- Boost-users - lists.boost.org

Re: [Boost-users] Serializing boost::variant?
by Merrill Cornish 24 Nov '05

24 Nov '05

Robert, >>> or you just do: >>> #include <boost/serialization/variant.hpp> I knew that a boost::variant data member in serialize() would compile, but--as the mathematicians say--the a necessary, but not sufficient condition. Thanks for verifying it. Merrill

1 0

about installation
by Johnny Wong 24 Nov '05

24 Nov '05

Hi, I am using Borland C++ compiler 5.5, and I need to use the Boost Graph Library. But I am new to C++, so I downloaded the boost_1_33_0 but I don't know how to install and make it work. I have read the getting started guide, but I cannot follow the instructions. Can anybody kindly tell me how to do? Thanks. _______________________________________ �x��ӍϢ - Yahoo! Messenger ��]��ϾW��Կ��ӍϢ�o�㣬��ϾW�r��κ��fԒ��ʧ�� http://messenger.yahoo.com.hk

1 0

Using Boost 1.33 with STLport 5.0 on MSVC71
by Wu Yinghui, Freddie 24 Nov '05

24 Nov '05

Dear all, I'm having a problem linking the simplest Boost unit_test program on the platform described above. More details here: Compiler: MSVC71 OS: WinXP Pro SP2 STLport 5.0 config: All default as obtained from CVS STLPORT_5_0 branch, except that I use /Zc:wchar_t to use wchar_t as built-in type. Boost 1.33/34: All default as obtained from CVS, built with the following command line: bjam --without-python "-sBUILD=debug release <runtime-link>static <threading>multi <stlport-version>5.0 <native-wchar_t>on <define>BOOST_SIGNALS_NAMESPACE=siglib" stage My simple unit_test program: ===== #include <string> #define BOOST_AUTO_TEST_MAIN #include <boost/test/auto_unit_test.hpp> BOOST_AUTO_UNIT_TEST( Test_Function ) { std::wstring string = L"some test string"; BOOST_CHECK( string.length() == 16 ); } ===== My project setting uses "Multi-threaded DLL (/MD)" in code gen, and links to "libboost_unit_test_framework-vc71-mt-sgdp.lib" in debug mode. My purpose was to use STLport in a DLL, but the whole Boost statically linked. (I know I can link with "libboost_unit_test_framework-vc71-mt-gdp.lib," but I want Boost to be statically linked.) The linker gives the following errors: stlportstld.5.0.lib(stlportstld.5.0.dll) : error LNK2005: "public: __thiscall stlpd_std::__owned_list::~__owned_list(void)" (??1__owned_list@stlpd_std@@QAE@XZ) already defined in libboost_unit_test_framework-vc71-mt-sgdp.lib(unit_test_suite.obj) stlportstld.5.0.lib(stlportstld.5.0.dll) : error LNK2005: "public: __thiscall stlpd_std::basic_string<char,class stlpd_std::char_traits<char>,class stlpd_std::allocator<char> >::~basic_string<char,class stlpd_std::char_traits<char>,class stlpd_std::allocator<char> >(void)" (??1?$basic_string@DV?$char_traits@D@stlpd_std@@V?$allocator@D@2@@stlpd_std@@QAE@XZ) already defined in libboost_unit_test_framework-vc71-mt-sgdp.lib(unit_test_suite.obj) stlportstld.5.0.lib(stlportstld.5.0.dll) : error LNK2005: "public: static void __cdecl stlpd_std::__node_alloc<1,0>::deallocate(void *,unsigned int)" (?deallocate@?$__node_alloc@$00$0A@@stlpd_std@@SAXPAXI@Z) already defined in libboost_unit_test_framework-vc71-mt-sgdp.lib(unit_test_suite.obj) stlportstld.5.0.lib(stlportstld.5.0.dll) : error LNK2005: "public: class stlpd_std::basic_string<char,class stlpd_std::char_traits<char>,class stlpd_std::allocator<char> > & __thiscall stlpd_std::basic_string<char,class stlpd_std::char_traits<char>,class stlpd_std::allocator<char> >::assign(char const *,unsigned int)" (?assign@?$basic_string@DV?$char_traits@D@stlpd_std@@V?$allocator@D@2@@stlpd_std@@QAEAAV12@PBDI@Z) already defined in libboost_unit_test_framework-vc71-mt-sgdp.lib(unit_test_log.obj) <more similar errors removed...> Is there any kind soul here can enlighten me a bit? Is it at all recommended (or even possible) to use STLport in DLL while statically linking with Boost? Cheers, Freddie -- Wu Yinghui, Freddie Research & Development Software Engineer Volume Interactions Pte Ltd 1 Kim Seng Promenade, #12-01 Great World City East Tower Singapore 237994 Tel: +65 62226962 (Ext 216) Fax: +65 62226215 Email: yhwu(a)volumeinteractions.com URL: http://www.volumeinteractions.com Important: This message is intended for the recipient(s) addressed above. It contains privileged and confidential information. If you are not the intended recipient, please notify the sender immediately by replying to this message and then delete it from your system. You must not read, copy, use, or disseminate this communication in any form. Thank you.

2 1

Boost IRIX build problems
by Jeroen Akershoek 24 Nov '05

24 Nov '05

Hello, I'm trying to build boost on our SGI Onyx4 and am running into some problems. I've tried looking up some documentation for configuring the build-system a bit better, but couldn't find much. I don't have the time to dig into it and find it out myself, especially since building boost isn't my ultimate goal for this exercise, so I'm hoping you can help me out here I'm getting (a lot of occurances of) three errors throughout the build: 1) the indicated type is incomplete std::basic_istream<charT> m_is; // stream for parsing numbers. Not sure about this one. Missing include somewhere or maybe SGIs STL is f00bar 2) The indicated enumeration value is out of "int" range. enum npos_type { npos = (size_type)-1 }; I assume it's the same problem as with IBMCPP. Don't know if MIPSpro can have 64-bits enums. 3) Threading is disabled: _SGI_MP_SOURCE is undefined If somebody could help me out with these, I'd be very grateful. My environment is IRIX 6.5.27 with MIPSpro 7.41, bjam 3.1.11 and I'm trying boost 1.33.0. I've modified the tools/build/v1/mipspro-tool.jam to use -64 and -mips4 as default. Probably not the way to do it, but I couldn't find the "proper" way within decent time Kind regards, Jeroen -- Jeroen Akershoek SARA - High Performance Computing & Visualization dep. tel: +31 20 5923000 fax: +31 20 6683167 Draco dormiens nunquam titillandus

4 5

Complaint about Current Boost Build System
by Wu Yinghui, Freddie 24 Nov '05

24 Nov '05

Hi all, Sorry for this e-mail, since it is more of a complaint. But I do feel the urge to raise this problem to the community so that some solution could be derived soon. The current build system (1.33/34 from CVS) creates subdirectories for each variant/ subvariant specified in the BUILD environment. When more user customizations are specified, the consequence of this is very deep directory structures. Nonetheless, some OSes imposes stringent limits on how long a path can go (e.g., Win2k/XP limits a path name to be 259 characters long maximum). There causes big problem with the building system when error can occur from nowhere in Boost source code itself. This constraint effectively limits how much customization a user can make (e.g., I had to remove some additional optimization/ tuning parameters for subvariants, and SUBST'ed the Boost directory in order to reduce the path name length). So my question is, why can we not be smarter in this to derive alternatives to make the directory structure shallower (or, simply reduce the length of the names of each subvariant directory)? Sorry again if this sound too demanding, but I do consider it an important usability limitation of the current Boost build system. Cheers, Freddie -- Wu Yinghui, Freddie Research & Development Software Engineer Volume Interactions Pte Ltd 1 Kim Seng Promenade, #12-01 Great World City East Tower Singapore 237994 Tel: +65 62226962 (Ext 216) Fax: +65 62226215 Email: yhwu(a)volumeinteractions.com URL: http://www.volumeinteractions.com Important: This message is intended for the recipient(s) addressed above. It contains privileged and confidential information. If you are not the intended recipient, please notify the sender immediately by replying to this message and then delete it from your system. You must not read, copy, use, or disseminate this communication in any form. Thank you.

2 1

Serializing boost::variant?
by Merrill Cornish 24 Nov '05

24 Nov '05

If I have a boost::variant data member I want to serialize, it looks like I'm going to have to use boost::static_visitor and boost::apply_visitor in the serialize() method. However, there is some black magic going on with the ar & dataMember; statements. Any pointers as to how I do it? Merrill

2 1

Bind, lambda and how to chain relative function calls
by Kerner, Jacques 24 Nov '05

24 Nov '05

Hi, I would like to create a functor for calling functions in a chain like this : car.GetRigidBody().GetPosition().GetX() What is the best way to do that using bind (and lambda ?) The car.GetRigidBody() is easy : bind(&Car::GetRigidBody, &car) , but then I am a bit stuck. Note that I don't want to store the result of car.GetRigidBody(). I want the final functor to call all the functions everytime its operator() is called. So bind ( &Position::GetX, &bind ( &RigidBody::GetPosition, &bind ( &Car::GetRigidBody, &car )() )() ) would not be acceptable... Any help greatly appreciated Jacques Kerner

3 2

Tuple operator==
by Josue Gomes 23 Nov '05

23 Nov '05

Hi, Take this code: #include <string> #include <boost/tuple/tuple.hpp> struct A {}; bool operator==(A, A) { return true; } int main() { using namespace boost; tuple<std::string, int, A> t1(std::string("same?"), 2, A()); tuple<std::string, long, A> t2(std::string("same?"), 2, A()); tuple<std::string, long, A> t3(std::string("different"), 3, A()); t1 == t2; // error: no match for operator== t1 == t3; // error: no match for operator== } This is similar to the example shown in Boost.Tuple docs. However the compiler(*) refuses to compile it. I tried it with versions 1.33.0 and 1.33.1 (Beta) Any ideas? Compiler outputs below: MSC C/C++ Compiler 13.10;3077 cl /GR /EHsc /IE:\projects\libraries\boost_1_33_1 tuple_error.cpp Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 13.10.3077 for 80x86 Copyright (C) Microsoft Corporation 1984-2002. All rights reserved. tuple_error.cpp tuple_error.cpp(17) : error C2678: binary '==' : no operator found which takes a left-hand operand of type 'boost::tuples::tuple<T0,T1,T2>' (or there is no acce ptable conversion) with [ T0=std::string, T1=int, T2=A ] tuple_error.cpp(18) : error C2678: binary '==' : no operator found which takes a left-hand operand of type 'boost::tuples::tuple<T0,T1,T2>' (or there is no acce ptable conversion) with [ T0=std::string, T1=int, T2=A ] gcc 3.4.4 g++ -I E:\projects\libraries\boost_1_33_1 tuple_error.cpp -o tuple_error_gcc.exe tuple_error.cpp: In function `int main()': tuple_error.cpp:17: error: no match for 'operator==' in 't1 == t2' tuple_error.cpp:7: note: candidates are: bool operator==(A, A) tuple_error.cpp:18: error: no match for 'operator==' in 't1 == t3' tuple_error.cpp:7: note: candidates are: bool operator==(A, A) Regards, Josue

2 1

Re: [Boost-users] Collections of shared_ptr<Base> and shared_ptr<Derived>
by Keith Weintraub 23 Nov '05

23 Nov '05

Michael, Thanks for the reply. You are absolutely right. I'm coming back to C++ after years of Java and I haven't been careful enough about reading the stl docs. One quibble with your solution is that I am doing an upcast not a down cast, but your solution worked perfectly for me. Thanks all, KW ________________________________________________ Date: Mon, 21 Nov 2005 15:29:29 -0800 From: Michael Rickert <pandamojo(a)gmail.com> Subject: Re: [Boost-users] Collections of shared_ptr<Base> and shared_ptr<Derived> To: boost-users(a)lists.boost.org Message-ID: <561a17f20511211529l1a528314od7905edcc2aa2ab7(a)mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 On 11/21/05, Keith Weintraub <Keith(a)archeuscap.com> wrote: > > > > Folks, > > I have come up with the following solution in template form. I would be ecstatic to receive any feedback on either my solution or original question. I am not an expert in these things and I can use all the help any generous soul could provide. > > > > Here is my solution: *snip* It would seem you're overengineering things. It'd be simpler just to use std::vector's iterator constructors: std::vector< boost::shared_ptr< derived > > data; ... std::vector< boost::shared_ptr< base > > downcast_data( data.begin() , data.end() ); If you wanted to go with your original version, I'd suggest returning a managed pointer of some sort (std::auto_ptr or boost::shared_ptr) to decrease the likelyhood of a leak - or having the user pass a nonconstant reference to a vector to fill. Sorry if this has allready been covered by the time this is recieved, my posts seem to still be awaiting moderator approval. -Mike

2 1

[Announce] Boost Graph Library Python Bindings Version 0.9
by Douglas Gregor 23 Nov '05

23 Nov '05

Version 0.9 of the Boost Graph Library Python bindings has been released. The BGL-Python bindings provide access to many of the features of the C++ Boost Graph Library from Python, blending the rapid prototyping and application development capabilities of Python with the power and performance of C++. BGL-Python is the successor to the experimental Python bindings found in Boost 1.33.0, providing more algorithms, better data structures, and drastically improved documentation. The BGL-Python bindings are available at: http://www.osl.iu.edu/~dgregor/bgl-python/ Doug

1 0