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

shared_ptr in API
by Philippe Lambert 03 Aug '06

03 Aug '06

Hello, We are a large team of C++ developers, using Boost at large. We have a philosophical issue about shared_ptr usage and I'm looking for external inputs. The 2 clans are opposed about a simple question: "Should we use shared_ptr in the API function signatures of a module (dll or else)". Both clan use a lot of Sutter/Meyers/other gurus and authors citation to support their view. 1- Pros: shared_ptr are everywhere in the current code. It is the best way to prevent resources leaks, it is easy to use, convenient, it is mostly safe. A lot of gurus are saying "use shared_ptr". Preventing them in API will means a lot of .get() in the existing code. 2- Cons: A shared_ptr in the API give the right to the called function to keep a copy (virtually permanently) of the object. It is a bad indication in the contract if the function doesn't need to keep a copy. It tell to the user "take care, I may keep a copy of this and influence his life cycle". If it is not the case, you should tell it in the signature by asking for a const reference or something like it. Also, if the caller have a stack object and want to use an API function which ask for a shared_ptr, he will need to create a copy, or create a shared_ptr with a deactivated destructor... It is not a beautiful thing, both ways. Of course, I thing that the correct solution is in between, but how to state it? If it is a real issue, I would like to read other team opinion, standard, decision algo or policy... If we are technically totally disoriented, I would like to be redirected... Thanks to all Philippe Lambert

3 2

Re: [Boost-users] replace_all_regex on c-string help
by llwaeva＠21cn.com 03 Aug '06

03 Aug '06

> > > I got the string (char*s), which is too big (>10M), from an external > > library, it's too slow to convert the raw string to std::string before > > it is handled by replace_all_regex > > Here's your fundamental problem: regex does not do in-place search and > replace. It is boost::algorithm::replace_all_regex, not boost:regex_replace

3 2

regular expression on table data
by Winson Yung 03 Aug '06

03 Aug '06

Hello all, I have a follow table: OPERATING REVENUES: publishing $ 42,419 $ 44,754 $ 46,203 collegiate marketing and production services 97 ASSOCIATION MANAGEMENT SERVICES 16 wireless 8,883 8,129 7,507 51,302 52,883 53,823 All I know is that OPERATING REVENUES: will be always there, question is how to write a regular expression to capture the total (which is 51,302 here) There might be more/less than four rows in the table. Would really appreciated if anyone has good suggestion on this. /Winson

2 5

[date_time] Warning in date_time/constrained_value.hpp
by Vsevolod Vlaskin 03 Aug '06

03 Aug '06

Hi, We run into a strange compilation warning, which we see on and off in one particular spot in our code: std::string str( /* initialized with a sane value */ ); ... double dbl( boost::lexical_cast<double>( str ) ); return dbl; The warning shows up only in release mode, NOT in debug: /usr/include/boost/date_time/constrained_value.hpp: In function int main(int, char**): /usr/include/boost/date_time/constrained_value.hpp:64: warning: year$D120988$value_ is used uninitialized in this function If we comment definition of dbl, the warning disappears. It is not a big deal, but maybe, someone could shed light on it. Best regards, Vsevolod Vlaskine __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

Re: [Boost-users] (de)serializing map<const A*>
by Cheok Yan Cheng 03 Aug '06

03 Aug '06

Hi Robert, I am using VC++ 2003 I must say, the following workaround just work :) :) :) void load(Archive & ar, const unsigned int version) { void * xptr; xptr = & fiducialPairSearchRegionMap; map<FiducialPair *, vector<RectPair> > *mptr; mptr = static_cast<map<FiducialPair *, vector<RectPair> > >(xptr); ar >> *mptr; } Btw, may I know what is the reason behind of using static_cast instead of reinterpret_cast (reinterpert_cast can be compiled too)? Is it by following a thumb of rule, "use static_cast if you can and use reinterpret_cast if you have to"? Thank you very much! yccheok

2 1

boost::bind for returning a reference?
by Paul 03 Aug '06

03 Aug '06

Hi all, I understand why boost::bind() must return a const-reference or a value, but what I don't understand is, why must I do boost::bind<R&>(...) if I want it to return by reference? Why must I (the user) deduce the return type myself? Can't we have a boost::bind_return_reference(...) function that is the equivalent of boost::bind<R&>(...), so I don't have to do all the hard result_of type deduction manually? cheers Paul

2 3

Boost.Signals doc_view example application question?
by Allen Gooch 02 Aug '06

02 Aug '06

I have a quick question regarding connecting a bound member function. In the doc_view example application signals are defined as requiring a function taking a bool and returning nothing. class Document { public: typedef boost::signal< void (bool) > signal_t; // ... void append(const char* s) { m_text += s; m_sig(true); } // ... }; The View abstract base class requires concrete bases to provide a concrete implementation of the refresh() member function matching the slot signature required. class View { public: virtual void refresh(bool) const = 0; // ... }; Here's my problem. If I change the signal to take no args using the preferred or portable syntax I get compilation errors: class Document { public: typedef boost::signal< void () > signal_t; // ... void append(const char* s) { m_text += s; m_sig(true); } // ... }; class View { public: virtual void refresh() const = 0; // ... }; Compilation output (msvc-7.1): Compiling... SignalTest.cpp c:\Boost\include\boost-1_33_1\boost\bind.hpp(63) : error C2825: 'F::result_type': cannot form a qualified name c:\Boost\include\boost-1_33_1\boost\bind\bind_template.hpp(15) : see reference to class template instantiation 'boost::_bi::result_traits<R,F>' being compiled with [ R=boost::_bi::unspecified, F=void (__thiscall View::* )(void) const ] c:\development\SignalTest\SignalTest.cpp(49) : see reference to class template instantiation 'boost::_bi::bind_t<R,F,L>' being compiled with [ R=boost::_bi::unspecified, F=void (__thiscall View::* )(void) const, L=boost::_bi::list2<boost::_bi::list_av_2<View *,boost::arg<1>>::B1,boost::_bi::list_av_2<View *,boost::arg<1>>::B2> ] c:\Boost\include\boost-1_33_1\boost\bind.hpp(63) : error C2039: 'result_type' : is not a member of 'operator``global namespace''' c:\Boost\include\boost-1_33_1\boost\bind.hpp(63) : error C2146: syntax error : missing ';' before identifier 'type' c:\Boost\include\boost-1_33_1\boost\bind.hpp(63) : error C2955: 'boost::_bi::type' : use of class template requires template argument list c:\Boost\include\boost-1_33_1\boost\bind.hpp(112) : see declaration of 'boost::_bi::type' c:\Boost\include\boost-1_33_1\boost\bind.hpp(63) : fatal error C1903: unable to recover from previous error(s); stopping compilation Observed with boost_1_33_1 release on cygwin/gcc-3.4.4 and msvc-7.1platforms. Any ideas? Many thanks in advance, -allen

1 0

boost::filesystem::ifstream and .eof()
by pktao＠uwm.edu 02 Aug '06

02 Aug '06

Dear Boost users --I have a simple question about how to test for eof. I usually use: std::ifstream inputDataStream("path to file"); and test for EOF by: inputDataStream.eof() I tried the same thing using: boost::filesystem::basic_ifstream<char> inputDataStream("path to file", std::ios_base::in); and test for EOF by: inputDataStream.eof() but its not working. Am I using the .eof() correctly in the boost version of basic_ifstream? Here is the broken code using the boost version of basic_ifstream: int main() { boost::filesystem::path dataPath("/tao/data"); if ( !exists(dataPath) ) { std::cerr << "Path does not exist!"; return EXIT_FAILURE; }; for (boost::filesystem::directory_iterator fileIterator(dataPath); fileIterator != boost::filesystem::directory_iterator(); fileIterator++) { const boost::filesystem::path InputFileName = dataPath/fileIterator -> leaf(); boost::filesystem::basic_ifstream<char> inputDataStream(InputFileName, std::ios_base::in); int i = 1; int previousIterator = 0; trialOnsetWrite(); while ( !inputDataStream.eof() ) { if ( (previousIterator + 3) == i) { immobileWrite(i); previousIterator = previousIterator + 3; } else { dataStructureWrite(i); }; i++; }; trialOffsetWrite(i); }; return (0); }; Very appreciative, Patrick

2 1

Lack of release() in scoped_ptr and scoped_array
by Milutin Jovanovic 02 Aug '06

02 Aug '06

I finally decided to switch to boost from my own implementation of some helper classes, and I noticed that scoped_ptr and scoped_array in boost lack the release method(). I had to fall back to std::auto_ptr. However, this seems like such a basic and fundamental problem that I have started to fear that I am missing something obvious... The usage I got used to was something like this: { scoped_ptr<A> temp( new A() ); verify( temp.get() ); // might throw store_in_a_very_special_place( temp.get() ); // might throw; and I should keep ownership if thrown temp.release(); } I though this was good practice to make code exception safe. Sure I could do it with try-catch but that gets messy if I allocate multiple instances etc. So the question is: 1) is there something wrong with the concept of release()? 2) if release() is not evil, is there any other reason why is it not in boost::scoped_*? Regards, Milutin Jovanovic

3 2

boost::mpl::if_
by Noel Yap 02 Aug '06

02 Aug '06

I'd like to do something like: typedef typename boost::mpl::if_< typename base::is_container<T>::type, typename T::element_type, T>::type ElementType_; such that T may be a primitive type. The above doesn't work since, for example, there's no such thing as int::element_type. Is there a simple way (eg don't have to do template specialization for all container types) to do the above? Thanks, Noel

2 1