- Boost-users - lists.boost.org

[enable_if] Type choices
by Robert Dailey 25 Mar '08

25 Mar '08

Hi, I have a simple template function that takes one argument. For example: template< typename t_type > void foo( t_type param ) { } I want to restrict the type of t_type to a certain subset of types, such as 'char' or 'wchar_t'. If the user passes in any type other than those two types, the compiler should not be able to find an overload for foo() that matches the argument types. Note that I would also be using boost::enable_if to test this condition, however I do not see anything in type_traits or anywhere else to perform such a test. I thought of using mpl::vector to create a list of types the function would take and using enable_if to check if the type is in the container, but this seems like a lot of work. I want to see if there is a simpler, more compact solution. Thanks.

3 5

[python] Extending: Understanding ownership of pointers
by Robert Dailey 25 Mar '08

25 Mar '08

Hi, I had a bit of trouble deciding which mailing list to post this question (Python or boost). I decided to post it here since the way I'm extending is boost-specific, however the fundamental question is in regards to how Python internally handles ownership. I do not feel there is a perfect location for this question, so I hope you'll forgive the semi-off topicness of the question. Right now I'm exposing a function to Python that returns a pointer to a custom object owned by a third party library. In most cases I can manage to use boost::shared_ptr when I'm returning pointers, however since this is a third party library I cannot do this. Since the pointer I'm returning is owned and managed by another object, I gave my def() the following return policy: return_value_policy<reference_existing_object>() I simply assumed this was correct. How does Python handle the pointer to this object? Is there some documents in either boost or the python documentation that I can read to understand a little more about how Python handles object ownership when the Python objects really represent C++ pointers? Perhaps someone could give me a brief rundown on the concept. Thanks in advance.

2 3

[python] Py_Finalize(): To call, or not to call?
by Robert Dailey 25 Mar '08

25 Mar '08

Hi, I remember reading in the boost documentation a while back that Py_Finalize() should not be called when using Boost.Python to embed or extend python. I cannot find that documentation again, unfortunately. Is this a true statement or is it safe to call Py_Finalize()? If not, why can this not be called? How does Boost.Python handle finalizing the interpreter? Thanks.

2 3

serialization & registration questions
by Jerry 25 Mar '08

25 Mar '08

Is there a direct way to check if a class is registered correctly? And why would this differ when using binary/XML mediums? My binary serialization problem persists (sorry!) when using fast_binary_archive though the same works when using the XML serializer. Also I thought I'd try and see what happens when using an archive implemented from scratch according to the boost docs. Is the implementation section of the archive class reference up to date with 1.34.1 ? I was getting hard to decipher template error messages using VC++ 7.1 SP1. Thanks.

5 12

BGL: Copy constructor for Boost graphs
by Ethan Kim 25 Mar '08

25 Mar '08

Hi, I have a question on the copy constructor for boost graphs. When I make a copy G of a graph D as follows: Graph G(D); It makes a deep copy of D, correct? If so, is there a way to maintain a map from vertices in G to vertices in D, and vice versa? Or is this something I have to do manually with std::map? It seems that the copy constructor (as above) is quite time-consuming.. Thanks for any help in advance! -- Ethan DH Kim

1 0

tokenizing with xpressive
by Dilts, Daniel D. 24 Mar '08

24 Mar '08

I'm trying to tokenize lines of a file using the included static regex. I only care about the tokens indicated by s* = ... When I use the sregex_token_iterator to parse the lines, I only get the last match for s2 and s3. How should I change things so that I can get every match for s2 and s3 rather than just the last match? sregex whitespace_regex = *_s; sregex line_regex = whitespace_regex >> (s1 = +_d) >> whitespace_regex >> ( +( '\"' >> (s2 = *~as_xpr('\"')) >> '\"' >> whitespace_regex >> ':' >> whitespace_regex >> '\"' >> (s3 = *~as_xpr('\"')) >> '\"' >> whitespace_regex ) | +( '\"' >> (s2 = *~as_xpr('\"')) >> '\"' >> whitespace_regex ) ); This e-mail transmission contains information that is confidential and may be privileged. It is intended only for the addressee(s) named above. If you receive this e-mail in error, please do not read, copy or disseminate it in any manner. If you are not the intended recipient, any disclosure, copying, distribution or use of the contents of this information is prohibited. Please reply to the message immediately by informing the sender that the message was misdirected. After replying, please erase it from your computer system. Your assistance in correcting this error is appreciated.

2 1

pool allocator performance with multiple threads
by Michael Dickey 23 Mar '08

23 Mar '08

Does anyone know if the fast_pool_allocator is supposed to work in a multithreaded environment? I know that it is "thread-safe," but I'm getting some really discouraging test results from it. Namely, using fast_pool_allocator is consistently performing about 10 times SLOWER for me than new/delete whenever multiple threads are involved. I've tested and come up with the same results on three different platforms: Mac, Linux and Windows (MSVC). I get the impression that I just need to roll my own allocator if I'm using more than one thread, but I'm hoping that this is not the case. If it helps, I can put together and post some source code that demonstrates these results. Thanks, -Mike

2 1

[Filesystem] Unicode support
by David Bertrand 23 Mar '08

23 Mar '08

When compiling unicode filesystem functions, I get linking errors : Test error LNK2019: unresolved external symbol "class boost::filesystem::file_status __cdecl boost::filesystem::detail::status_api(class std::basic_string<unsigned short,struct std::char_traits<unsigned short>,class std::allocator<unsigned short> > const &,unsigned int &)" (...) I can reproduce it easily by compiling following line of code: fs::exists(L"test.gif"); The non-unicode variants links fine, it's really only the unicode stuff. Since "\libs\filesystem\test\wide_test.cpp" must be working, I assume I'm missing a compiler macro or something ? Anyone seen this already ? My setup : Windows XP SP2 MSVC 7.1 SP1 Boost 1.34.1 downloaded from boost-consulting.com This email and any attachments is confidential, may be legally privileged and is intended for the use of the addressee only. If you are not the intended recipient, please note that any use, disclosure, printing or copying of this email is strictly prohibited and may be unlawful. If received in error, please delete this email and any attachments and confirm this to the sender.

3 2

phoenix compile problems
by Olaf Peter 22 Mar '08

22 Mar '08

Hello, here my compile problem using phoenix (for use with spirit): ---8<--- template<typename ActionsT> struct tag_keyword_action { template <typename ScanT, typename Iter1T, typename Iter2T> struct result { typedef void type; }; explicit tag_keyword_action( const ActionsT& action ) : action( action ) { } template <typename ScanT, typename Iter1T, typename Iter2T> void operator()( const ScanT& scan, const Iter1T& bgn, const Iter2T& end ) const { action.on_tag( // the real action / assign class std::string( bgn, end ), scan.first.get_position().line ); } const ActionsT& action; }; template<typename ActionsT> struct ver_parser { typedef ::phoenix::function< tag_keyword_action<ActionsT> > keyword_tag_type; explicit ver_parser( const ActionsT& action ) : tag_keyword( action ) { } template <typename ScannerT> std::ptrdiff_t operator()(const ScannerT& scan, result_t& result) const { ... rule_t rule = lexeme_d[ ... ][tag_keyword( var( scan ), arg1, arg2 )] ... } const keyword_tag_type tag_keyword; }; template <typename ActionsT = parse_actions, typename ErrorHandlerT = parse_error_handler> struct grammar : sp::grammar< grammar<ActionsT, ErrorHandlerT> > { explicit grammar(const ActionsT& actions = ActionsT(), const ErrorHandlerT& handler = ErrorHandlerT()) : actions( actions ), error_handler( handler ) { } template <typename ScannerT> struct definition { ... explicit definition( const grammar& self ) : ver_p( self.actions ) { ... } ... const sp::functor_parser< ver_parser<ActionsT> > ver_p; }; ... }; --->8--- The (beautified) error message is: error: no matching function for call to 'boost::spirit::functor_parser<ver_parser<parse_actions> >::functor_parser(const parse_actions&)' candidates are: boost::spirit::functor_parser<FunctorT>::functor_parser(const FunctorT&) [with FunctorT = ver_parser<parse_actions>] boost::spirit::functor_parser<FunctorT>::functor_parser() [with FunctorT = ver_parser<parse_actions>] boost::spirit::functor_parser<ver_parser<parse_actions> >::functor_parser(const boost::spirit::functor_parser<ver_parser<parse_actions> >&) What is the correct syntax, is this way a good choice? I want to tag the keyword in the given data files with their lines to display them in a browser later. Thanks, Olaf

1 1

[lambda] Confusion!
by Robert Dailey 22 Mar '08

22 Mar '08

Hi, I have defined a global overloaded operator<<. I want to use this overloaded operator in std::for_each(), but it must be properly deduced since there are many global overloads for this operator. For example: class Stream { // This is an abstract class (cannot instantiate) }; Stream& operator<< ( Stream& stream, char data ) { stream.write( data ); return stream; } Stream& operator<< ( Stream& stream, std::vector<char>& buffer ) { using namespace boost::lambda; std::for_each( buffer.begin(), buffer.end(), var( stream ) << _1 ); // This lambda should simply be binding to the overloaded operator<< above, right? return stream; } When I compile this, MSVC9 tells me: 1>c:\program files\microsoft visual studio 9.0\vc\include\algorithm(29) : error C2064: term does not evaluate to a function taking 1 arguments 1> class does not define an 'operator()' or a user defined conversion operator to a pointer-to-function or reference-to-function that takes appropriate number of arguments What am I doing wrong? Do I need to be using boost bind instead? If so, how can I use boost.bind to map a global operator? Will it correctly deduce which overload to use?

2 5