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

[polygon] Compiling polygon under Windows
by Josh Pieper 29 Nov '10

29 Nov '10

Hello, I have been compiling boost::polygon under Windows (VS2008, MSVC9.0 specifically) and have run into a minor problem with boost::polygon's detection of Microsoft's compiler. Specifically, it tests for the WIN32 preprocessor define instead of _WIN32. WIN32 isn't actually defined by the compiler, but by windows.h, and not all translation units actually need to include that header. _WIN32 is defined by the compiler, and boost has the similar BOOST_MSVC preprocessor define which could also be used. While there is some disagreement among the existing boost libraries about the best mechanism to check this, some check _WIN32, some check BOOST_MSVC, and some check both or others at the same time. I propose the following patch to the 1.45.0 boost::polygon sources to make the detection of the MSVC compiler more robust. Regards, Josh Pieper Index: boost/polygon/gtl.hpp =================================================================== --- boost/polygon/gtl.hpp +++ boost/polygon/gtl.hpp @@ -12,7 +12,7 @@ #pragma warning (disable:1125) #endif -#ifdef WIN32 +#ifdef _WIN32 #pragma warning( disable: 4996 ) #pragma warning( disable: 4800 ) #endif Index: polygon/isotropy.hpp =================================================================== --- boost/polygon/isotropy.hpp +++ boost/polygon/isotropy.hpp @@ -48,7 +48,7 @@ #include <boost/mpl/or.hpp> #else -#ifdef WIN32 +#ifdef _WIN32 #define BOOST_POLYGON_MSVC #endif #ifdef __ICC @@ -290,7 +290,7 @@ template <typename T> struct gtl_if { -#ifdef WIN32 +#ifdef _WIN32 typedef gtl_no type; #endif };

4 4

[interprocess] file_mapping/mapped_region on empty file throws?
by Jeff Flinn 29 Nov '10

29 Nov '10

The following appears to throw an interprocess_exception with boost_1_44_0 compiled with MSVC8 running on windows XP. using namespace boost::interprocess; file_mapping fm("empty.txt", read_only); mapped_region mr( fm, read_only); Is there a reason I can't just get back a null address and zero size? I've wrapped this to get back a begin/end range and an empty range is a valid condition. Thanks, Jeff

2 2

[Units] Defining units with conversions local to class instances
by Andrew Hundt 29 Nov '10

29 Nov '10

I'm looking to define local units specific to each instance of a class. In this specific case I need to define a unit called "ticks" that is an angle defined relative to degree_base_unit and radian_base_unit. Local unit definitions are required because I am dealing with varying physical motors that measure their position in "ticks". The number of ticks per radian is arbitrarily defined by the manufacturer, so each class needs an independent "ticks" definition for that specific device loaded by a config file at runtime (so the user is not required to recompile every time they use a new motor). Some of the pieces I need look to be there within a combination of the examples: http://www.boost.org/doc/libs/1_45_0/libs/units/example/runtime_conversion_… - Defines a runtime conversion for global units defined to exist at compile time <http://www.boost.org/doc/libs/1_45_0/libs/units/example/runtime_conversion_…> http://www.boost.org/doc/libs/1_45_0/libs/units/example/radar_beam_height.c… - Defines a global unit defined at compile time <http://www.boost.org/doc/libs/1_45_0/libs/units/example/radar_beam_height.c…> http://www.boost.org/doc/libs/1_45_0/libs/units/example/runtime_unit.cpp - Only works for units defined at compile time The key problem here is that I need to define my units locally since they are irrelevant anywhere else while retaining the conversion and type safety of boost.units. It looks like the macros BOOST_UNITS_DEFINE_CONVERSION_FACTOR and BOOST_UNITS_DEFINE_CONVERSION_FACTOR_TEMPLATE both require definition in the global namespace. Any thoughts would be greatly appreciated. Cheers! Andrew Hundt

1 0

[Interprocess] Minimum allocation size for managed_shared_memorty
by Aditya Gadre 29 Nov '10

29 Nov '10

I have a question about Boost.Interprocess library. I am sharing a fixed size boost::interprocess::vector using managed_shared_memory. The vector length is 100 and the type of element in the vector is boost::array<float, 2>. That makes (vector::size)*sizeof(boost::array<float, 2> = 100*8 = 800. . If I create a managed_shared_memory segment with length 800 bytes and create a vector in that segment, then boost::interprocess::bad_alloc is thrown. If however I create a segment with a length of say 8000 and then create the vector, then everything works well. So my question is what is the minimum size required for a managed_shared_memory segment apart from what the size of a vector? Thanks, - Aditya

2 3

[Boost-users] [message_queue.hpp]boost::interprocess_exception::library_error
by rohini chandra pallapothu 29 Nov '10

29 Nov '10

Hi, I am trying to use boost message queues for interprocess communincation between a command and a server. I had been receiving the below exception in the server. boost::interprocess_exception::library_error Below is the server code I had problems with. The exception was consistent when the maximum message size argument passed to the constructor while creating a message queue was greater than the size of the data I was trying to receive.I am trying to receive an integer. When I assigned the max_msg_size variable a size greater than the size of int, it threw an exception. When I assigned the size of int, it is working fine. But I want to understand the correct reason behind the exception. Is it not allowed to receive a message which is less than the maximum message size because the name maximum message size itself says that it should only not exceed. I am working on a 64 bit SUSE 11. I had even tried to change the /proc/sys/kernel/shmmax value after reading one of the discussion on the same exception in the BOOST archives which did not work. Could anyone please help. Please let me know if you need more information. #include <boost/interprocess/sync/named_mutex.hpp> #include <boost/interprocess/ipc/message_queue.hpp> #include <iostream> using namespace std; using namespace boost::interprocess; int main(){ cout<<"in main"<<endl; cout<<"size of int is "<<sizeof(int)<<endl; //4 on my platform unsigned int priority; std::size_t recvd_size; std::size_t max_msg_size=sizeof(int); //not working when assigned a larger value than I was sending. std::size_t max_no_msgs=10; int i; message_queue::remove("TestQueue"); cout<<"message queue removed if any"<<endl; try{ //create a message queue. message_queue mq(create_only,"TestQueue",max_no_msgs,max_msg_size); cout<<"message queue created"<<endl; mq.receive(&i,sizeof(i),recvd_size,priority); } catch(interprocess_exception &ex){ std::cout <<ex.what() << std::endl; return 1; } cout<<"message received"<<endl; cout<<"message received is"<<i<<endl; return 0; } Thanks in advance. Best Regards, Rohini Chandra

1 0

Usage of Boost.Optional
by John Dlugosz 29 Nov '10

29 Nov '10

I have a situation where 'optional' is a good fit. However, I can't quite figure out how to use it for what I want. It appears to be geared around passing and returning instances, but I want to use it as a member. Consider: class C { D* d; // ... real code has much more stuff, naturally public: D(params) : d(0) { // ... if (I_ended_up_needing_d) d= new D (blahblah); // ... } ~D() { delete d; // harmless if zero } }; There is a single ownership of d by the instance, and it is not polymorphic but an exact type known at compile time. Making it a pointer adds overhead of using the heap, makes it harder to write EH-safe code for the constructor in more realistic situations, and requires adding explicit code to the destructor. Using a smart pointer only adds overhead, might take care of the EH and housekeeping, but still uses the heap. The 'optional' class provides for in-place storage, which it can use for an initialized instance, or it could be nil. That's exactly what I want. But, how do I tell optional<D> to construct at some point within the 'if' statement of my code? Assigning requires duplicating the object, and this type is not meant to be copied and has copy constructing and assignment disabled. Based on how the class works, I should be able to say "turn it on, now" and use an in-place factory. But the function to do that is not present in the class. class C { optional<D> d; public: D(params) : d(boost::none_t) { if (I_ended_up_needing_d) d.reset (in_place(blahblah)); } }; Am I missing something, or is the class missing something? I'm using Boost 1.43 in the project, and basing the above on the documentation for that version. The 'news' for 1.44 does not list Optional as having changed. Thanks, --John (sorry about the corporate footer) TradeStation Group, Inc. is a publicly-traded holding company (NASDAQ GS: TRAD) of three operating subsidiaries, TradeStation Securities, Inc. (Member NYSE, FINRA, SIPC and NFA), TradeStation Technologies, Inc., a trading software and subscription company, and TradeStation Europe Limited, a United Kingdom, FSA-authorized introducing brokerage firm. None of these companies provides trading or investment advice, recommendations or endorsements of any kind. The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and delete the material from any computer.

6 11

[BGL] BFS < given distance.
by Adam Spargo 29 Nov '10

29 Nov '10

Hi, I have finally managed to implement my genome assembly algorithm! Except one part is too slow. I'm doing a breadth first search starting from each vertex i in an undirected graph, looking for its mate-pair (another given vertex j, of approximately known distance, where distance is the same as summed edge weight). Ideally I want to follow the BFS until either (i) I find the mate-pair vertex j (ii) I have visited all vertices up to some given distance away from the source vertex i. I want to bail out of the algorithm at the first occurrance of either (i) or (ii) and not carry on to search through the rest of the graph, which is very large. So I'm wondering if anybody has had a similar requirement and if they implemented a BFS visitor which accomplishes it? I have managed to limit it to the same connected component, but it still takes too long. The limiting distance is small and so I could just write something, but it would obviously be more elegant to use a BFS visitor. Thanks for any advice, Adam. -- Dr Adam Spargo High Performance Assembly Group email: aws(a)sanger.ac.uk Wellcome Trust Sanger Institute Tel: +44 (0)1223 834244 x7728 Hinxton, Cambridge CB10 1SA Fax: +44 (0)1223 494919 -- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.

2 4

thread group
by Lloyd Kanakkassery 29 Nov '10

29 Nov '10

Hi, I am creating "n" number of thread in a loop like this. for(int i=0;i<n;++i) { boost::shared_ptr<boost::thread> th(new boost::thread(boost::bind(&Enum::Execute,shared_from_this(),i,x))); } How can I join all these threads? The approach I used is given below... Before the loop I have created a thread_group object boost::thread_group tg; for(int i=0;i<n;++i) { boost::shared_ptr<boost::thread> th(new boost::thread(boost::bind(&Enum::Execute,shared_from_this(),i,x))); tg.add_thread(th.get()); } tg.join_all(); But the threads are not joined, it is exiting without completion. What could be the mistake I am doing? Thanks, Lloyd

3 2

[thread][repost] recursive shared mutex
by JOAQUIN M. LOPEZ MUÑOZ 29 Nov '10

29 Nov '10

[Trying my luck again with this query.] Hi, Boost.Threads provides a recursive mutex and a shared mutex, but no recursive *and* shared mutex. Would it be a good idea to add this to the library? How hard is it to construct a recursive shared mutex from boost::sahred_mutex (i.e. adding reentrancy to boost::shared_mutex)? Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

5 7

Visiting an element of fusion sequence, selected at run-time
by Max Motovilov 29 Nov '10

29 Nov '10

Here's a seemingly simple problem that I have spent most of my Thanksgiving weekend on, and still don't have enough to boast: Given: 1. A Boost.Fusion-compatible sequence; 2. An integral value i computed at runtime, known to be within 0..size(sequence); implement an algorithm applying some sort of a generic visitor to the i-th element of the sequence. However long I looked through the documentation, I didn't find a ready solution. Here's the best I was able to come up with on my own: ====================================================================== struct at_v { template<typename Seq, typename N> struct result : public fusion::result_of::at<Seq,N> {}; template<typename Seq, typename N> typename result<Seq,N>::type operator()( Seq& seq, N ) const { return fusion::at<N>( seq ); } }; template<typename F, typename Seq, typename IntV> void apply_at( F f, Seq& seq, IntV i ) { fusion::for_each( fusion::as_vector( mpl::range_c< IntV, 0, IntV( fusion::result_of::size<Seq>::type::value ) >() ), phoenix::if_( phoenix::arg_names::arg1 == i ) [ phoenix::function<F>( f )( phoenix::function<at_v>()( phoenix::ref( seq ), phoenix::arg_names::arg1 ) ) ] ); } ====================================================================== Drawbacks: 1. Requires a Boost.Phoenix-compliant visitor -- even though I managed to hide the Phoenix dependency inside apply_at(), it still requires a full-blown templatized result metafunction as part of the visitor even though the latter is not expected to return a value. Be nice to have simpler requirements for a visitor. 2. Can't figure out an easy way of returning a value from the visitor. I tried some constructs using fusion::fold() but only halfheartedly: the big problem is that there is no obvious way to specify that the visitor always returns results of the same type, no matter what argument type it is given. Note that the function calls the visitor only once, so, in principle, there should always be exactly one value to return (index out of bounds should naturally be a contract violation, which is not checked in my version). 3. The complexity is O( size(sequence) ) which is quite atrocious for random access (even though in practice it should not be much of a problem given that it simply means N integer comparisons with a relatively small N). Still, a O( log N ) or even a constant-time implementation should be quite possible... Even though the solution above should be sufficient for my purposes I'm still openly wondering whether I have overlooked something in the Boost libraries and/or whether someone more versed in Fusion/Phoenix/MPL might suggest a better approach. Thanks in advance for any responses, ...Max... PS. Without fusion::as_vector() around the index sequence constructor, GCC 4.4.3 complained of not being able to match operator(), suggestive of a forwarding function problem: looks like operator* on an iterator over mpl::range_c<> returns something that can't be digested by a Phoenix actor. I wonder if this should really be taken care of by the Fusion adapter for MPL sequences? PPS. Is there better documentation for Boost.Phoenix than what's at http://www.boost.org/doc/libs/1_45_0/libs/spirit/phoenix ? Not even a straightforward reference manual?

2 2