- Boost-users - lists.boost.org

[Multi-Index] crash on iterator copy or assignment
by Dominique Devienne 12 Mar '09

12 Mar '09

Hi, I've been using B.MI for a few days, and until now it's worked great for me. I've included below in Exhibit A a simplified (and a bit obfuscated, sorry) version of what I've been using (a proxy, indexed by name and type, and a method returning all instances of a given type). Filling in a list of all entries matching a given type (in exhibit A) works great. But now I was trying to return a custom (non-standard) iterator, where the client gets a pointer to a abstract base iterator (allows to hide the B.MI stuff in the concrete derived iterator), wrapped in an auto_ptr. The private concrete iterator internally stores the B.MI iterator from the index I'm interested in (exhibit B). This compiles (vs2005 sp1 on winXP 32), but it crashes at runtime on the call to boost::tie, during iterator assignment apparently. (call stack in exhibit C). So at this point I have no clue what's going on, and I would appreciate some help. The level of template complexity is more than stretching my abilities I'm afraid, just reading the type names makes my end spin ;) We're going to try it using GCC on Linux to see if it might be a compiler issue, and possibly try vs2008 as well, but if anyone knows if taking copies of the iterators is allowed (I know the index themselves shouldn't be copied, and we don't), and if yes why it's failing in this particular case, I'd appreciate. Thanks, --DD PS: Earlier I had a version of ByTypeFooIterator that took the begin/end iterator as param to its Ctor, taking a *copy*, but that also crashed w/o a stack trace this time. PPS: What I was trying to test in fact was whether I could get two equal ranges for two different MyTypeInfo, and iterate on those but interleaving the calls to ++iter, to verify nested or interleaved iteration was valid with B.MI. But I never reached that point. === Exhibit A: the code that works === struct FooProxy { Foo*const foo_; FooProxy(Foo* foo) : foo_(foo) {} const MyString& get_name() const { return foo_->name(); } const MyTypeInfo& get_type() const { return foo_->type_info(); } }; struct ByName {}; struct ByType {}; typedef boost::multi_index_container< FooProxy , indexed_by< hashed_non_unique< tag<ByName>, BOOST_MULTI_INDEX_CONST_MEM_FUN( FooProxy , const MyString&, get_name ) >, hashed_non_unique< tag<ByType>, BOOST_MULTI_INDEX_CONST_MEM_FUN( FooProxy , const MyTypeInfo&, get_type ) > > > FooIndex; FooIndex foo_index_; typedef FooIndex::index<ByName>::type FoosByName; typedef FooIndex::index<ByType>::type FoosByType; size_t select_foos(const MyTypeInfo& tinfo, MyList<Foo*>& list) { size_t count = 0; FoosByType::iterator range_begin, range_end; FoosByType& by_type = get<ByType>(foo_index_); boost::tie(range_begin, range_end) = by_type.equal_range(tinfo); for (FoosByType::iterator i = range_begin; i != range_end; ++i) { const FooProxy& proxy = *i; list.append(proxy.foo_); ++count; } return count; } === Exhibit B: the code that crashes === // public abstract base iterator class FooIterator { public: virtual ~FooIterator() {} virtual bool more() const = 0; virtual Foo* cur() const = 0; virtual void next() = 0; protected: FooIterator() {} }; // private concrete iterator struct ByTypeFooIterator : public FooIterator { FoosByType::iterator cur_, end_; virtual bool more() const { return cur_ != end_; } virtual Foo* cur() const { return cur_->foo_; } virtual void next() { ++cur_; } }; // public method std::auto_ptr<FooIterator> iterate(const MyTypeInfo& tinfo) { FoosByType& by_type = get<ByType>(foo_index_); std::auto_ptr<ByTypeFooIterator > iter_ptr = new ByTypeFooIterator; boost::tie(iter_ptr->cur_, iter_ptr->end_) = by_type.equal_range(tinfo); return iter_ptr; } // part of unit test Foo*const expected_bar = ...; // Bar derives from Foo for (std::auto_ptr<FooIterator> i = iterate(Bar::tinfo()); i->more(); i->next()) { Foo*const curr = i->cur(); CPPUNIT_ASSERT_EQUAL(expected_bar, curr); // only one Bar here } === Exhibit C: call stack of crash === (actual FooIndex contains more indices than shown in exhibit A above). > my_lib.dll!std::_Iterator_base::operator=(const std::_Iterator_base & _Right={...}) Line 374 + 0x8 bytes C++ my_lib.dll!std::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &>::operator=(const std::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::detail::iterator_base<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &>::operator=(const boost::detail::iterator_base<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &>::operator=(const boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::dereferenceable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> >::operator=(const boost::dereferenceable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> > & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::incrementable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::dereferenceable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> > >::operator=(const boost::incrementable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::dereferenceable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> > > & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::equality_comparable1<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::incrementable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::dereferenceable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> > > >::operator=(const boost::equality_comparable1<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::incrementable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::dereferenceable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<& __that={...}) + 0x18 bytes C++ my_lib.dll!boost::input_iteratable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> >::operator=(const boost::input_iteratable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> > & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::forward_iteratable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> >::operator=(const boost::forward_iteratable<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy const *,boost::iterator<std::forward_iterator_tag,FooProxy,int,FooProxy const *,FooProxy const &> > & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::forward_iterator_helper<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy,int,FooProxy const *,FooProxy const &>::operator=(const boost::forward_iterator_helper<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,FooProxy,int,FooProxy const *,FooProxy const &> & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >::operator=(const boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > > & __that={...}) + 0x18 bytes C++ my_lib.dll!boost::tuples::tuple<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > > &,boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > > &,boost::tuples::null_type,boost::tuples::null_type,boost::tuples::null_type,boost::tuples::null_type,boost::tuples::null_type,boost::tuples::null_type,boost::tuples::null_type,boost::tuples::null_type>::operator=<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > > >(const std::pair<boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > >,boost::multi_index::detail::hashed_index_iterator<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::hashed_index_node<boost::multi_index::detail::index_node_base<FooProxy,std::allocator<FooProxy> > > >,boost::multi_index::detail::bucket_array<std::allocator<FooProxy> > > > & k=({node=0x04680080 buckets=0x04667bac },{node=0x04667c68 buckets=0x04667bac })) Line 638 C++ my_lib.dll!iterate(const MyTypeInfo & tinfo={...}) Line 510 + 0x51 bytes C++

2 3

Build Errors with Exceptions Support Disabled
by Gabi Davar 12 Mar '09

12 Mar '09

Hi All, While compiling the signals library (Boost v1.38) to a platform with exception support disabled (ARM RVDS v3.0Sp1 [EDG]) I've encountered build errors in the function libraries due to usage of try/catch keywords. See attached file for a fix where the keywords were replaced by BOOST_TRY/CATCH macros. -gabi

1 0

tokenizer problem
by Medora Schauer 12 Mar '09

12 Mar '09

Hi All! I'm having a problem passing strings into a boost tokenizer. I have been able to get the code to work but I don't think I should have had to do what I did. I don't know if it is tokenizer problem or a compiler problem or my problem. Basically, I start with a char[] and depending on how I convert it to an std::string, I may or may not get garbage tokens. I've attached a complete, but small, program to demonstrate the problem. I hope someone can tell what is going on. I'm using a Fedora 8 system with g++ v 4.1.2. Thanks in advance! Medora #include <iostream> #include <string> #include <boost/tokenizer.hpp> using namespace std; char *in_strs[] = { "S,1001,101,39.6511915386591,-108.199469861370,1240151.6,127361.5", "S,1001,111,39.6511191997175,-108.199521016802,1240136.4,127335.6", "S,1001,121,39.6510468691093,-108.199571817158,1240121.3,127309.7", "S,1001,131,39.6509745384787,-108.199622617407,1240106.2,127283.8", "S,1001,141,39.6509021994691,-108.199673772516,1240091.0,127257.9", "S,1001,151,39.6508298687935,-108.19972457255,1240075.9,127232.0", "S,1001,161,39.6507575380953,-108.199775372477,1240060.8,127206.1", "S,1001,171,39.6506851990178,-108.199826527263,1240045.6,127180.2", "S,1001,181,39.6506125938775,-108.199877316164,1240030.5,127154.2", NULL }; int main(int argc, char* argv[]) { typedef boost::tokenizer<boost::char_separator<char> > str_tokenizer; boost::char_separator<char> sep(","); cout << endl; #if 1 // Here I'm just tokenizing the strings and writing them out. // This version works. char **str = in_strs; while (*str){ string file1_str = *str; str_tokenizer file1_tokens(file1_str, sep); for(str_tokenizer::iterator iter=file1_tokens.begin(); iter!=file1_tokens.end();++iter){ cout << *iter << " "; } ++str; cout << endl; } #endif #if 0 // This version results in garbage in some tokens. char **str = in_strs; while (*str){ str_tokenizer file1_tokens(string(*str), sep); // <-- difference for(str_tokenizer::iterator iter=file1_tokens.begin(); iter!=file1_tokens.end();++iter){ cout << *iter << " "; } ++str; cout << endl; } #endif //==================================================================================================== #if 0 // Now I want to skip the first 2 characters of the string. // This version works. string file1_str; char **str = in_strs; while (*str){ string file1_str = *str; string sub1 = file1_str.substr(2); // <-- difference str_tokenizer file1_tokens(sub1, sep); // <-- difference for(str_tokenizer::iterator iter=file1_tokens.begin(); iter!=file1_tokens.end();++iter){ cout << *iter << " "; } ++str; cout << endl; } #endif #if 0 // This version results in garbage in some tokens. string file1_str; char **str = in_strs; while (*str){ string file1_str = *str; str_tokenizer file1_tokens(file1_str.substr(2), sep); // <--- difference for(str_tokenizer::iterator iter=file1_tokens.begin(); iter!=file1_tokens.end();++iter){ cout << *iter << " "; } ++str; cout << endl; } #endif return(0); }

2 2

A problem while building 1.38.0 for msvc
by Younie, Bradford 12 Mar '09

12 Mar '09

I downloaded the 1.38 build zip file and unzipped it. I then downloaded the latest bjam and unzipped that. I then used bjam to build boost, as follows: bjam --build-dir=c:\users\myuser\boost-build --toolset=msvc stage It all built successfully, but when I referenced libboost-thread-vc90-mt-gd-1_38_0.lib, I got a linker error where it is looking for libboost-regex-vc80-mt-gd-1_38_0.lib. The problem is that it is requiring the vc80 lib, but bjam only built vc90 libraries. When I renamed the 90 regex library to say 80, I got a host of unresolved externals when linking. (I didn't really expect that to work, but I tried anyway). Can someone advise what I should do to fix this problem? Here are the stats about my system: Windows Vista Business 32-bit Visual C++ 6.0 installed Visual Studio 2003 installed Visual Studio 2005 installed Visual Studio 2008 installed (we develop an SDK, so I need to build in the various environments.) Brad ---------- Learn more about Chase Paymentech Solutions,LLC payment processing services at www.chasepaymentech.com. THIS MESSAGE IS CONFIDENTIAL. This e-mail message and any attachments are proprietary and confidential information intended only for the use of the recipient(s) named above. If you are not the intended recipient, you may not print, distribute, or copy this message or any attachments. If you have received this communication in error, please notify the sender by return e-mail and delete this message and any attachments from your computer.

2 1

boost asio: Probelm with closing a blocking ip::udp::socket receive
by Lennyk 12 Mar '09

12 Mar '09

Hi, I'm using boost 1.35 on Linux. I have a thread which initiates a ip::udp::socket and has a blocking receive. When I try to terminate this thread from a different thread, in order to close the udp::socket I use the following: // m_socket is of type ip::udp::socket boost::system::error_code ec; m_socket::shutdown(boost::asio::ip::udp::socket::shutdown_receive, ec); std::cout << "error code = " << ec << std::endl; m_socket.close(); After shutdown I ALWAYS get system:107 error code - which means "Transport endpoint is not connected". Does anybody have any idea why I'm receiving such an error? Needless to say, the receive is working great - and I get the packets from the port I'm listening to. Thanks!

2 3

[asio] 1.37.0: trouble with multiple timers and serial line port
by Markus Werle 12 Mar '09

12 Mar '09

Hi! Using boost-1.37.0 Having some trouble with multiple timers in one class. Looks like the timers are going berserk when MemFun1() calls Process(): The expiration wait times then are ignored and the whole loop seesm to run as if timers are all set to 0. Any idea what went wrong? A little code sketch (does not compile) class Communicator { // the main IO service that runs this connection boost::asio::io_service io_service_; // the serial port this instance is connected to boost::asio::serial_port SerialPort_; // maximum amount of data to read in one operation static const int max_read_length = 256; // data read from the socket char read_msg_[max_read_length]; std::deque<char> received_data_; // buffered write data std::deque<char> write_msgs_; boost::thread * pThreadRead_; // etc. // timers boost::asio::deadline_timer timer1_; boost::asio::deadline_timer timer2_; // etc. // missing: // the whole async_write/async_read_some to serial line stuff void Process() { // MISSING CODE: // ... fill the send queue of the serial line cache and fire // async_write on serial port ... // Q: does this probaly interfere StartMemfun(0); } void StartMemFun1(int counter) { timer1_.cancel(); // Q: Is this necessary here? timer1_.expires_from_now(boost::posix_time::milliseconds(1000)); timer1_.async_wait(bind(MemFun1, this, counter)); } void MemFun1(int counter) { if (!(counter < 5)) Process(); bool success; // ... process data from read cache if (success) StartMemfun2(0); else StartMemFun1(counter + 1); // this here works fine! } void StartMemFun2(int counter) { timer2_.cancel(); // Q: Is this necessary here? timer2_.expires_from_now(boost::posix_time::milliseconds(1000)); timer2_.async_wait(bind(MemFun2, this, counter)); } void MemFun2(int counter) { if (!(counter < 5)) Process(); bool success; // ... process data from read cache if (success) StartMemfun2(counter + 1); else StartMemFun1(counter + 1); } };

2 2

Testing algorithms don't use more than a concept provides
by athor 12 Mar '09

12 Mar '09

I'm trying to verify an algorithm doesn't use more than a certain concept provides. For example, if the algorithm checks for an ForwardIterator it must not use the i-- operator. I'm currently testing this by creating a minimal abstract class for each concept and then testing if the algorithm compiles using only the minimal as input. I have a few issues with this approach. 1. I have to maintain two definitions, one concept and one minimal implementation. 2. I have to write minimal implementations for all the boost/stl containers/concepts I want to use. I guess someone must have had the same problem before, is there a better way to do this? Maybe even some existing boost functionality I am unaware of? In code: ------------------------ #define CONCEPT_USER(function, call, targ1) \ struct ConceptUser_##function##_##targ1 \ { \ targ1& _1; \ void Check() \ { \ function##call; \ } \ }; struct Interface { virtual void Foo() const = 0; }; template<typename T> void f(int n, const T& t) { t.Foo(); } CONCEPT_USER(f, (47, _1), Interface); // Compiles template<typename T> void g(double x, const T& t) { t.Foo(); t.Bar(); } CONCEPT_USER(g, (3.14, _1), Interface); // Compile Error. g also uses the function Bar. ------------------------ -- View this message in context: http://www.nabble.com/Testing-algorithms-don%27t-use-more-than-a-concept-pr… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

[preprocessor] error: macro "BOOST_PP_SEQ_ELEM_III"
by Reimund Klemm 12 Mar '09

12 Mar '09

I'm having a problem with the preprocessor and currently I'm stuck. I tried to reduce the problem from my program as far as possible. Why does this example compile with an error?: #include <boost/preprocessor/seq/subseq.hpp> #include <boost/preprocessor/seq/seq.hpp> #include <boost/preprocessor/control/if.hpp> #define SEQ (b)(o)(o)(s)(t) #define i 0 int main( int argc, char* argv[]){ BOOST_PP_IF(i,BOOST_PP_SEQ_HEAD(BOOST_PP_SEQ_SUBSEQ(SEQ, 0, i)),0); return 0; } The error is the following: ./test.cxx:23:1: error: macro "BOOST_PP_SEQ_ELEM_III" requires 2 arguments, but only 1 given I'm using gcc 4.3.1 under Linux. As the condition is 0, the preprocessor shouldn't expand BOOST_PP_SEQ_HEAD(BOOST_PP_SEQ_SUBSEQ(SEQ, 0, i)). The preprocessed file (g++ -P -E) has expands to 0; but the preprocessor still throws an error. Could also someone give me a hint how to debug efficiently such a problem, as I'm a newbie to boost. --- Reimund

2 2

[multi-index] runtime indices
by Neil Hunt 12 Mar '09

12 Mar '09

Hi all, I have been using multi-index with great success, but have struck a requirement that sort of fits into the remit of multi-index, but isn't catered for. Here's what I want to do (pseudocode) typedef multi_index<MyClass> MyIndex; MyIndex mi; int index_id = 0; mi.add_sorted_index( index_id++, MyIndex::sorted_index_predicate_type( &some_predicate ) ); mi.add_sorted_index( index_id++, MyIndex::sorted_index_predicate_type( &some_other_predicate ) ); So basically, add indices at runtime, to take advantage of multi_index's great features, such as projection, index updating on modification, etc. My use case, if it's relevant, is to expose multi_index behaviour on objects from a dll. Where obviously the calling module can't specify at compile time the required indices. Is this a sensible addition to multi-index, or is it best implement using other mechanisms. (Having to re-implement many of the great features of multi-index doesn't seem desireable.) Neil

2 1

[bind][factory] Nested binds with functional/factory
by Rutger ter Borg 12 Mar '09

12 Mar '09

Dear all, I'm trying to use boost::bind in combination with boost/functional/factory.hpp (and value_factory) this works fine, boost::bind( boost::factory<T*>(), arg1, _1 ) with arg1 not being an lvalue (e.g., 2.0). however, with nested binds it is not accepted, boost::bind( boost::factory<T*>(), boost::bind( f, _1 ), _2 ) the compiler complains that it needs an lvalue for the nested bind. Is this desired behaviour? Is there some trick around this? Many thanks, Kind regards, Rutger ter Borg

1 0