcc3.2 nd the > operating system is SuSE 9.6...I think the regex++ > version is 331... > the following are the errors: <snip> > ../../../boost/regex/v3/regex_match.hpp:1937: Internal compiler error > in expand_call, at calls.c:3049 > Please submit a full bug report, > with preprocessed source if appropriate. > See <URL:http://www.gnu.org/software/gcc/bugs.html> > for instructions. Why didn't you just do what the error message said? You should have found this existing bug report: <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8701>. Anyway, this turns out to be a bug that SuSE introduced to its version of GCC. Go to them for a fix.

1 0

RE: [Boost-users] boost::thread, boost::bind, and member functions
by Ben Hutchings 25 May '04

25 May '04

Mike Feldmeier <pooky_bunny(a)hotmail.com> wrote: > I've seen variations of this come up in this newsgroup and elsewhere, > but I have yet to find an answer that works for me. > > class threaded_manager > { > public: > void start(void); > }; > > threaded_manager mgr; > > boost::thread(boost::bind(&threaded_manager::start, &mgr)); > > This consistently produces the linking error: > > converter.obj : error LNK2019: unresolved external symbol > "__declspec(dllimport) public: __thiscall boost::thread::thread(class > boost::function0<void,class > _STL::allocator<class boost::function_base> > const &)" > (__imp_??0thread@boost@@QAE@ABV?$function0@XV?$allocator@Vfunction_bas > e@boost@@@_STL@@@1@@Z) referenced in function "public: void __thiscall > mytest::test::converter::run(void)" > (?run@converter@test@mytest@@QAEXXZ) I think you aren't linking with the thread library. Many of the Boost libraries (such as Boost.Bind) include their implementation entirely in header files, so they will be compiled into your program as needed. Boost.Thread doesn't; you need to build it as a library (and on Windows it must be a DLL) and link it with your program. > Separating the two boost functions on to separate lines works fine: > > boost::bind(&threaded_manager::start, &mgr); > boost::thread(); > > ..which indicates to me that bind is returning an object that is not > compatible with thread, but I've seen multiple examples of just this > type of call. <snip> If that were the case you should see a compiler error. The two lines above may work because the default constructor for boost::thread is compiled inline whereas the other constructor must be linked in. Having said that, though, neither v1.30 nor v1.31 has the default constructor definition in the header, so I don't entirely understand what's going on. Did you perhaps mistakenly declare a function returning a boost::thread? Ben.

1 0

Boost.Test release notes typos
by Ross Boylan 25 May '04

25 May '04

I noticed a couple of typos in the latest release notes for Boost.Test, at http://cvs.sourceforge.net/viewcvs.py/boost/boost/libs/test/doc/release_not… It now reads, near the top, <LI>Floating point comaprison algorithm reworked (Is not backward compatible!!!) <UL> <LI><u>tolerance now expected in persent units</u></LI> That should be comparison in the first line and percent in the last.

2 1

boost::thread, boost::bind, and member functions
by Mike Feldmeier 24 May '04

24 May '04

I've seen variations of this come up in this newsgroup and elsewhere, but I have yet to find an answer that works for me. class threaded_manager { public: void start(void); }; threaded_manager mgr; boost::thread(boost::bind(&threaded_manager::start, &mgr)); This consistently produces the linking error: converter.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: __thiscall boost::thread::thread(class boost::function0<void,class _STL::allocator<class boost::function_base> > const &)" (__imp_??0thread@boost@@QAE@ABV?$function0@XV?$allocator@Vfunction_base@boos t@@@_STL@@@1@@Z) referenced in function "public: void __thiscall mytest::test::converter::run(void)" (?run@converter@test@mytest@@QAEXXZ) Separating the two boost functions on to separate lines works fine: boost::bind(&threaded_manager::start, &mgr); boost::thread(); ..which indicates to me that bind is returning an object that is not compatible with thread, but I've seen multiple examples of just this type of call. This is probably something simple, my two days of experience with Boost just doesn't appear to be sufficient. What in the world am I doing wrong? Thanks, Mike _________________________________________________________________ Stop worrying about overloading your inbox - get MSN Hotmail Extra Storage! http://join.msn.click-url.com/go/onm00200362ave/direct/01/

1 0

Win32 Non-greedy Regex "\w*?" matches whitespace (tab or space)
by John F Meinel Jr 24 May '04

24 May '04

When changing the "search for word characters" from greedy to non-greedy, it starts to match whitespace characters. Attached is an example program. If you remove the '?' to make the search greedy, then it (correctly) doesn't match. I am using boost 1.31.0, on WinXP using VC7.1. I have confirmed that this does not occur on my Linux machine. Also attached is my Jamfile to make sure I didn't just compile it wrong. It occurs in both debug and release mode. John =:-> project-root ; exe test : test.cpp #unit_test_example1.cpp : <sysinclude>$(BOOST_ROOT) <stlport-iostream>on <library-path>$(BOOST_ROOT)/../../lib <vc7.1><debug><find-library>libboost_regex-vc71-mt-gd <vc7.1><release><find-library>libboost_regex-vc71-mt <gcc><debug><find-library>boost_regex-gcc-mt-d <gcc><release><find-library>boost_regex-gcc-mt <threading>multi ; # vim: ts=4 #include <boost/regex.hpp> #include <iostream> boost::regex re("(\\w*?)"); void doesMatch(std::string s) { boost::smatch m; std::cout << "String: \"" << s << "\"" << std::endl; if(boost::regex_match(s, m, re)) { std::cout << "Matched: "; std::cout << "\"" << std::string(m[1]) << "\"" << std::endl; } else { std::cout << "Didn't match" << std::endl; } std::cout << std::endl; } int main(int argc, char **argv) { if(argc > 1) for(int i = 1; i < argc; ++i) doesMatch(argv[i]); else { char *arg[] = {"", "should", "should\tnot", "shou_ld", "should not" ," doesn't " }; main(sizeof(arg)/sizeof(char*), arg); } }

2 1

boost/test/detail/algorithm.hpp: request for promotion and/or fast review
by Gennadiy Rozental 23 May '04

23 May '04

Hi, I added several algorithms that extent STL algorithms set. I use them in my Boost.test development. But since they seems to be generic and IMO should be present in STL I propose to push them in boost/algorithm.hpp (or under boost/algorithm). I remember there was a discussion already about these algorithms. Here is the list (I hope name and spec is self -explanatory): template <class InputIter1, class InputIter2> inline std::pair<InputIter1, InputIter2> mismatch( InputIter1 first1, InputIter1 last1, InputIter2 first2, InputIter2 last2 ); template <class InputIter1, class InputIter2, class Predicate> inline std::pair<InputIter1, InputIter2> mismatch( InputIter1 first1, InputIter1 last1, InputIter2 first2, InputIter2 last2, Predicate pred ); template<class ForwardIterator1, class ForwardIterator2> inline ForwardIterator1 find_first_not_of( ForwardIterator1 first1, ForwardIterator1 last1, ForwardIterator2 first2, ForwardIterator2 last2 ); template<class ForwardIterator1, class ForwardIterator2, class Predicate> inline ForwardIterator1 find_first_not_of( ForwardIterator1 first1, ForwardIterator1 last1, ForwardIterator2 first2, ForwardIterator2 last2, Predicate pred ); template<class BidirectionalIterator1, class ForwardIterator2> inline BidirectionalIterator1 find_last_of( BidirectionalIterator1 first1, BidirectionalIterator1 last1, ForwardIterator2 first2, ForwardIterator2 last2 ) template<class BidirectionalIterator1, class ForwardIterator2, class Predicate> inline BidirectionalIterator1 find_last_of( BidirectionalIterator1 first1, BidirectionalIterator1 last1, ForwardIterator2 first2, ForwardIterator2 last2, Predicate pred ); template<class BidirectionalIterator1, class ForwardIterator2> inline BidirectionalIterator1 find_last_not_of( BidirectionalIterator1 first1, BidirectionalIterator1 last1, ForwardIterator2 first2, ForwardIterator2 last2 ); template<class BidirectionalIterator1, class ForwardIterator2, class Predicate> inline BidirectionalIterator1 find_last_not_of( BidirectionalIterator1 first1, BidirectionalIterator1 last1, ForwardIterator2 first2, ForwardIterator2 last2, Predicate pred ); Implementation together with doxigen docs is here: http://tinyurl.com/35pfj Test is here: http://tinyurl.com/yvlgu If we don't find this header eligible for review I would like to ask then for permision to at least put this algorithm in boost namespace (currently boost::unit_test), for it inconvenience for me to keep it in deep namespace. Gennadiy.

1 0

Problems with
by Pete Cilliers 23 May '04

23 May '04

Hi I am having problems running on MSVC 7.1 the example given in The parameterized boost::function based test case #include <boost/test/unit_test_ex.hpp> using boost::unit_test_framework::test_suite; #include <boost/bind.hpp> void test_mask( int arg, int mask ) { BOOST_CHECK( (arg & mask) != 0 ); } struct sub_test_suite : public test_suite { sub_test_suite() { parameters_list.push_back( 1 ); parameters_list.push_back( 5 ); parameters_list.push_back( 6 ); parameters_list.push_back( 7 ); parameters_list.push_back( 15 ); add( BOOST_PARAM_TEST_CASE( bind( &test_mask, _1, 0x80 ), parameters_list.begin(), parameters_list.end() ) ) ); } std::list<int> parameters_list; } //__________________________________________________________________________ __// test_suite* init_unit_test_suite( int argc, char* argv[] ) { test_suite* test = BOOST_TEST_SUITE( "Example" ); test->add( new sub_test_suite ); return test; } I get the following error: error C2661: 'boost::unit_test_framework::create_test_case' : no overloaded function takes 4 arguments I have had to make a number of changes to the example to correct typo's but the example given here is as it appears in the documentation. Has anyone experienced the same type of problem and can you please advise on how to correct? Thanks ********************************************************************** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept by MIMEsweeper for the presence of computer viruses. **********************************************************************

2 1

check() function?
by Mark Dawson 21 May '04

21 May '04

I'm trying to compile boost 1.3.0 using the CodeWarrior 9.2 compiler on Mac OS X (10.3.3). I've run into a weird error: Error : unexpected argument while expanding macro 'check' (wanted 1) (included from: boost/type_traits.hpp:29 is_base_and_derived.hpp line 112 s_type check(D const volatile *, T); is_base_and_derived.hpp: template <typename B, typename D> struct bd_helper { template <typename T> static type_traits::yes_type check(D const volatile *, T); static type_traits::no_type check(B const volatile *, int); }; When I look up check() its in my /usr/include/AssertMacros.h file, it DOES takes one parameter. Am I somehow including the wrong paths, and (should be) getting check() from somewhere else? Thanks, mark mcdawson(a)mac.com

2 1