- Boost-users - lists.boost.org

Looking for tangent of an arbitrary 2d curve.
by Atanassov Tervel-PT1672 17 Sep '08

17 Sep '08

Hello, I am looking for code which will approximate the tangent at a point of an arbitrary 2D curve. Does boost have such functionality? Thanks, Tervel Atanassov

1 0

[fusion] How to transform types of fusion::map ?
by Zeljko Vrba 17 Sep '08

17 Sep '08

A very concrete question: I have map<pair<K1, V1>, pair<K2, V2>, ..> as input. How do I produce map<pair<K1, std::vector<V1> >, pair<K2, std::vector<V2>, .. > on output? I do not see an equivalent of LISPs MAPCAR in the list of algorithms.. Is this something that should be done by combining fold and push_back algorithms? If so, I'm currently in the dark on how to do it (no example for the fold metafunction algorithm) === Few more questions now that I'm already posting: - what is the difference between accumulate and fold? - what is the purpose of for_each metafunction when the docs say that it always returns void? (and no side-effects are possible at compile-time) Thanks.

4 4

Problem with iterator_facade
by Robert Dailey 17 Sep '08

17 Sep '08

Hi, I have the following iterator_facade class: class PassIterator : public boost::iterator_facade< PassIterator, Pass, boost::forward_traversal_tag > { public: PassIterator(); PassIterator( Effect const& effect ); private: void increment(); Pass& dereference() const; Pass m_pass; friend class boost::iterator_core_access; }; Note how my Pass object is stored by-value in my PassIterator class. When I attempt to compile this class, it fails because dereference() is const and it is trying to return a reference to a const member. However, if I make dereference() non-const, it still fails because the base class is calling into PassIterator::dereference() through a const member function as well, so dereference() *must* be const in order to be callable by boost. How can I overcome this issue? const_cast? mutable? Thanks.

3 5

[EnableIf?] Seeking an example....
by Robert Jones 17 Sep '08

17 Sep '08

Somewhere in the Boost headers, or maybe the Boost documentation I have seen a bit of code which uses SFINAE, and also possibly the enable_if header to select either a find class method or find global template depending on what's available. It might have been written by Eric. I can't find it now,maybe it's not there any more, but if anyone can point me to it I'd really appreciate it! Thanks in advance, - Rob.

3 2

Tuples: Can someone help me generalize this?
by Derrick Hathaway 17 Sep '08

17 Sep '08

Hello, boost-users, what I am trying to accomplish is a generic collection of N objects of type Type. I need for the N objects to be tightly packed and I would like the collection to be constructible in this way: uniform_vector<3, float> my_point(1.0f, 2.1f, 3.2f); I'm very new to the template (meta)programming scene and I would like to get better at it, so any advice you could give that would be instructive in that regard would be much appreciated. The code below is what I have come up with. Question 1: Am I reinventing the wheel at all? In other words is there something I can just typedef to get this functionality? Question 2: If not, how could I generalize this? Is there a template metaprogramming way? I'm sure it can be done with the preprocessor but I'm no good with macros... but then again I'm not much good with templates either at least not by a boost mailing list standard anyway. :) Thanks, Derrick. // -- uniform_vector.hpp ------------------------------------------------------------------------ #ifndef DJH_UNIFORM_VECTOR_HPP #define DJH_UNIFORM_VECTOR_HPP #include <boost/tuple/tuple.hpp> namespace djh { template <unsigned int N, class Type> struct uniform_vector; template <class Type> struct uniform_vector<0, Type>; template <class Type> struct uniform_vector<1, Type> { static const unsigned int size = 1; typedef boost::tuple<Type> type; }; template <class Type> struct uniform_vector<2, Type> { static const unsigned int size = 2; typedef boost::tuple<Type, Type> type; }; template <class Type> struct uniform_vector<3, Type> { static const unsigned int size = 3; typedef boost::tuple<Type, Type, Type> type; }; template <class Type> struct uniform_vector<4, Type> { static const unsigned int size = 4; typedef boost::tuple<Type, Type, Type, Type> type; }; } #endif

2 3

Merging sorted iterators?
by Dr. Roland Bock 17 Sep '08

17 Sep '08

Hi, I have a bunch of iterators over some sorted input from several big files which do not fit into memory. What I would like to have is one iterator which allows me to iterate over the combined sorted input (of course, the result should be sorted, too). Something like this: -------\ --------\ ... ---- --------/ -------/ Is there a class in boost which allows me to do this in a simple way? Thanks and regards, Roland

3 2

[filesystem] Strange problems on Cygwin with 1.36.0
by Boris 17 Sep '08

17 Sep '08

According to http://www.boost.org/doc/libs/1_36_0/libs/filesystem/doc/index.htm#Cgywin Boost.Filesystem seems to be supported on Cygwin. However with the latest version 1.36.0 (which I've built with BOOST_POSIX_API defined) I don't seem to get any Boost.Filesystem function to work on Cygwin. I'm currently running some test cases and had to modify them for Cygwin like this: #if defined(__CYGWIN__) BOOST_CHECK(a == b); #else BOOST_CHECK(bfs::equivalent(a, b)); #endif If a and b contain the very same path you would expect both a == b and boost::filesystem::equivalent() to return true. However on Cygwin boost::filesystem::equivalent() returns false. If I use BOOST_CHECK(bfs::equivalent(a, b)) on Cygwin I get this error: unknown location(0): fatal error in "test_find_param": std::runtime_error: boost::filesystem::equivalent: No such file or directory: "\home\Boris\boost_1_36_0\libs\process\test\..\..\..\bin.v2\libs\process\test\gcc-3.4.4\debug\link-static\runtime-link-static\helpers.exe", "\home\Boris\boost_1_36_0\libs\process\test\..\..\..\bin.v2\libs\process\test\gcc-3.4.4\debug\link-static\runtime-link-static\helpers.exe" Both strings are exactly the same and contain valid Cygwin paths. Is Boost.Filesystem broken on Cygwin? Boris PS: I've a similar problem with boost::filesystem::create_directory() which can't create a directory for no obvious reason.

3 6

[smart_ptr] Problem with circular reference detection
by Jim Barry 17 Sep '08

17 Sep '08

I am trying to detect circular references using the find_unreachable_objects function in sp_collector.cpp. However, it reports false positives in the simple case of one object being owned by another. To illustrate: #define BOOST_SP_ENABLE_DEBUG_HOOKS #include <boost/shared_ptr.hpp> // sp_collector.cpp exported functions std::size_t find_unreachable_objects(bool report); struct X {}; struct Y { boost::shared_ptr<X> pX; Y() : pX(new X) {} }; int main() { boost::shared_ptr<Y> y(new Y); find_unreachable_objects(true); return 0; } The output is as follows: ... 2 objects in m. ... 1 objects in m2. ... 1 objects in open. Unreachable object at 0036BA60, 12 bytes long. Clearly this is incorrect as there are no circular references. Looking at the code, there is what appears to be a typo at line 124: std::cout << "... " << m2.size() << " objects in open.\n"; should be: std::cout << "... " << open.size() << " objects in open.\n"; But the real problem seems to be at line 121: if(p->use_count() != i->second) open.push_back(p); I'm not entirely sure what's going on in this section of code, but it ends up determining (incorrectly) that the "X" object is unreachable. Is there any way to get this working properly? Thanks, - Jim

3 5

Boost Serialize and derived classes
by John Yamokoski 17 Sep '08

17 Sep '08

I can't seem to figure out if what I want is possible. I am using the Boost Serialization library to save the state of my program. I have a manager-type class which holds onto pointers of a base class. The Boost Serialization website has a lot to say about base and derived classes. In a tutorial they show one way to deal with them. But it assumes a priori knowledge of the derived classes: class bus_route { friend class boost::serialization::access; friend std::ostream & operator<<(std::ostream &os, const bus_route &br); typedef bus_stop * bus_stop_pointer; std::list<bus_stop_pointer> stops; template<class Archive> void serialize(Archive &ar, const unsigned int version) { // in this program, these classes are never serialized directly but rather // through a pointer to the base class bus_stop. So we need a way to be // sure that the archive contains information about these derived classes. //ar.template register_type<bus_stop_corner>(); ar.register_type(static_cast<bus_stop_corner *>(NULL)); // TYPES KNOWN A PRIORI //ar.template register_type<bus_stop_destination>(); ar.register_type(static_cast<bus_stop_destination *>(NULL)); // TYPES KNOWN A PRIORI // serialization of stl collections is already defined // in the header ar & BOOST_SERIALIZATION_NVP(stops); } public: bus_route(){} void append(bus_stop *_bs) { stops.insert(stops.end(), _bs); } }; In my code, however, I don't know exactly what the derived classes might be. The derived objects are created and passed back to my manager class through the base pointer. One option I was considering was that when plugins register with the program, they would have to also register the derived class names with the archive. How this would actually be implemented, I have no idea. Can an archive object be created independently from the Serialization library, modified and then sent back to Serialization? Here is a simple test case I am working on, but I just can't get it to write out the derived classes to file: class base { friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int file_version) { ar & BOOST_SERIALIZATION_NVP(val); } public: base() : val(0.0) {} base(double b) : val(b) {} double val; }; class derived : public base { friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int file_version) { boost::serialization::void_cast_register<derived, base>(static_cast<derived *>(NULL),static_cast<base *>(NULL)); ar & BOOST_SERIALIZATION_NVP(val); } public: derived() : base(0.0), val(0) {} derived(int a, double b) : base(b), val(a) {} int val; }; BOOST_CLASS_EXPORT_GUID(derived, "derived"); class manager { friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int file_version) { ar & BOOST_SERIALIZATION_NVP(objs); } public: std::list<base*> objs; }; void main() { manager mgr; // Creat some data to save derived *d1, *d2, *d3, *d4; d1 = new derived(1, 1.0); d2 = new derived(2, 3.0); mgr.objs.push_back( d1 ); mgr.objs.push_back( d2 ); { std::ofstream ofs("out.xml"); assert(ofs.good()); boost::archive::xml_oarchive oa(ofs); oa << BOOST_SERIALIZATION_NVP(mgr); ofs.close(); } manager new_mgr; { std::ifstream ifs("out.xml"); assert(ifs.good()); boost::archive::xml_iarchive ia(ifs); ia >> BOOST_SERIALIZATION_NVP(new_mgr); ifs.close(); } { std::ofstream ofs("check.xml"); assert(ofs.good()); boost::archive::xml_oarchive oa(ofs); oa << BOOST_SERIALIZATION_NVP(new_mgr); ofs.close(); } delete d1, d2; } I apologize if this commonly asked but I was just looking for some insight as to the best approach. J.D.

2 3

How to implement a reader-writer lock with quick response to writer?
by Bill David 17 Sep '08

17 Sep '08

Question: How to implement a reader-writer lock with quick response to writer? I am using boost1.35.0. I know 1.36.0 is available, but I failed to build regex in 1.36.0 with VC8. I have the following requirements: 1.There are several reader threads and a single writer thread. The writer thread will update shared data periodically. And reader threads may read shared data at any time. 2.As long as writer thread try to update data, all reader thread should be blocked as soon as possible. 3.And reader thread should be blocked during data updating. 4.But reader threads can't be interrupted when they are reading. (i.e. all upcoming reader threads should be blocked, but writer still need wait till the last reader thread in reading finished.) How to implement above scenario with quick response to writer thread? I tried to write the following program to test shared_mutex, but it seems it can't meet my requirement: #include <iostream> using namespace std; #include <boost/thread/shared_mutex.hpp> #include <boost/thread/thread.hpp> #include <boost/thread/xtime.hpp> using namespace boost; boost::shared_mutex rw_mutex; int g_i = 1; void sleep(int iSec) { boost::xtime xt; boost::xtime_get(&xt, boost::TIME_UTC); xt.sec += iSec; boost::thread::sleep(xt); } void read() { while (true) { rw_mutex.lock_shared(); cout << g_i << endl; sleep(3); rw_mutex.unlock_shared(); } } int main() { boost::thread_group threads; for (int i = 0; i < 3; i++) threads.create_thread(read); sleep(1); rw_mutex.lock(); cout << "lock" << endl; g_i = 2; cout << "update g_i to " << g_i << endl; sleep(5); cout << "unlock" << endl; rw_mutex.unlock(); threads.join_all(); } The problem to above solution is it can't respond to writer thread as soon as possible, and only in rare case, it can respond to writer thread in about 6 seconds. The following is a possible output I have gotten: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 lock update g_i to 2 unlock 222 2 2 2 The writer thread should wait no more than 6 seconds, but it seems it can wait more than 20 seconds. Can shared_mutex meet such kind of scenario? If not, do you have any other suggestion?

1 0