- Boost-users - lists.boost.org

[addressof] Can't obtain the address of an instance of an anonymous struct
by Jean-Francois Bastien 20 Mar '08

20 Mar '08

We're running into an issue with addressof as described at: https://savannah.cern.ch/bugs/?33071 Basically this fails under GCC 3.4.4 as well as GCC 4.3: struct // NoName { } noName; void * fail() { return addressof(noName); } The error message is "no matching function for call to `addressof(<anonymous struct>&)'". MSVC 2005 SP1 compiles and runs this with no problem. The problem seems to be that GCC can't deduce the templated type because the struct is anonymous. I'm not sure whether this is because it's non-standard behavior that MSVC accepts and GCC doesn't or because GCC has a bug. Either way it would be nice if addressof had a workaround for the "broken" GCC versions, but I'm not sure what the best workaround would be. Axel suggested in the above link to C-style cast to char &, but this causes a problem in the (unlikely) event where operator char &() is declared in the anonymous struct. We could use template magic to see if operator char &() is declared, but I'm not very comfortable with such a huge workaround. If the discussion points towards a bug in GCC I'll file a bug report with them, but either way addressof isn't working with GCC 3.4.4 and 4.3 (at least), and it would be nice of Boost to offer a workaround.

1 0

Re: [Boost-users] Boost-users Digest, Vol 1577, Issue 2
by Abir Basak 20 Mar '08

20 Mar '08

> > I don't have access to that compiler. I tried running similar code in > Visual C++ 6.5 and Visual C++ express 8, and they both worked. I can > try adding something to the unit tests but that will take me a while > to work through. At the end of this email is the code I tested with, > can you try it on your compiler? If it works, I'm doing something > different to you, do you have any idea what? > > thanks, > > Daniel > > #include <boost/unordered_map.hpp> > #include <boost/tuple/tuple.hpp> > #include <iostream> > > int main() { > typedef boost::unordered_multimap<char, int> Mymap; > Mymap c1; > > c1.insert(Mymap::value_type('a', 1)); > c1.insert(Mymap::value_type('a', 2)); > c1.insert(Mymap::value_type('b', 3)); > c1.insert(Mymap::value_type('c', 4)); > > Mymap::iterator it,end; > boost::tie(it,end) = c1.equal_range('a'); > for( ; it!=end; ++it) { > std::cout<<(*it).second<<" "; > } > } Yes, MSVC 9 (i.e 2008) runs it properly, but MSVC 7.1 (2003 . NET) fails to compile it. It looks like, for some reason node_ = data::next_group(node_); from increment_group() tries to construct data, which is a typedef to BOOST_UNORDERED_TABLE_DATA even while using static function for that. just changing the statement to node_ = next_group(node_); (hash_table_impl.hpp cvs line no 301) works. so it seems, msvc 7. either doesn't support calling static function by class name typedef, or the data typedef is confused with some other (like BOOST_UNORDERED_TABLE) note node_ = BOOST_UNORDERED_TABLE_DATA::next_group(node_); also works fine. i tried to produce some simple test case for this, like class outer{ public: typedef outer outer_t; inline static void func(){} class inner{ void func1(){ outer_t::func();// func() this works;or outer::func() works } }; }; this doesn't work in msvc 7.1 but works in 9.0 . i am not sure if it is valid in standard though. The error is, E:\lib\boost\trunk\boost\unordered\detail\hash_table_impl.hpp(301): error C2039: 'hash_table_data_equivalent_keys<std::allocator<std::pair<char const ,int> > >' : is not a member of 'boost::unordered_detail::hash_table_data_equivalent_keys<Alloc>::iterator_base' with [ Alloc=boost::unordered_detail::hash_types_equivalent_keys<std::pair<const char,int>,char,boost::hash<char>,std::equal_to<char>,std::allocator<std::pair<const char,int>>>::value_allocator ] thanks abir

2 1

Re: [Boost-users] unordered_set insert returns const_iterator ?
by Abir Basak 20 Mar '08

20 Mar '08

> > I don't have access to that compiler. I tried running similar code in > Visual C++ 6.5 and Visual C++ express 8, and they both worked. I can > try adding something to the unit tests but that will take me a while > to work through. At the end of this email is the code I tested with, > can you try it on your compiler? If it works, I'm doing something > different to you, do you have any idea what? > > thanks, > > Daniel > > #include <boost/unordered_map.hpp> > #include <boost/tuple/tuple.hpp> > #include <iostream> > > int main() { > typedef boost::unordered_multimap<char, int> Mymap; > Mymap c1; > > c1.insert(Mymap::value_type('a', 1)); > c1.insert(Mymap::value_type('a', 2)); > c1.insert(Mymap::value_type('b', 3)); > c1.insert(Mymap::value_type('c', 4)); > > Mymap::iterator it,end; > boost::tie(it,end) = c1.equal_range('a'); > for( ; it!=end; ++it) { > std::cout<<(*it).second<<" "; > } > } Yes, MSVC 9 (i.e 2008) runs it properly, but MSVC 7.1 (2003 . NET) fails to compile it. It looks like, for some reason node_ = data::next_group(node_); from increment_group() tries to construct data, which is a typedef to BOOST_UNORDERED_TABLE_DATA even while using static function for that. just changing the statement to node_ = next_group(node_); (hash_table_impl.hpp cvs line no 301) works. so it seems, msvc 7. either doesn't support calling static function by class name typedef, or the data typedef is confused with some other (like BOOST_UNORDERED_TABLE) note node_ = BOOST_UNORDERED_TABLE_DATA::next_group(node_); also works fine. i tried to produce some simple test case for this, like class outer{ public: typedef outer outer_t; inline static void func(){} class inner{ void func1(){ outer_t::func();// func() this works;or outer::func() works } }; }; this doesn't work in msvc 7.1 but works in 9.0 . i am not sure if it is valid in standard though. The error is, E:\lib\boost\trunk\boost\unordered\detail\hash_table_impl.hpp(301): error C2039: 'hash_table_data_equivalent_keys<std::allocator<std::pair<char const ,int> > >' : is not a member of 'boost::unordered_detail::hash_table_data_equivalent_keys<Alloc>::iterator_base' with [ Alloc=boost::unordered_detail::hash_types_equivalent_keys<std::pair<const char,int>,char,boost::hash<char>,std::equal_to<char>,std::allocator<std::pair<const char,int>>>::value_allocator ] thanks (sorry for double posting ... last message subject line was wrong) abir

1 0

unordered_set insert returns const_iterator ?
by Abir Basak 20 Mar '08

20 Mar '08

Hi while i am using boost unordered_set , and insert a new key value pair, it returns the iterator as a const one. In fact when i looked at the code, the iterator is a typedef to implementation::const_iterator ... so there is no way to gen an mutable iterator from the set, including begin() and end() is it a standard behavior or i am missing something ? thanks abir

3 3

ignoring function arguments with Boost.Bind?
by Christoph Duelli 20 Mar '08

20 Mar '08

Please consider the following scenario: class Caller { public: Caller(...); typedef boost::function<void (int)> MyCallback; MyCallback callback; }; class Callee { public: Callee(); void do_it() {...}; }; Caller caller; Callee callee; // does not work as do_it does not match MyCallback caller.callback = boost::bind(&Callee::do_it, &callee, ...); The function do_it() is not interested in the arguments passed to the callback. I obviously could make the function take these arguments and just ignore them... but I was wondering: is it possible to leave do_it() as it is and work some magic with Boost.Bind that 'drops' the unwanted argument? Basically create an anonymous function object that calls callee without the argument it gets passed itself. (Similarly to Qt signals/slots: the slot may have fewer arguments than the signal.) Thank you and best regards Christoph

2 1

Re: [Boost-users] boost serialization & module boundaries
by Robert Ramey 20 Mar '08

20 Mar '08

I'm not sure from the title of the post what the question is. Never-the-less, I'll provide an answer. Up until now, there have problems when serialization code was spread accross and/or repeated in distinct runtime modules. Also, there was a problem when serialization was invoked simultaneously from different threads. Both these problems are related to the fact that the library has to keep a table of which types have the serialization code loaded. This has now been addressed in the recently checked in version in the trunk. Serialization code can be in dlls. Also code can be repeated accross DLLs. Personally, I would prefer to organize my code so that this doesn't happen but apparently some people found this inconvenient. I have tested this by hand with my VC 7.1 system. The tests consist of making separate DLLS and running the executable all from the same directory. I have been unable to figure out how to do such a thing with bjam so as far as linux is concerned its untested. Robert Ramey

1 0

Q and Embedded Systems
by Thanh Tran 19 Mar '08

19 Mar '08

I'm new to boost and would like to know if boost can be used in embedded systems. Obviously, it'll work if the system is using derivative of Linux. What if I don't have a real OS running on the system. Can I still simply fold the source code into my project and build it? I'm sure it can be done. I'm not sure how difficult it will be. Thanks. ____________________________________________________________________________________ Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

2 1

boost serialization & module boundaries
by Christian Widmer 19 Mar '08

19 Mar '08

Hi, I'm trying to use boost to serialize a machine learning library that I'm working with. It all worked out very conveniently, but it seems now, that I have hit a dead end. I've been looking on the net but couldn't find any information that worked for me, so here we go: The problem is, that the library consists of several modules, which are dynamically linked. For instance the class KernelMachine has a field kernel of type Kernel, which is contained in a separate module (in the sense of a shared object). The following code compiles, but when I try to serialize a CKernelMachine object I get an "unregistered_class" error. (CKernel does serialize fine, when called directly) class CKernelMachine : public CClassifier { private: friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & boost::serialization::base_object<CClassifier>(*this); ar.register_type(static_cast<CGaussianKernel *>(NULL)); ar & kernel; std::cout << "done CKernelMachine" << std::endl; } protected: CKernel* kernel; }; I found some threads that pointed me towards BOOST_CLASS_EXPORT, but I'm not exactly sure how to use it. So my question is - is it in principle even possible to serialize across module (shared-object) boundaries. If so, how should I go about? Any pointers would be greatly appreciated. - Chris

3 7

Boost Regex - Specify multiple expressions and formats with regex_replace?
by Jeff 19 Mar '08

19 Mar '08

Dear Regex Users: I'd like to know if it is possible to check multiple expressions and specify multiple patters within a single regex_replace statement. For example, I want to replace .html extensions with .php and replace .gif with the .srf extension Match: http://localhost/index.gif replace with: http://localhost/index.srf Match: http://localhost/index.html replace with: http://localhost/index.php Can this be done within a single regex_replace statement? Although this does not work, it demonstrates what I am trying to achieve: // trying to match either condition const boost::regex sExp("(?:^(.*/)index\.html$) | (?:^(.*/)index\.gif$)"); // trying to replace according to matched extension const std::string sRewrite("$1index.php"); std::string rewrite(const std::string& s) { return boost::regex_replace(s, sExp, sRewrite, boost::match_perl | boost::format_perl | boost::format_no_copy | boost::format_all); } int main() { string s[5] = { "http://localhost/abc.png", "http://localhost/abc.jpb", "http://localhost/index.gif", "http://localhost/hambone.srf", "http://localhost/index.html" }; for(int i = 0; i < 5; ++i) { if (rewrite(s[i])!="") { cout << rewrite(s[i]) << endl; } } return 0; } I don't know if regex_replace can be used to do this or not. Thank you, Jeff

2 1

Boost.Interprocess Documentation
by Alexander Dong Back Kim 19 Mar '08

19 Mar '08

Hi all, Is there any document or resource that I can learn about Boost.Interprocess? I'm currently being interesting with Shared Memory stuff. Any suggestions or hints are welcomed. -- Alexander Dong Back Kim Australia +61 433 469 100 / Korea +82 10 6357 8840

2 1