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

converting a string to a function name
by Fred J. 09 Nov '06

09 Nov '06

Hi is there a way to convert a string to a function name and fire it. like void his_fun(){ cout << "his is here" << endl; } vector<string> vec; vec.push_back("his"); vec.push_back("me"); for(i=0; i<vec.size(); i++) string var = vec[i] + _fun; fire(var); and it will fire the routine "his_fun? is so, how? what whole idea is to avoid typing vector<func_ptr> vec; vec.push_back( &func_1 ); vec.push_back( &func_n ); ... thanks --------------------------------- Everyone is raving about the all-new Yahoo! Mail.

2 1

got a bad_any_cast exception
by ClassyK_Danad 09 Nov '06

09 Nov '06

#include <iostream> #include <boost/any.hpp> template<typename A, typename B> boost::any min(const A& a,const B& b) { return a < b ? a : b; } void main() { //got a bad_any_cast exception ! std::cout << boost::any_cast<int>( min( 1, 1.1)) << std::endl; } how could this happen?

3 3

Re: [Boost-users] Socket and iostream
by Nat Goodspeed 09 Nov '06

09 Nov '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Jakub Stachowski > Sent: Wednesday, November 08, 2006 2:22 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Socket and iostream > > > nothing seems to be written nor readen to/from the socket except for > > the first getline. What am I doing wrong here ? > > Try flushing the stream after each write. [Nat] I wondered about that -- but doesn't std::endl implicitly flush?

3 2

Re: [Boost-users] Sequence of Types, What For?
by james.jones＠firstinvestors.com 09 Nov '06

09 Nov '06

From: "Andrew Holden" <aholden(a)charteroaksystems.com> > 3: Look up the value in the first array, and then look up the code block > number in the second array. > > 4: Go to the pointer you just got. > > The actual blocks of code are laid out in order right after the goto in > step 4, in the same order as the source code. Breaks become goto end of > switch. If you leave out a break, the processor really does JUST fall > through to the next case. > > All in all, if you have a monster switch statement, the > compiler-generated code probably is just about as fast as you can get. > It takes less than a dozen machine instructions to implement it, and it > has constant complexity. Actually this is not particularly good. For one thing, it's O(n), not constant time, because doing an array lookup is O(n). For many switch use-cases this is no big deal, but it certainly isn't unheard of to have a switch with dozens (or more) of cases, each of which does almost nothing (like return a value). In this case, the lookup can actually dominate your runtime. There's a theorem in hash theory (proved by Komlos, one of my old profs) that says that you can always generate a perfect hash function if you know the values you need to hash. That is, you can generate a hash function that requires exactly one lookup in every case. But I don't know any compiler that actually does this. Since all cases are known at compile time, doing this should be a beneficial runtime optimization for large switches (if the generated hash function is sufficiently cheap). Maybe this could be done using the MPL? I don't know TMP well enough to know where to begin on something like this. See http://en.wikipedia.org/wiki/Perfect_hash_function for more detail. - James Jones Administrative Data Mgmt. Webmaster 375 Raritan Center Pkwy, Suite A Data Architect Edison, NJ 08837

5 5

BOOST_FOREACH VC8 warning: change in behavior
by Nico Galoppo 08 Nov '06

08 Nov '06

Hi, I'm getting the following warning in VC8, using BOOST_FOREACH on a container of shared_ptr<>s. The container itself is also a shared_ptr. warning C4686: 'boost::foreach_detail_::make_probe' : possible change in behavior, change in UDT return calling convention The warning happens at the following line shared_ptr<Scene> scene(new Scene()); typedef shared_ptr<Character> character_ptr; BOOST_FOREACH(character_ptr & actor, *scene) { ... } Unfortunately, I can't reproduce it in a simple program, it doesn't give me the warning there. --nico

2 2

Re: [Boost-users] [Boost.Filesystem] problem to link an executabl e to boost_filesys tem static library on Windows XP
by Johan.GUIHENEUF＠external.thalesgroup.com 08 Nov '06

08 Nov '06

Hello Rush and everybody, Thank you for your answer. Unfortunately, it does NOT match the name. The name computed "by auto_link.hpp" is boost_filesystem-vc71-mt-s-1_33.lib and the name of the static library on my computer is boost_filesystem-vc71-mt-1_33.lib (without the '-s'). So my problem is a problem of name mismatch, not a problem of project configuration for linking. Thank you. Johan GUIHENEUF -----Message d'origine----- De : boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] De la part de Rush Manbert Envoyé : mardi 7 novembre 2006 20:40 À : boost-users(a)lists.boost.org Objet : Re: [Boost-users] [Boost.Filesystem] problem to link an executable to boost_filesys tem static library on Windows XP Johan.GUIHENEUF(a)external.thalesgroup.com wrote: > Hello, > > I try to build an console executable that uses boost_fileystem library. > I have got the file boost_filesystem-vc71-mt-1_33.lib. My IDE is Visual C++ 7.1. > I set the preprocessor definition _MT. > At build step, I have this error: "could not find boost_filesystem-vc71-mt-s-1_33.lib". > I had a look to boost/config/auto_link.hpp to see what "#define" I forgot but I do not understand why it chooses "#define BOOST_LIB_RT_OPT -s". > > hat preprocessor macro should I define to make it choose "#define BOOST_LIB_RT_OPT" (i.e. empty) ? Is there a blocking problem with Visual C++ 7.1 on Windows XP ? > Hi Johan, If you have verified that the library name generated by the auto link (boost_filesystem-vc71-mt-s-1_33.lib) matches what was installed by the boost build, then it looks like you didn't define the path to the library for the linker. This should be settable via the linker preferences, but I use VS 2005, so don't know how you would do it. - Rush _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

3 2

unit_test_result: is it deprecated?
by Stephen Torri 08 Nov '06

08 Nov '06

Is the class unit_test_result deprecated? I have boost 1.33.1 installed and it is not a part of the installation. Is there something else I am suppose to be using? Stephen

2 5

[regex] boost, problem compiling doc example of boost::match_extra
by Rick Weiss 08 Nov '06

08 Nov '06

I am trying to use the repeat capture mechanism in regex i thought i would start with the documentation example, but im having problems compiling it. Any help is appreciated. (There is likely a problem between my seat and the keyboard) Here is my compile output: and below that is the actual example. /usr/bin/g++ -Wno-ctor-dtor-privacy -O -o file2AssocTable.o -I/afs/apd.pok.ibm.com/func/vlsi/eclipz/common/integration/OA_migration/OpenAccess/2.2.4/include/oa \ -I/afs/apd.pok.ibm.com/func/vlsi/eclipz/common/integration/OA_migration/OpenAccess/2.2.4/include -I/afs/apd.pok.ibm.com/func/vlsi/eclipz/common/integration/GNU/boost/boost_1_33_1/include/boost-1_33_1 \ -c file2AssocTable.cpp file2AssocTable.cpp: In function `void print_captures(const std::string&, const std::string&)': file2AssocTable.cpp:171: no matching function for call to ` boost::match_results<__gnu_cxx::__normal_iterator<const char*, std::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<const char*, std::basic_string<char, std::char_traits<char>, std::allocator<char> > > > > >::captures(unsigned int&)' file2AssocTable.cpp:177: no matching function for call to ` boost::match_results<__gnu_cxx::__normal_iterator<const char*, std::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<boost::sub_match<__gnu_cxx::__normal_iterator<const char*, std::basic_string<char, std::char_traits<char>, std::allocator<char> > > > > >::captures(unsigned int&)' make: *** [file2AssocTable.o] Error 1 void print_captures(const std::string& regx, const std::string& text) { boost::regex e(regx); boost::smatch what; std::cout << "Expression: \"" << regx << "\"\n"; std::cout << "Text: \"" << text << "\"\n"; if(boost::regex_match(text, what, e, boost::match_extra)) { unsigned i, j; std::cout << "** Match found **\n Sub-Expressions:\n"; for(i = 0; i < what.size(); ++i) std::cout << " $" << i << " = \"" << what[i] << "\"\n"; std::cout << " Captures:\n"; for(i = 0; i < what.size(); ++i) { std::cout << " $" << i << " = {"; for(j = 0; j < what.captures(i).size(); ++j) { if(j) std::cout << ", "; else std::cout << " "; std::cout << "\"" << what.captures(i)[j] << "\""; } std::cout << " }\n"; } } else { std::cout << "** No Match found **\n"; } }

2 3

[lambda] if-then-else construct gets broken by lambda-lib
by Sebastian Weber 08 Nov '06

08 Nov '06

Hello everybody! I'm trying to store a reference to a function depending upon a condition within a function object like this: boost::function< double (std::size_t) > wf = (sIndex < m_maxIndex+1) ? ( boost::lambda::bind(boost::ref(m_cf), sDeg, boost::lambda::_1) ) : ( boost::ref(*m_wf) ); However, this does not compile. The error message is: error: no match for ternary 'operator?:' in '(sIndex < (((boost::cor_cm_iterator4<RandomGenerator, Graph>*)this)->boost::cor_cm_iterator4<RandomGenerator, Graph>::m_maxIndex + 1)) ? boost::lambda::bind(boost::ref(((boost::cor_cm_iterator4<RandomGenerator, Graph>*)this)->boost::cor_cm_iterator4<RandomGenerator, Graph>::m_cf), sDeg, boost::lambda::<unnamed>::_1) : boost::ref((*((boost::cor_cm_iterator4<RandomGenerator, Graph>*)this)->boost::cor_cm_iterator4<RandomGenerator, Graph>::m_wf))' But the lengthy code boost::function< double (std::size_t) > wf; if(sIndex < m_maxIndex+1) wf = boost::lambda::bind(boost::ref(m_cf), sDeg, boost::lambda::_1); else wf = boost::ref(*m_wf); compiles fine. Any idea? How can lambda eat up the ? : construct? How do I prevent that? My system is a ubunut edgy which means, I have g++ 4.1 and use boost 1.33.1. Greetings, Sebastian Weber

3 3

Dependencies between boost libraries
by Loïc Joly 08 Nov '06

08 Nov '06

Hello, I was curious to know wether there was somewhere available a graph of dependencies between boost libraries ? Best regards, -- Loïc

3 4