June 2008 - Boost-users - lists.boost.org

[Graph] Problem Compiling BGL-Python
by ZSS 11 Jun '08

11 Jun '08

Hi, I'm facing difficulty in compiling BGL-Python (which I'm not familiar with). I followed all the instructions in the README file, however, when I reach the stage where I need to run bjam I recieve this error: error: Could not find parent for project at '.' error: Did not find Jamfile.jam or Jamroot.jam in any parent directory. However, when I move the bgl-python-0.9 directory and put it under boost_1_35_0 I get a different error message: Jamfile:13: in modules.load rule project-root unknown in module Jamfile I don't know what to do, any hints are highly appreciated. Thank you, Zee

1 0

graph visualization
by Dennis Yurichev 10 Jun '08

10 Jun '08

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, I'm looking for some open-source solution which is able to vizualize boost::graph structures. I mean, I develope software which use boost::graph extensively and I wish to offer end-user to see it visually and edit it. It's not a problem to export to some GDL file for example, and use external tool, but user interaction should be present. Are there anything ready-made or I doomed to develope my own? P.S. Target is Win32 GDI API. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) iD8DBQFITurt1YPmFmJG++MRAmYTAKDe0rEYszytT7f70y6up3f5js45ngCfTQHP uJ5SJhSb0tKlNG62DwwMJoI= =193P -----END PGP SIGNATURE-----

3 2

darwin universal static libs for release and debug for 1.35.0
by Jeff Flinn 10 Jun '08

10 Jun '08

Trying to build universal static libs for darwin. bjam --toolset=darwin --architecture=combined stage appears to only build libboost_*_mt-1_35.dylib's and libboost_test_exec_monitor-mt-1_35.a. Whereas the default for 1.34.1 was to build all libs. I need to build release and debug static libs. I'm having problems finding the command line flags in the docs. This is my first foray into mac/unix builds for boost. Any help is appreciated. Thanks, Jeff

1 2

boost filesystems and case issues
by brad＠16systems.com 10 Jun '08

10 Jun '08

I've looked through the documentation and Googled, but I cannot find any info on the topic. If I want to find a file named 'System.dll', but on some systems the file is named 'system.DLL' or 'SYSTEM.DLL', etc. If I could could do something like this: if (boost::filesystem::is_regular(itr->path())) { std::cout << itr->path().lower() << std::endl; } Notice the .lower() In this way, each path would be lowercase and I could search for only 'system.dll' and .lower() would find any occurrence of that filename, no matter the case. I wonder how others do this today? Do you use transform or tolower from another library? Thank you

4 3

A question about boost::bind function
by 风舞影天 10 Jun '08

10 Jun '08

I'm a newbie in boost library and sorry for my poor English. I have a project now, it was created in Visual Studio.Net 2003 and the version of the boost library is 1.33.1, all thing is good. But, When I replaced the boost library with the new version 1.34.1 or 1.35.0, the project can't compiled correct. The error message is 'can't convert const sometype& to sometype&'. So, i force myself to read the source code of boost library and compare them betwwen version 1.33.1 and version 1.44.1. at last, I found the problem and written a test program. The following code is the test program: #include "stdafx.h" #include <boost/bind.hpp> #include <vector> #include <algorithm> struct _sUserData { int m_nUserId; int m_nUserAge; _sUserData() : m_nUserId( rand() ) , m_nUserAge( rand()%100+1 ) { } }; void AddUserAge( int& age ) { ++age; } int _tmain(int argc, _TCHAR* argv[]) { std::vector<_sUserData> vUserData; for( int i = 0; i < 10; ++i ) vUserData.push_back( _sUserData() ); std::for_each( vUserData.begin(), vUserData.end(), boost::bind( AddUserAge, boost::bind( &_sUserData::m_nUserAge, _1 ) ) ); return 0; } If i change the function AddUserAge( int& age ) to AddUserAge( const int& age ), all things is good. but I need change this parameter. In the version 1.33.1, the data member formal boost::bind function was defined as following: template<class R, class T, class A1> _bi::bind_t< R, _mfi::dm<R, T>, typename _bi::list_av_1<A1>::type > BOOST_BIND(R T::*f, A1 a1) { typedef _mfi::dm<R, T> F; typedef typename _bi::list_av_1<A1>::type list_type; return _bi::bind_t<R, F, list_type>( F(f), list_type(a1) ); } In the version 1.35.0, the boost::bind function was defined as following: template< class A1, class M, class T > _bi::bind_t< typename _bi::dm_result< M T::*, A1 >::type, _mfi::dm<M, T>, typename _bi::list_av_1<A1>::type > BOOST_BIND( M T::*f, A1 a1 ) { typedef typename _bi::dm_result< M T::*, A1 >::type result_type; typedef _mfi::dm<M, T> F; typedef typename _bi::list_av_1<A1>::type list_type; return _bi::bind_t< result_type, F, list_type >( F( f ), list_type( a1 ) ); } Although the code that is similar to the version 1.33.1 is existed, but it is written in a macro condition, and in my environment is invalid, so ignore it. The difference is the return type. The boost::_bi::dm_result is defined as following: namespace _bi { template< class Pm, int I > struct add_cref; template< class M, class T > struct add_cref< M T::*, 0 > { typedef M type; }; template< class M, class T > struct add_cref< M T::*, 1 > { typedef M const & type; }; template<class R> struct isref { enum value_type { value = 0 }; }; template<class R> struct isref< R& > { enum value_type { value = 1 }; }; template<class R> struct isref< R* > { enum value_type { value = 1 }; }; template<class Pm, class A1> struct dm_result { typedef typename add_cref< Pm, 1 >::type type; }; template<class Pm, class R, class F, class L> struct dm_result< Pm, bind_t<R, F, L> > { typedef typename bind_t<R, F, L>::result_type result_type; typedef typename add_cref< Pm, isref< result_type >::value >::type type; }; } the compiler will convert the result type to const reference, then pass it to user's function, the error is caught. Now, all things is come out. But I don't understand why add const reference to the result type and the behavior is difference with the version 1.33.1. Wait for your answer and thanks for your time, and sorry for my poor in English:)

4 4

[boost-users][tuple, bind] binding tuple element
by Igor R 10 Jun '08

10 Jun '08

Hello, Given a vector of tuples, like this: vector<tuple<int, std::string> > tuples_; I want create a binder to find a tuple by one of its "attributes". If it would be a struct: struct Tuple { int id_; string name_; }; ...I'd use it like this: find_if(tuples_.begin(), tuples_.end(), bind(&Tuple::name_, _1) == "kuku"); How can bind tuples::get<1>() function? Thanks a lot, Igor'.

2 1

More Granularity in boost::serialization "stream error"?
by eric 10 Jun '08

10 Jun '08

Hello, How can I get more detail when boost::serialization raises the exception "stream error"? I find that this one error can result from a wide variety of dissimilar root causes, but regardless of the nature of the problem (or the size of the buffer) the only information provided by boost::serialization is that one opaque string. It is easy to lose a lot of time tracking down the problem. I have worked with other more traditional XML parsers that raise exceptions in the form "Unexpected character in line x column y." I am aware that the architecture of boost::serialization is completely different to that of a traditional XML parser but the bottom line is the same, if I could get more granularity on this exception it would save a lot of time troubleshooting. Debugging seems not to be an option as my debugger (gdb) gets lost in boost header files and templates. I am left with trial and error, chopping up the stream into smaller pieces until I can isolate the invalid content. Any ideas? Thanks, Eric

1 0

Create boost_1_33_1 boost::archive::text_iarchive with an empty stream
by Gancho Manev 10 Jun '08

10 Jun '08

Hi, The following code does not throw archive_exception always: std::stringstream ss; boost::archive::text_iarchive ia(ss); The archive_exception is thrown always on VC++2005/Windows while on GCC3.2.2/Redhat8 and GCC3.4.6/CentOS a std::length_error: basic_string::resize is thrown sometimes. That means the constructor tries to resize some string with value > basic_string.max_size(). The GCC defines the max_size() to be only about a quarter of the max length the allocator can get. That explains why that exception happens to be raised on GCC. A search of the source code lead me to the overloaded methods 'load' of text_iarchive_impl<Archive>. They look suspicious because do not check the input stream state before using the string size just loaded. The local variable is not initialized nor loaded, so it contains some arbitrary value! In the method load(std::string&) that might cause std::length_error while in load(char *s) - even program crash. The code in the next Boost versions does not seem to be changed. Has anyone encountered such behavior? I've not found that bug reported. Thanks in advance, Gancho

2 2

Using boost::thread with windows events
by Sean Farrow 10 Jun '08

10 Jun '08

Hi: I am neding to use boost::thread with windows events, dos anyone know whether this is possible/whether there are any examples. Basically, I need to exit the thread when a variable becomes fales. Any help apreciated. Sean. __________ Information from ESET NOD32 Antivirus, version of virus signature database 3171 (20080610) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com

1 0

yet another logging library
by Mojmir Svoboda 10 Jun '08

10 Jun '08

Hello, i'm trying to create a logging library framework. It is my second attempt, the first can be seen at http://www.codeplex.com/flog . There is some documentation in the doc directory that is perhaps worth reading. The first version used boost::iostreams and not very smart locking mechanism in mt environment (in terms of performance mainly), but served more as a verification 'it can be done and it is not that ugly as i thought' than a real library candidate. Now i'd like to focus on performance and code size, so I made a sketch of the next version. It is really small and is rather a model than anything usable, but before I start implementation, I'd like to know opinion of library writers and users - what do you think about the concept? is it worth continuing, or i'd better go doing something better? :) The sketch of the next version can be seen at http://www.codeplex.com/fflog It's meant to be illustration of the interface and it's usage itself. Short summary of the design: User specifies compile-time vector of requested filters that apply to the logged text. Each filter requiring user's input is tagged and when logging function is invoked, the supplied types are checked against expected type, set to filter and later written to a buffer. Buffers are managed by the sink_manager and likely is one of few places that has to be mt-safe, as the rest is on the stack. I thought about some lock-free queues of buffers that are assigned to caller and filled returned back to write. The actual write of the buffer can be performed in dedicated lower-priority thread. Run-time filtering like level or context is possible, but not implemented yet. The main idea can be seen in the 1st version in runtime_filtering.hpp. Again, runtime values are shared among threads, so they have to be protected in mt environment. Many thanks for your attention, Mojmir

6 13