December 2006 - Boost-users - lists.preview.boost.org

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Ovanes Markarian > Sent: Friday, December 01, 2006 2:23 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > Just a small question: > > Isn't it so, that linked lists (double or single), can be spreaded in the > memory, so that cache > misses and page faults are subject to happen? I think deque is a better > choice (which can be > better optimized in terms of locality)... [Nat] My suggestion was based on the (implied) intrusive_ptr constraint of no additional heap resources. But if we were to use a policy-based approach, a deque implementation could be one of the policies.

1 0

Problem building Boost on Windows
by Ronald Levine 01 Dec '06

01 Dec '06

C:\kits\boost_1_33_1>bjam "-sTOOLS=vc-8_0" install ****************************************************** Building Boost.Iostreams with bzip2 support disabled. To enable bzip2, consult the Boost.Iostreams documentation ****************************************************** ****************************************************** Building Boost.Iostreams with zlib and gzip support disabled. To enable zlib and gzip, consult the Boost.Iostreams documentation ****************************************************** Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). ...patience... ...patience... ...found 13371 targets... ...updating 4941 targets... MkDir1 C:\Boost spawn: Invalid argument

1 0

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Steven T. Hatton > Sent: Friday, December 01, 2006 7:03 AM > To: Boost-users(a)lists.boost.org > Subject: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > Intrusive pointers are, IMO, the correct mechanism to hold a > tree together. They are efficient, effective, and easy to > understand. They work better than shared pointers for > recursive data structures. Unfortunately the lack of a > corresponding weak_ptr to go with boost::intrusive_ptr means I > either use a raw pointer to the parent, devise a strategy for > decrementing the pointers myself, or create my own weak > pointer. [Nat] I've thought about this problem to some extent. This is the way I see it so far. The one big advantage of a weak_ptr over a raw pointer is that the weak_ptr can tell you when its referent becomes invalid. If your data structure requires storing a pointer in a place which itself becomes invalid as soon as the referent becomes invalid (e.g. recursive deletion of a subtree), then I'm not convinced that weak_ptr is significantly better than a raw pointer. (Even "snipping loose" a subtree, but preserving it, doesn't seem to argue for weak_ptr as a parent pointer. In such a case, you'd want to clear the parent pointer anyway, wouldn't you?) Does this rationale make sense, or am I overlooking something? I'm not rejecting the idea of a weak_ptr that works with intrusive_ptr. I'm just saying that for certain use cases, raw pointers seem to address the requirements equally well -- with less runtime overhead. This case looks like one of them.

2 1

serialization problem
by nico35＠noos.fr 01 Dec '06

01 Dec '06

Hello, I am using boost1.33.1 and spirit1.6 and borland c++ builder. I have a compilation problem with the following source code. The error message is [C++ Error] void_cast.hpp(154): E2451 Undefined symbol 'get_instance' [C++ Error] void_cast.hpp(155): E2451 Undefined symbol 'get_instance' [C++ Error] void_cast.hpp(154): E2285 Could not find a match for 'void_caster::void_caster(undefined,undefined)' Can anybody help me fixing this compilation problem ? Than you in advance. Marc. //--------------------------------------------------------------------------- #include <vcl.h> #pragma hdrstop #include <boost/serialization/utility.hpp> // includes for xml serialization #include <boost/archive/xml_oarchive.hpp> #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/basic_xml_oarchive.hpp> #include <boost/archive/basic_xml_iarchive.hpp> #include <boost/archive/impl/xml_iarchive_impl.ipp> #include <boost/archive/impl/xml_oarchive_impl.ipp> #include <boost/archive/impl/basic_xml_oarchive.ipp> #include <boost/archive/impl/basic_xml_iarchive.ipp> #include <boost/archive/basic_text_oprimitive.hpp> #include <boost/archive/basic_text_iprimitive.hpp> #include <boost/archive/impl/basic_text_oprimitive.ipp> #include <boost/archive/impl/basic_text_iprimitive.ipp> #include <boost/archive/impl/basic_xml_grammar.hpp> #include <libs/serialization/src/basic_xml_grammar.ipp> #include <boost/serialization/nvp.hpp> #include <boost/serialization/base_object.hpp> #include <boost/serialization/export.hpp> #include <boost/serialization/type_info_implementation.hpp> #include <boost/serialization/extended_type_info_no_rtti.hpp> #include <boost/serialization/extended_type_info_typeid.hpp> #include<iostream> #include<fstream> #include <vector> #include <boost/serialization/vector.hpp> class BaseClass { protected: friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & BOOST_SERIALIZATION_NVP(m_degrees); ar & BOOST_SERIALIZATION_NVP(m_minutes); ar & BOOST_SERIALIZATION_NVP(m_seconds); } int m_degrees; int m_minutes; float m_seconds; public: BaseClass() { m_degrees = 0; m_minutes = 0; m_seconds = 0.0; } BaseClass(int degrees, int minutes, float seconds) { m_degrees = degrees; m_minutes = minutes; m_seconds = seconds; } virtual void display() const { std::cout << "------------------------------------" << std::endl; std::cout << "m_degrees= " << m_degrees << std::endl; std::cout << "m_minutes= " << m_minutes << std::endl; std::cout << "m_seconds= " << m_seconds << std::endl; std::cout << "------------------------------------" << std::endl; } }; BOOST_CLASS_TYPE_INFO( BaseClass, extended_type_info_no_rtti<BaseClass> ) BOOST_CLASS_EXPORT(BaseClass) class DerivedClass1: public BaseClass { protected: friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & BOOST_SERIALIZATION_BASE_OBJECT_NVP(BaseClass); // ar & BOOST_SERIALIZATION_NVP(m_vector); } mutable std::vector<int> m_vector; public: DerivedClass1(): BaseClass() { } DerivedClass1(int degrees, int minutes, float seconds): BaseClass(degrees, minutes, seconds) { } void push_back(int arg) const { m_vector.push_back(arg); } virtual void display() const { std::cout << "***********************************************" << std::endl; BaseClass::display(); for (std::vector<int>::const_iterator iter=m_vector.begin(); iter!=m_vector.end(); iter++) { std::cout << " value= " << *iter << std::endl; } std::cout << "***********************************************" << std::endl; } }; BOOST_CLASS_TYPE_INFO( DerivedClass1, extended_type_info_no_rtti<DerivedClass1> ) BOOST_CLASS_EXPORT(DerivedClass1) BOOST_CLASS_TYPE_INFO( std::vector<double>, extended_type_info_no_rtti<std::vector<double> > ) BOOST_CLASS_EXPORT(std::vector<double>) int main() { const DerivedClass1 g(35, 59, 24.567f); g.push_back(10); g.push_back(20); g.push_back(30); g.display(); { std::ofstream ofs("filename"); boost::archive::xml_oarchive oa(ofs); oa << boost::serialization::make_nvp("g1", g); } DerivedClass1 newg; { std::ifstream ifs("filename"); boost::archive::xml_iarchive ia(ifs); ia >> boost::serialization::make_nvp("g1", newg); } newg.display(); return 0; } //--------------------------------------------------------------------------- /* ------------------------------------------------------------- Noos, votre bureau virtuel sur Internet : Mail... Web : webmail.noos.fr

2 1

Re: [Boost-users] [bind, function, mem_fn] Connection to a base class
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Johannes Brunen > Sent: Friday, December 01, 2006 7:10 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] [bind,function,mem_fn] Connection to a base class [Nat] This is top-of-the-head, no tests. > I have a base class X with two functions f and g > > class X { > virtual bool f() const; > virtual bool g(int) const; > }; > > and another class Y with functions f1, f2 and F > > class Y { > void f1(); > void f2(); > void F(???); [Nat] Seems like you'd want to make F accept boost::function<bool()>, and call it without arguments. > }; > > Y::f1() { > // call F so that F works on X::f > F(??? X::f); [Nat] F(boost::bind(&X::f, this)) > } > > Y::f2() { > // call F so that F works on Y::g but with g's int parameter known > and determined in Y::f2 > int v = 5; > F(??? X::g, v); [Nat] F(boost::bind(&X::g, this, v)) > }

1 0

how to clear a Graph
by Ufuk Okuyucu 01 Dec '06

01 Dec '06

Hi, I've just been handed a code using boost's cuthill-mckee algorithm. I found a memory leak within the code block where a number of edges being added to a graph but never cleaned at the end. Now, graph has a "clear" method where it's supposed to delete vertices and edges but it does not effect the memory leak, when I check the used memory size its still there. I need a way to delete a graph object and free all its internal mallocations if any. ...and I haven't used boost before... Here is the code block: //create incidence graph G2 from adj_list //apply cuthill-mckee on G2 //write out resulting permutation to perm_result Graph G2(matrix_size); //Graph data structure used in Cuthill-Mckee //add all the edges in adj_list to G2 for(int row=0;row<matrix_size;row++) { for(ROW_ADJ::iterator col=adj_list[row].begin(); col != adj_list[row].end(); col++) { add_edge (row, *col, G2); //<<-----this is where it gets all that memory } } cout <<"edges added\n"; print_system_stats(); graph_traits<Graph>::vertex_iterator ui, ui_end; property_map<Graph,vertex_degree_t>::type deg = get(vertex_degree, G2); for (boost::tie(ui, ui_end) = vertices(G2); ui != ui_end; ++ui) deg[*ui] = degree(*ui, G2); property_map<Graph, vertex_index_t>::type index_map = get(vertex_index, G2); //printf("original bandwidth: %d\n",bandwidth(G2)); printf("# vertices: %d\n",num_vertices(G2)); printf("# edges: %d\n",num_edges(G2)); vector<Vertex> inv_perm(num_vertices(G2)); vector<size_type> perm(num_vertices(G2)); Vertex start_vertex = vertex(0,G2); cuthill_mckee_ordering(G2, start_vertex, inv_perm.rbegin(), get(vertex_color, G2), get(vertex_degree, G2)); for (size_type c = 0; c != inv_perm.size(); ++c) perm[index_map[inv_perm[c]]] = c; //cout << " bandwidth: " << bandwidth(G2, make_iterator_property_map(&perm[0], index_map, perm[0])) << endl; //perm has the new ordering, copy it to perm_result for (int i=0;i<matrix_size;i++) perm_result[i] = perm[i]; G2.clear(); //<<-------this statement does not release all the memory inv_perm.clear(); perm.clear(); -- * Ufuk Okuyucu** *

1 0

configure boost on cygwin
by Shan 01 Dec '06

01 Dec '06

Hi, I'm trying to use boost in a cygwin enviroment. I've installed the boost-1_33_1 libraries using the cygwin setup program. I have a makefile to compile a simple program but I keep getting the error. $ make g++ -c -O3 cash_flow.cpp In file included from cash_flow.cpp:6: cash_flow.h:12: error: `boost' has not been declared cash_flow.h:12: error: ISO C++ forbids declaration of `shared_ptr' with no type cash_flow.h:12: error: expected `;' before '<' token make: *** [cash_flow.o] Error 1 I'm new to Unix like environments and am not able to setup the makefile properly. Below is what I have TARGET = main CPP = g++ CPPFLAGS = -O3 INCLUDES = -I/usr/include/boost-1_33_1 LDFLAGS = LIBS = -lboost_1_33_1 SOURCES = $(wildcard *.cpp) DEPENDS = $(SOURCES:%.cpp=%.d) OBJECTS = $(DEPENDS:%.d=%.o) TMPS = $(wildcard *.hpp~) $(wildcard *.cpp~) $(TARGET) : $(OBJECTS) $(CPP) -o $@ $(LDFLAGS) $(OBJECTS) $(LIBS) clean : rm -rf $(TARGET) $(DEPENDS) $(OBJECTS) $(TMPS) $(DEPENDS) : %.d : %.cpp set -e; $(CPP) -MM $(CPPFLAGS) $< \ | sed 's/$$*$\.o[ :]*/\1.o $@ : /g' > $@; \ [ -s $@ ] || rm -f $@ $(OBJECTS) : %.o : %.cpp $(CPP) -c $(CPPFLAGS) $< .PHONY : clean include $(DEPENDS) I also tried using the includes/library names I tried that as well INCLUDES = -I/usr/include/boost-1_33_1 LDFLAGS = -L/usr/lib LIBS = -lboost_1_33_1 and $(OBJECTS) : %.o : %.cpp $(CPP) -c $(CPPFLAGS) $(INCLUDES) $< Still get the same error .. Can someone tell me which libboost*.a should I point to as there are many .. for eg to use boost::shared_ptr? Thanks S.

2 2

#include <boost/lambda/lambda.hpp>
by Andreas Wißmeier 01 Dec '06

01 Dec '06

> Dear all, > > if I include this file I get the following error-message: > > d:\cvsatSixt\sixt_offer_src\boost\boost\lambda\detail\arity_code.hpp(66): > error C2065: 'T' : undeclared identifier > > > I am using > Microsoft Visual C++ .NET version 1.0.3705 > > I am probably missing something very basic. > > Kind regards > Andreas > > > > > Andreas Wißmeier > GDS-Projekt > Sixt GmbH & Co. Autovermietung KG > Zugspitzstr. 1 > D-82049 Pullach > Tel.: +49 (0)89 74444 4229 > E-Mail.: andreas.wissmeier(a)sixt.de >

1 0

[bind,function,mem_fn] Connection to a base class
by Johannes Brunen 01 Dec '06

01 Dec '06

Hi, I have the following problem and don't know how to design it with the boost tools at hand: I have a base class X with two functions f and g class X { virtual bool f() const; virtual bool g(int) const; }; and another class Y with functions f1, f2 and F class Y { void f1(); void f2(); void F(???); }; Now, f1 and f2 should call F, but f1 should determine that F calls X::f polymorphically and f2 should determine that F calls X::g polymorphically. Y::f1() { // call F so that F works on X::f F(??? X::f); } Y::f2() { // call F so that F works on Y::g but with g's int parameter known and determined in Y::f2 int v = 5; F(??? X::g, v); } Y::F(???) { smart_ptr<Y> sp1(new Y); smart_ptr<Y> sp2(new Y); // Either call Y::f on sp1 and sp2, or call Y::g(5) on sp1 and sp2 } Is this possible? I have looked into the documentation of boost::bind, boost::mem_fn and boost::function but didn't find a way to do that. Could someone give me some hint? With best regards Johannes Brunen

1 0

boost::weak_ptr and boost::intrusive_ptr
by Steven T. Hatton 01 Dec '06

01 Dec '06

Why is there no counterpart for boost::weak_ptr that works with boost::intrusive_ptr? It's certainly possible to set up cycles with boost::intrusive_ptr. A simple example where intrusive_ptr is the right thing to use, and where the problem arrises is a tree structure. It is sometimes useful to maintain a bidirectional pointer relationship between children and parent nodes in a tree. Ideally, deleting the root of a tree will cascade down to delete all intermediate and leaf nodes which are not referenced by some other object. If both the downward and upward pointers are intrusive, removing the pointer to the root will result in orphaned data because the upward pointers prevent the root node from being deleted, while the downward pointers prevent the child nodes (and their upward pointers) from being removed. Intrusive pointers are, IMO, the correct mechanism to hold a tree together. They are efficient, effective, and easy to understand. They work better than shared pointers for recursive data structures. Unfortunately the lack of a corresponding weak_ptr to go with boost::intrusive_ptr means I either use a raw pointer to the parent, devise a strategy for decrementing the pointers myself, or create my own weak pointer. All of these options detract significantly from the value of boost::intrusive_ptr. I have the impression that in the past people have used intrusive pointers for the "wrong things", and this has caused them to get an overall bad reputation. IMO, they do have a place in good software design. Any thoughts on this topic? Steven

1 0