Subject: CGAL users discussion list
List archive
- From: "Sebastien Loriot (GeometryFactory)" <>
- To:
- Subject: Re: [cgal-discuss] Stack overflow in Alpha_shape_2
- Date: Tue, 31 May 2011 10:08:41 +0200
Here is a patched file that corrects the overflow problem.
S.
Marc Glisse wrote:
On Wed, 25 May 2011, GIS Worx wrote:
Thanks for the thread on the other issue. I think I may try avoiding the
Lazy_kernel, to see how performance is affected.
Note that when you disable the lazy kernel, IIRC, Epeck still uses a lazy number type, which could make it even worse. If you don't want any laziness, you could try something like: Filtered_kernel<Simple_cartesian<Gmpq> >
Many thanks for the suggestions, with a nudge to the right stack reserve
size, it works.
Using /F didn't produce results when set to an unreasonably low value. I
learned you have to actually modify the .exe for any /STACK option to work
(rather than my little CGAL related DLL). Using a VS2008 build environment,
a stack reserve size of 1750000B succeeds and 1500000B fails. The default
for the linker is 1MB.
Less than 2MB, so it's not too wild.
Sometimes, GMP can use quite a bit of stack space itself. I don't know how you got your GMP, but replacing 65536 by something much smaller (say 1024) in the definition of TMP_ALLOC in gmp-impl.h and recompiling can help.
I know that the Lazy_kernel is supposed to be more efficient, but I wonder
if recursion this deep really be more efficient that simply outright
evaluating on the spot.
Well the whole point is that in many cases you don't need the evaluation, and the approximate value stored in each node is sufficient. So in some sense, the deeper the DAG, the larger the potential gain.
I also wonder if there is something about my scenario that produces a particularly pathological case for the lazy evaluation.
If you post a full reproducible example, maybe someday someone will look into it.
// Copyright (c) 1999-2007 INRIA Sophia-Antipolis (France). // All rights reserved. // // This file is part of CGAL (www.cgal.org); you can redistribute it and/or // modify it under the terms of the GNU Lesser General Public License as // published by the Free Software Foundation; version 2.1 of the License. // See the file LICENSE.LGPL distributed with CGAL. // // Licensees holding a valid commercial license may use this file in // accordance with the commercial license agreement provided with the software. // // This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE // WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. // // $URL: svn+ssh://scm.gforge.inria.fr/svn/cgal/branches/CGAL-3.8-branch/Number_types/include/CGAL/Lazy_exact_nt.h $ // $Id: Lazy_exact_nt.h 59568 2010-11-08 14:52:56Z sloriot $ // // // Author(s) : Sylvain Pion #ifndef CGAL_LAZY_EXACT_NT_H #define CGAL_LAZY_EXACT_NT_H #define CGAL_int(T) typename First_if_different<int, T>::Type #define CGAL_double(T) typename First_if_different<double, T>::Type #define CGAL_To_interval(T) To_interval<T> #include <CGAL/number_type_basic.h> #include <boost/iterator/transform_iterator.hpp> // for Root_of functor #include <boost/static_assert.hpp> #include <boost/operators.hpp> #include <CGAL/Interval_nt.h> #include <CGAL/Handle.h> #include <CGAL/NT_converter.h> #include <CGAL/Profile_counter.h> #include <CGAL/Lazy.h> // #include <CGAL/Root_of_traits.h> // TODO /* * This file contains the definition of the number type Lazy_exact_nt<ET>, * where ET is an exact number type (must provide the exact operations needed). * * Lazy_exact_nt<ET> provides a DAG-based lazy evaluation, like LEDA's real, * Core's Expr, LEA's lazy rationals... * * The values are first approximated using Interval_base. * The exactness is provided when needed by ET. * * Lazy_exact_nt<ET> is just a handle to the abstract base class * Lazy_exact_nt_rep which has pure virtual methods .approx() and .exact(). * From this class derives one class per operation, with one constructor. * * The DAG is managed by : * - Handle and Rep. * - virtual functions to denote the various operators (instead of an enum). * * Other packages with vaguely similar design : APU, MetaCGAL, LOOK. */ /* * TODO : * - Generalize it for constructions at the kernel level. * - Add mixed operations with ET too ? * - Interval refinement functionnality ? * - Separate the handle and the representation(s) in 2 files (?) * maybe not a good idea, better if everything related to one operation is * close together. * - Add a CT template parameter ? * - Add a string constant to provide an expression string (a la MetaCGAL) ? * // virtual ostream operator<<() const = 0; // or string, like Core ? * - Have a template-expression (?) thing that evaluates a temporary element, * and allocates stuff in memory only when really needs to convert to the * NT. (cf gmp++, and maybe other things, Blitz++, Synaps...) */ /* * Interface of the rep classes: * - .approx() returns Interval_nt<> (assumes rounding=nearest). * [ only called from the handle, and declared in the base ] * - .exact() returns ET, if not already done, computes recursively * * - .rafine_approx() ?? */ namespace CGAL { template <class NT> class Lazy_exact_nt; // Abstract base representation class for lazy numbers template <typename ET> struct Lazy_exact_nt_rep : public Lazy_exact_nt<ET>::Self_rep { typedef typename Lazy_exact_nt<ET>::Self_rep Base; Lazy_exact_nt_rep (const Interval_nt<false> & i) : Base(i) {} #ifdef CGAL_LAZY_KERNEL_DEBUG void print_dag(std::ostream& os, int level) const { this->print_at_et(os, level); } #endif }; // int constant template <typename ET> struct Lazy_exact_Int_Cst : public Lazy_exact_nt_rep<ET> { Lazy_exact_Int_Cst (int i) : Lazy_exact_nt_rep<ET>(double(i)) {} void update_exact() const { this->et = new ET((int)this->approx().inf()); } }; // double constant template <typename ET> struct Lazy_exact_Cst : public Lazy_exact_nt_rep<ET> { Lazy_exact_Cst (double d) : Lazy_exact_nt_rep<ET>(d) {} void update_exact() const { this->et = new ET(this->approx().inf()); } }; // Exact constant template <typename ET> struct Lazy_exact_Ex_Cst : public Lazy_exact_nt_rep<ET> { Lazy_exact_Ex_Cst (const ET & e) : Lazy_exact_nt_rep<ET>(CGAL_NTS to_interval(e)) { this->et = new ET(e); } void update_exact() const { CGAL_error(); } }; // Construction from a Lazy_exact_nt<ET1> (which keeps the lazyness). template <typename ET, typename ET1> class Lazy_lazy_exact_Cst : public Lazy_exact_nt_rep<ET> { mutable Lazy_exact_nt<ET1> l; public: Lazy_lazy_exact_Cst (const Lazy_exact_nt<ET1> &x) : Lazy_exact_nt_rep<ET>(x.approx()), l(x) { this->set_depth(l.depth() + 1); } void update_exact() const { this->et = new ET(l.exact()); this->at = l.approx(); prune_dag(); } void prune_dag() const { l = Lazy_exact_nt<ET1>(); } }; // Unary operations: abs, sqrt, square. // Binary operations: +, -, *, /, min, max. // Base unary operation template <typename ET> struct Lazy_exact_unary : public Lazy_exact_nt_rep<ET> { mutable Lazy_exact_nt<ET> op1; Lazy_exact_unary (const Interval_nt<false> &i, const Lazy_exact_nt<ET> &a) : Lazy_exact_nt_rep<ET>(i), op1(a) { this->set_depth(op1.depth() + 1); } void prune_dag() const { op1 = Lazy_exact_nt<ET>(); } #ifdef CGAL_LAZY_KERNEL_DEBUG void print_dag(std::ostream& os, int level) const { this->print_at_et(os, level); if(this->is_lazy()){ msg(os, level, "Unary number operator:"); print_dag(op1, os, level+1); } } #endif }; // Base binary operation template <typename ET, typename ET1 = ET, typename ET2 = ET> struct Lazy_exact_binary : public Lazy_exact_nt_rep<ET> { mutable Lazy_exact_nt<ET1> op1; mutable Lazy_exact_nt<ET2> op2; Lazy_exact_binary (const Interval_nt<false> &i, const Lazy_exact_nt<ET1> &a, const Lazy_exact_nt<ET2> &b) : Lazy_exact_nt_rep<ET>(i), op1(a), op2(b) { this->set_depth((std::max)(op1.depth(), op2.depth()) + 1); } void prune_dag() const { op1 = Lazy_exact_nt<ET1>(); op2 = Lazy_exact_nt<ET2>(); } #ifdef CGAL_LAZY_KERNEL_DEBUG void print_dag(std::ostream& os, int level) const { this->print_at_et(os, level); if(this->is_lazy()){ msg(os, level, "Binary number operator:"); print_dag(op1, os, level+1); print_dag(op2, os, level+1); } } #endif }; // Here we could use a template class for all operations (STL provides // function objects plus, minus, multiplies, divides...). But it would require // a template template parameter, and GCC 2.95 seems to crash easily with them. // Macro for unary operations #define CGAL_LAZY_UNARY_OP(OP, NAME) \ template <typename ET> \ struct NAME : public Lazy_exact_unary<ET> \ { \ typedef typename Lazy_exact_unary<ET>::AT::Protector P; \ NAME (const Lazy_exact_nt<ET> &a) \ : Lazy_exact_unary<ET>((P(), OP(a.approx())), a) {} \ \ void update_exact() const \ { \ this->et = new ET(OP(this->op1.exact())); \ if (!this->approx().is_point()) \ this->at = CGAL_NTS to_interval(*(this->et)); \ this->prune_dag(); \ } \ }; CGAL_LAZY_UNARY_OP(opposite, Lazy_exact_Opp) CGAL_LAZY_UNARY_OP(CGAL_NTS abs, Lazy_exact_Abs) CGAL_LAZY_UNARY_OP(CGAL_NTS square, Lazy_exact_Square) CGAL_LAZY_UNARY_OP(CGAL_NTS sqrt, Lazy_exact_Sqrt) // A macro for +, -, * and / #define CGAL_LAZY_BINARY_OP(OP, NAME) \ template <typename ET, typename ET1 = ET, typename ET2 = ET> \ struct NAME : public Lazy_exact_binary<ET, ET1, ET2> \ { \ typedef typename Lazy_exact_binary<ET,ET1,ET2>::AT::Protector P; \ NAME (const Lazy_exact_nt<ET1> &a, const Lazy_exact_nt<ET2> &b) \ : Lazy_exact_binary<ET, ET1, ET2>((P(), a.approx() OP b.approx()), a, b) {} \ \ void update_exact() const \ { \ this->et = new ET(this->op1.exact() OP this->op2.exact()); \ if (!this->approx().is_point()) \ this->at = CGAL_NTS to_interval(*(this->et)); \ this->prune_dag(); \ } \ }; CGAL_LAZY_BINARY_OP(+, Lazy_exact_Add) CGAL_LAZY_BINARY_OP(-, Lazy_exact_Sub) CGAL_LAZY_BINARY_OP(*, Lazy_exact_Mul) CGAL_LAZY_BINARY_OP(/, Lazy_exact_Div) // Minimum template <typename ET> struct Lazy_exact_Min : public Lazy_exact_binary<ET> { Lazy_exact_Min (const Lazy_exact_nt<ET> &a, const Lazy_exact_nt<ET> &b) : Lazy_exact_binary<ET>((CGAL::min)(a.approx(), b.approx()), a, b) {} void update_exact() const { this->et = new ET((CGAL::min)(this->op1.exact(), this->op2.exact())); if (!this->approx().is_point()) this->at = CGAL_NTS to_interval(*(this->et)); this->prune_dag(); } }; // Maximum template <typename ET> struct Lazy_exact_Max : public Lazy_exact_binary<ET> { Lazy_exact_Max (const Lazy_exact_nt<ET> &a, const Lazy_exact_nt<ET> &b) : Lazy_exact_binary<ET>((CGAL::max)(a.approx(), b.approx()), a, b) {} void update_exact() const { this->et = new ET((CGAL::max)(this->op1.exact(), this->op2.exact())); if (!this->approx().is_point()) this->at = CGAL_NTS to_interval(*(this->et)); this->prune_dag(); } }; // The real number type, handle class template <typename ET_> class Lazy_exact_nt : public Lazy<Interval_nt<false>, ET_, Lazy_exact_nt<ET_>, To_interval<ET_> > , boost::ordered_euclidian_ring_operators2< Lazy_exact_nt<ET_>, int > , boost::ordered_euclidian_ring_operators2< Lazy_exact_nt<ET_>, double > { public: typedef Lazy_exact_nt<ET_> Self; typedef Lazy<Interval_nt<false>, ET_, Self, To_interval<ET_> > Base; typedef typename Base::Self_rep Self_rep; typedef typename Base::ET ET; // undocumented typedef typename Base::AT AT; // undocumented typedef typename Base::Exact_type Exact_type; typedef typename Base::Approximate_type Approximate_type; public : Lazy_exact_nt () {} Lazy_exact_nt (Self_rep *r) : Base(r) {} Lazy_exact_nt (const CGAL_int(ET) & i) : Base(new Lazy_exact_Int_Cst<ET>(i)) {} Lazy_exact_nt (unsigned i) : Base(new Lazy_exact_Cst<ET>(i)){} Lazy_exact_nt (const CGAL_double(ET) & d) : Base(new Lazy_exact_Cst<ET>(d)){} Lazy_exact_nt (const ET & e) : Base(new Lazy_exact_Ex_Cst<ET>(e)){} template <class ET1> Lazy_exact_nt (const Lazy_exact_nt<ET1> &x) : Base(new Lazy_lazy_exact_Cst<ET, ET1>(x)){} Self operator+ () const { return *this; } Self operator- () const { return new Lazy_exact_Opp<ET>(*this); } Self & operator+=(const Self& b) { return *this = new Lazy_exact_Add<ET>(*this, b); } Self & operator-=(const Self& b) { return *this = new Lazy_exact_Sub<ET>(*this, b); } Self & operator*=(const Self& b) { return *this = new Lazy_exact_Mul<ET>(*this, b); } Self & operator/=(const Self& b) { CGAL_precondition(b != 0); return *this = new Lazy_exact_Div<ET>(*this, b); } // Mixed operators. (could be optimized ?) Self & operator+=(CGAL_int(ET) b) { return *this = new Lazy_exact_Add<ET>(*this, b); } Self & operator-=(CGAL_int(ET) b) { return *this = new Lazy_exact_Sub<ET>(*this, b); } Self & operator*=(CGAL_int(ET) b) { return *this = new Lazy_exact_Mul<ET>(*this, b); } Self & operator/=(CGAL_int(ET) b) { CGAL_precondition(b != 0); return *this = new Lazy_exact_Div<ET>(*this, b); } Self & operator+=(CGAL_double(ET) b) { return *this = new Lazy_exact_Add<ET>(*this, b); } Self & operator-=(CGAL_double(ET) b) { return *this = new Lazy_exact_Sub<ET>(*this, b); } Self & operator*=(CGAL_double(ET) b) { return *this = new Lazy_exact_Mul<ET>(*this, b); } Self & operator/=(CGAL_double(ET) b) { CGAL_precondition(b != 0); return *this = new Lazy_exact_Div<ET>(*this, b); } // % kills filtering Self & operator%=(const Self& b) { CGAL_precondition(b != 0); ET res = this->exact(); res %= b.exact(); return *this = Lazy_exact_nt<ET>(res); } Self & operator%=(int b) { CGAL_precondition(b != 0); ET res = this->exact(); res %= b; return *this = Lazy_exact_nt<ET>(res); } Interval_nt<true> interval() const { const Interval_nt<false>& i = this->approx(); return Interval_nt<true>(i.inf(), i.sup()); } Interval_nt_advanced approx_adv() const { return this->ptr()->approx(); } static const double & get_relative_precision_of_to_double() { return relative_precision_of_to_double; } static void set_relative_precision_of_to_double(const double & d) { CGAL_assertion((0 < d) & (d < 1)); relative_precision_of_to_double = d; } bool identical(const Self& b) const { return ::CGAL::identical( static_cast<const Handle &>(*this), static_cast<const Handle &>(b)); } template < typename T > bool identical(const T&) const { return false; } private: static double relative_precision_of_to_double; }; template <typename ET> double Lazy_exact_nt<ET>::relative_precision_of_to_double = 0.00001; template <typename ET1, typename ET2> bool operator<(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); if (a.identical(b)) return false; Uncertain<bool> res = a.approx() < b.approx(); if (is_certain(res)) return get_certain(res); CGAL_BRANCH_PROFILER_BRANCH(tmp); return a.exact() < b.exact(); } template <typename ET1, typename ET2> bool operator==(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); if (a.identical(b)) return true; Uncertain<bool> res = a.approx() == b.approx(); if (is_certain(res)) return get_certain(res); CGAL_BRANCH_PROFILER_BRANCH(tmp); return a.exact() == b.exact(); } template <typename ET1, typename ET2> inline bool operator>(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { return b < a; } template <typename ET1, typename ET2> inline bool operator>=(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { return ! (a < b); } template <typename ET1, typename ET2> inline bool operator<=(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { return b >= a; } template <typename ET1, typename ET2> inline bool operator!=(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { return ! (a == b); } template <typename ET> inline Lazy_exact_nt<ET> operator%(const Lazy_exact_nt<ET>& a, const Lazy_exact_nt<ET>& b) { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); CGAL_precondition(b != 0); return Lazy_exact_nt<ET>(a) %= b; } // Mixed operators with int. template <typename ET> bool operator<(const Lazy_exact_nt<ET>& a, int b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain<bool> res = a.approx() < b; if (is_certain(res)) return res; CGAL_BRANCH_PROFILER_BRANCH(tmp); return a.exact() < b; } template <typename ET> bool operator>(const Lazy_exact_nt<ET>& a, int b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain<bool> res = b < a.approx(); if (is_certain(res)) return get_certain(res); CGAL_BRANCH_PROFILER_BRANCH(tmp); return b < a.exact(); } template <typename ET> bool operator==(const Lazy_exact_nt<ET>& a, int b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain<bool> res = b == a.approx(); if (is_certain(res)) return get_certain(res); CGAL_BRANCH_PROFILER_BRANCH(tmp); return b == a.exact(); } // Mixed operators with double. template <typename ET> bool operator<(const Lazy_exact_nt<ET>& a, double b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain<bool> res = a.approx() < b; if (is_certain(res)) return res; CGAL_BRANCH_PROFILER_BRANCH(tmp); return a.exact() < b; } template <typename ET> bool operator>(const Lazy_exact_nt<ET>& a, double b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain<bool> res = b < a.approx(); if (is_certain(res)) return res; CGAL_BRANCH_PROFILER_BRANCH(tmp); return b < a.exact(); } template <typename ET> bool operator==(const Lazy_exact_nt<ET>& a, double b) { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain<bool> res = b == a.approx(); if (is_certain(res)) return res; CGAL_BRANCH_PROFILER_BRANCH(tmp); return b == a.exact(); } template <typename ET1, typename ET2> Lazy_exact_nt< typename Coercion_traits<ET1, ET2>::Type > operator+(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Add<typename Coercion_traits<ET1, ET2>::Type, ET1, ET2>(a, b); } template <typename ET1, typename ET2> Lazy_exact_nt< typename Coercion_traits<ET1, ET2>::Type > operator-(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Sub<typename Coercion_traits<ET1, ET2>::Type, ET1, ET2>(a, b); } template <typename ET1, typename ET2> Lazy_exact_nt< typename Coercion_traits<ET1, ET2>::Type > operator*(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Mul<typename Coercion_traits<ET1, ET2>::Type, ET1, ET2>(a, b); } template <typename ET1, typename ET2> Lazy_exact_nt< typename Coercion_traits<ET1, ET2>::Type > operator/(const Lazy_exact_nt<ET1>& a, const Lazy_exact_nt<ET2>& b) { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); CGAL_precondition(b != 0); return new Lazy_exact_Div<typename Coercion_traits<ET1, ET2>::Type, ET1, ET2>(a, b); } // // Algebraic structure traits // namespace INTERN_LAZY_EXACT_NT { template< class NT, class Functor > struct Simplify_selector { struct Simplify : public std::unary_function<NT&, void> { void operator()( NT& ) const { // TODO: In the old implementation the Simplify-functor was the default // (which does nothing). But this cannot be the correct way!? } }; }; template< class NT > struct Simplify_selector< NT, Null_functor > { typedef Null_functor Simplify; }; template< class NT, class Functor > struct Unit_part_selector { struct Unit_part : public std::unary_function<NT, NT > { NT operator()( const NT& x ) const { return NT( CGAL_NTS unit_part( x.exact() ) ); } }; }; template< class NT > struct Unit_part_selector< NT, Null_functor > { typedef Null_functor Unit_part; }; template< class NT, class Functor > struct Is_zero_selector { struct Is_zero : public std::unary_function<NT, bool > { bool operator()( const NT& x ) const { return CGAL_NTS is_zero( x.exact() ); } }; }; template< class NT > struct Is_zero_selector< NT, Null_functor > { typedef Null_functor Is_zero; }; template< class NT, class Functor > struct Is_one_selector { struct Is_one : public std::unary_function<NT, bool > { bool operator()( const NT& x ) const { return CGAL_NTS is_one( x.exact() ); } }; }; template< class NT > struct Is_one_selector< NT, Null_functor > { typedef Null_functor Is_one; }; template< class NT, class Functor > struct Square_selector { struct Square : public std::unary_function<NT, NT > { NT operator()( const NT& x ) const { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Square<typename NT::ET>(x); } }; }; template< class NT > struct Square_selector< NT, Null_functor > { typedef Null_functor Square; }; template< class NT, class Functor > struct Integral_division_selector { struct Integral_division : public std::binary_function<NT, NT, NT > { NT operator()( const NT& x, const NT& y ) const { return NT( CGAL_NTS integral_division( x.exact(), y.exact() ) ); } CGAL_IMPLICIT_INTEROPERABLE_BINARY_OPERATOR( NT ) }; }; template< class NT > struct Integral_division_selector< NT, Null_functor > { typedef Null_functor Integral_division; }; template< class NT, class Functor > struct Is_square_selector { struct Is_square : public std::binary_function<NT, NT&, bool > { bool operator()( const NT& x, NT& y ) const { typename NT::ET y_et; bool result = CGAL_NTS is_square( x.exact(), y_et ); y = NT(y_et); return result; } bool operator()( const NT& x) const { typename NT::ET y_et; return CGAL_NTS is_square( x.exact(), y_et ); } }; }; template< class NT > struct Is_square_selector< NT, Null_functor > { typedef Null_functor Is_square; }; template <class NT, class AlgebraicStructureTag> struct Sqrt_selector{ struct Sqrt : public std::unary_function<NT, NT > { NT operator ()(const NT& x) const { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); CGAL_precondition(x >= 0); return new Lazy_exact_Sqrt<typename NT::ET>(x); } }; }; template <class NT> struct Sqrt_selector<NT,Null_functor> { typedef Null_functor Sqrt; }; template< class NT, class Functor > struct Kth_root_selector { struct Kth_root : public std::binary_function<int, NT, NT > { NT operator()( int k, const NT& x ) const { return NT( CGAL_NTS kth_root( k, x.exact() ) ); } }; }; template< class NT > struct Kth_root_selector< NT, Null_functor > { typedef Null_functor Kth_root; }; template< class NT, class Functor > struct Root_of_selector { private: struct Cast{ typedef typename NT::ET result_type; result_type operator()(const NT& lazy_exact) const { return lazy_exact.exact(); } }; public: struct Root_of { // typedef typename Functor::Boundary Boundary; typedef NT result_type; template< class Input_iterator > NT operator()( int k, Input_iterator begin, Input_iterator end ) const { Cast cast; return NT( typename Algebraic_structure_traits<typename NT::ET>:: Root_of()( k, ::boost::make_transform_iterator( begin, cast ), ::boost::make_transform_iterator( end, cast ) ) ); } }; }; template< class NT > struct Root_of_selector< NT, Null_functor > { typedef Null_functor Root_of; }; template< class NT, class Functor > struct Gcd_selector { struct Gcd : public std::binary_function<NT, NT, NT > { NT operator()( const NT& x, const NT& y ) const { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return NT( CGAL_NTS gcd( x.exact(), y.exact() ) ); } CGAL_IMPLICIT_INTEROPERABLE_BINARY_OPERATOR( NT ) }; }; template< class NT > struct Gcd_selector< NT, Null_functor > { typedef Null_functor Gcd; }; template< class NT, class Functor > struct Div_selector { struct Div : public std::binary_function<NT, NT, NT > { NT operator()( const NT& x, const NT& y ) const { return NT( CGAL_NTS div( x.exact(), y.exact() ) ); } CGAL_IMPLICIT_INTEROPERABLE_BINARY_OPERATOR( NT ) }; }; template< class NT > struct Div_selector< NT, Null_functor > { typedef Null_functor Div; }; template< class NT, class Functor > struct Inverse_selector { struct Inverse { typedef NT result_type; NT operator()( const NT& x ) const { return NT( 1 ) / x ; } }; }; template< class NT > struct Inverse_selector< NT, Null_functor > { typedef Null_functor Inverse; }; template< class NT, class Functor > struct Mod_selector { struct Mod : public std::binary_function<NT, NT, NT > { NT operator()( const NT& x, const NT& y ) const { return NT( CGAL_NTS mod( x.exact(), y.exact() ) ); } CGAL_IMPLICIT_INTEROPERABLE_BINARY_OPERATOR( NT ) }; }; template< class NT > struct Mod_selector< NT, Null_functor > { typedef Null_functor Mod; }; template< class NT, class Functor > struct Div_mod_selector { struct Div_mod { typedef void result_type; typedef NT first_argument_type; typedef NT second_argument_type; typedef NT& third_argument_type; typedef NT& fourth_argument_type; void operator()( const NT& x, const NT& y, NT& q, NT& r ) const { typename NT::ET q_et; typename NT::ET r_et; CGAL_NTS div_mod( x.exact(), y.exact(), q_et, r_et ); q = NT( q_et ); r = NT( r_et ); } template< class NT1, class NT2 > void operator()( const NT1& x, const NT2& y, NT& q, NT& r ) const { BOOST_STATIC_ASSERT((::boost::is_same< typename Coercion_traits< NT1, NT2 >::Type, NT >::value)); typename Coercion_traits< NT1, NT2 >::Cast cast; operator()( cast(x), cast(y), q, r ); } }; }; template< class NT > struct Div_mod_selector< NT, Null_functor >{ typedef Null_functor Div_mod; }; } // namespace INTERN_LAZY_EXACT_NT template <class ET> class Algebraic_structure_traits< Lazy_exact_nt<ET> > :public Algebraic_structure_traits_base < Lazy_exact_nt<ET>, typename Algebraic_structure_traits<ET>::Algebraic_category > { private: typedef Algebraic_structure_traits<ET> AST_ET; typedef typename AST_ET::Algebraic_category ET_as_tag; public: typedef typename AST_ET::Is_exact Is_exact; typedef typename AST_ET::Is_numerical_sensitive Is_numerical_sensitive; typedef typename INTERN_LAZY_EXACT_NT::Simplify_selector <Lazy_exact_nt<ET>, typename AST_ET::Simplify > ::Simplify Simplify; typedef typename INTERN_LAZY_EXACT_NT::Unit_part_selector <Lazy_exact_nt<ET>, typename AST_ET::Unit_part > ::Unit_part Unit_part; typedef typename INTERN_LAZY_EXACT_NT::Is_zero_selector <Lazy_exact_nt<ET>, typename AST_ET::Is_zero > ::Is_zero Is_zero; typedef typename INTERN_LAZY_EXACT_NT::Is_one_selector <Lazy_exact_nt<ET>, typename AST_ET::Is_one > ::Is_one Is_one; typedef typename INTERN_LAZY_EXACT_NT::Square_selector <Lazy_exact_nt<ET>, typename AST_ET::Square > ::Square Square; typedef typename INTERN_LAZY_EXACT_NT::Integral_division_selector <Lazy_exact_nt<ET>, typename AST_ET::Integral_division> ::Integral_division Integral_division; typedef typename INTERN_LAZY_EXACT_NT::Is_square_selector <Lazy_exact_nt<ET>, typename AST_ET::Is_square > ::Is_square Is_square; typedef typename INTERN_LAZY_EXACT_NT::Sqrt_selector <Lazy_exact_nt<ET>, typename AST_ET::Sqrt> ::Sqrt Sqrt; typedef typename INTERN_LAZY_EXACT_NT::Kth_root_selector <Lazy_exact_nt<ET>, typename AST_ET::Kth_root > ::Kth_root Kth_root; typedef typename INTERN_LAZY_EXACT_NT::Root_of_selector <Lazy_exact_nt<ET>, typename AST_ET::Root_of > ::Root_of Root_of; typedef typename INTERN_LAZY_EXACT_NT::Gcd_selector <Lazy_exact_nt<ET>, typename AST_ET::Gcd > ::Gcd Gcd; typedef typename INTERN_LAZY_EXACT_NT::Div_selector <Lazy_exact_nt<ET>, typename AST_ET::Div > ::Div Div; typedef typename INTERN_LAZY_EXACT_NT::Mod_selector <Lazy_exact_nt<ET>, typename AST_ET::Mod > ::Mod Mod; typedef typename INTERN_LAZY_EXACT_NT::Div_mod_selector <Lazy_exact_nt<ET>, typename AST_ET::Div_mod > ::Div_mod Div_mod; typedef typename INTERN_LAZY_EXACT_NT::Inverse_selector <Lazy_exact_nt<ET>, typename AST_ET::Inverse > ::Inverse Inverse; }; // // Real embeddalbe traits // template < typename ET > class Real_embeddable_traits< Lazy_exact_nt<ET> > : public INTERN_RET::Real_embeddable_traits_base< Lazy_exact_nt<ET> , CGAL::Tag_true > { // Every type ET of Lazy_exact_nt<ET> has to be real embeddable. BOOST_STATIC_ASSERT((::boost::is_same< typename Real_embeddable_traits< ET > ::Is_real_embeddable, Tag_true >::value)); public: typedef Lazy_exact_nt<ET> Type; class Abs : public std::unary_function< Type, Type > { public: Type operator()( const Type& a ) const { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Abs<ET>(a); } }; class Sgn : public std::unary_function< Type, ::CGAL::Sign > { public: ::CGAL::Sign operator()( const Type& a ) const { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); Uncertain< ::CGAL::Sign> res = CGAL_NTS sign(a.approx()); if (is_certain(res)) return get_certain(res); CGAL_BRANCH_PROFILER_BRANCH(tmp); return CGAL_NTS sign(a.exact()); } }; class Compare : public std::binary_function< Type, Type, Comparison_result > { public: Comparison_result operator()( const Type& a, const Type& b ) const { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); if (a.identical(b)) return EQUAL; Uncertain<Comparison_result> res = CGAL_NTS compare(a.approx(), b.approx()); if (is_certain(res)) return get_certain(res); CGAL_BRANCH_PROFILER_BRANCH(tmp); return CGAL_NTS compare(a.exact(), b.exact()); } CGAL_IMPLICIT_INTEROPERABLE_BINARY_OPERATOR_WITH_RT( Type, Comparison_result ) }; class To_double : public std::unary_function< Type, double > { public: double operator()( const Type& a ) const { CGAL_BRANCH_PROFILER(std::string(" failures/calls to : ") + std::string(CGAL_PRETTY_FUNCTION), tmp); const Interval_nt<false>& app = a.approx(); double r; if (fit_in_double(app, r)) return r; // If it's precise enough, then OK. if (has_smaller_relative_precision(app, Lazy_exact_nt<ET>::get_relative_precision_of_to_double())) return CGAL_NTS to_double(app); CGAL_BRANCH_PROFILER_BRANCH(tmp); // Otherwise we trigger exact computation first, // which will refine the approximation. a.exact(); return CGAL_NTS to_double(a.approx()); } }; class To_interval : public std::unary_function< Type, std::pair< double, double > > { public: std::pair<double, double> operator()( const Type& a ) const { CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return a.approx().pair(); } }; class Is_finite : public std::unary_function< Type, bool > { public: bool operator()( const Type& x ) const { return CGAL_NTS is_finite(x.approx()) || CGAL_NTS is_finite(x.exact()); } }; }; template <class ET1, class ET2, class F> class Lazy_exact_nt_coercion_traits_base { public: typedef Tag_false Are_explicit_interoperable; typedef Tag_false Are_implicit_interoperable; //typedef Null_type Type typedef Null_functor Cast; }; template <class ET1, class ET2> class Lazy_exact_nt_coercion_traits_base < Lazy_exact_nt<ET1>, Lazy_exact_nt<ET2>, Tag_true > { typedef Coercion_traits<ET1,ET2> CT; typedef Lazy_exact_nt<ET1> A; typedef Lazy_exact_nt<ET2> B; public: typedef Lazy_exact_nt<typename CT::Type> Type; typedef typename CT::Are_implicit_interoperable Are_explicit_interoperable; typedef typename CT::Are_implicit_interoperable Are_implicit_interoperable; class Cast{ private: template <class T> Type cast(const T& x) const{ return Type(x); } Type cast(const Type& x) const{ return x; } public: typedef Type result_type; Type operator()(const A& x) const { return cast(x);} Type operator()(const B& x) const { return cast(x);} }; }; CGAL_DEFINE_COERCION_TRAITS_FOR_SELF_TEM(Lazy_exact_nt<ET>, class ET) template<class ET1, class ET2 > struct Coercion_traits< Lazy_exact_nt<ET1>, Lazy_exact_nt<ET2> > :public Lazy_exact_nt_coercion_traits_base <Lazy_exact_nt<ET1>, Lazy_exact_nt<ET2>, typename Coercion_traits<ET1,ET2>::Are_implicit_interoperable>{}; #define CGAL_COERCION_TRAITS_LAZY_EXACT(NTX) \ template<class ET> \ struct Coercion_traits< NTX, Lazy_exact_nt<ET> >{ \ private: \ typedef Coercion_traits<NTX,ET> CT; \ typedef Lazy_exact_nt<ET> NT; \ public: \ typedef typename CT::Are_explicit_interoperable \ Are_explicit_interoperable; \ typedef typename CT::Are_implicit_interoperable \ Are_implicit_interoperable; \ private: \ static const bool interoperable \ =boost::is_same< Are_implicit_interoperable, Tag_false>::value; \ public: \ typedef typename boost::mpl::if_c <interoperable,Null_tag,NT> \ ::type Type; \ typedef typename boost::mpl::if_c <interoperable, Null_functor, \ INTERN_CT::Cast_from_to<NTX,NT> >::type Cast; \ }; \ \ template<class ET> \ struct Coercion_traits< Lazy_exact_nt<ET>, NTX > \ :public Coercion_traits<NTX, Lazy_exact_nt<ET> >{}; \ CGAL_COERCION_TRAITS_LAZY_EXACT(int) CGAL_COERCION_TRAITS_LAZY_EXACT(short) CGAL_COERCION_TRAITS_LAZY_EXACT(double) CGAL_COERCION_TRAITS_LAZY_EXACT(float) #undef CGAL_COERCION_TRAITS_LAZY_EXACT namespace INTERN_LAZY_EXACT_NT { template < typename NT, typename TAG > class Fraction_traits_base; template < class ET > class Fraction_traits_base <Lazy_exact_nt<ET> , CGAL::Tag_false> : public Fraction_traits<ET> { public: typedef Lazy_exact_nt<ET> Type; }; template < class ET > class Fraction_traits_base <Lazy_exact_nt<ET> , CGAL::Tag_true>{ typedef Fraction_traits<ET> ETT; typedef typename ETT::Numerator_type ET_numerator; typedef typename ETT::Denominator_type ET_denominator; public: typedef Lazy_exact_nt<ET> Type; typedef Tag_true Is_fraction; typedef Lazy_exact_nt<ET_numerator> Numerator_type; typedef Lazy_exact_nt<ET_denominator> Denominator_type; struct Common_factor : std::binary_function<Denominator_type,Denominator_type,Denominator_type>{ Denominator_type operator()(const Denominator_type& a, const Denominator_type& b) const { typename ETT::Common_factor common_factor; return Denominator_type(common_factor(a.exact(),b.exact())); } }; struct Compose : std::binary_function<Type,Numerator_type,Denominator_type>{ Type operator()(const Numerator_type& n, const Denominator_type& d) const { typename ETT::Compose compose; return Type(compose(n.exact(),d.exact())); } }; struct Decompose { typedef void result_type; typedef Type first_argument_type; typedef Numerator_type second_argument_type; typedef Denominator_type third_argument_type; void operator()(const Type& f, Numerator_type& n, Denominator_type& d) const { typename ETT::Decompose decompose; ET_numerator nn; ET_denominator dd; decompose(f.exact(),nn,dd); n = Numerator_type(nn); d = Denominator_type(dd); } }; }; } // namespace INTERN_LAZY_EXACT_NT template < class ET > class Fraction_traits< Lazy_exact_nt< ET > > :public INTERN_LAZY_EXACT_NT::Fraction_traits_base<Lazy_exact_nt<ET>, typename Fraction_traits<ET>::Is_fraction> {}; template < class ET > struct Min <Lazy_exact_nt<ET> > : public std::binary_function<Lazy_exact_nt<ET>,Lazy_exact_nt<ET>,Lazy_exact_nt<ET> > { Lazy_exact_nt<ET> operator()( const Lazy_exact_nt<ET>& x, const Lazy_exact_nt<ET>& y) const { if (x.identical(y)){ return x; } Uncertain<bool> res = x.approx() < y.approx(); if(is_certain(res)){ return res.make_certain() ? x : y; } CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Min<ET>(x, y); } }; template < class ET > struct Max <Lazy_exact_nt<ET> > : public std::binary_function<Lazy_exact_nt<ET>,Lazy_exact_nt<ET>,Lazy_exact_nt<ET> > { Lazy_exact_nt<ET> operator()( const Lazy_exact_nt<ET>& x, const Lazy_exact_nt<ET>& y) const { if (x.identical(y)){ return x; } Uncertain<bool> res = x.approx() > y.approx(); if(is_certain(res)){ return res.make_certain() ? x : y; } CGAL_PROFILER(std::string("calls to : ") + std::string(CGAL_PRETTY_FUNCTION)); return new Lazy_exact_Max<ET>(x, y); } }; template<typename ET> inline Lazy_exact_nt<ET> min BOOST_PREVENT_MACRO_SUBSTITUTION( const Lazy_exact_nt<ET> & x, const Lazy_exact_nt<ET> & y){ return CGAL::Min<Lazy_exact_nt<ET> > ()(x,y); } template<typename ET> inline Lazy_exact_nt<ET> max BOOST_PREVENT_MACRO_SUBSTITUTION( const Lazy_exact_nt<ET> & x, const Lazy_exact_nt<ET> & y){ return CGAL::Max<Lazy_exact_nt<ET> > ()(x,y); } template <typename ET> std::ostream & operator<< (std::ostream & os, const Lazy_exact_nt<ET> & a) { return os << CGAL_NTS to_double(a); } template <typename ET> std::istream & operator>> (std::istream & is, Lazy_exact_nt<ET> & a) { ET e; is >> e; if (is) a = e; return is; } template< class ET > class Is_valid< Lazy_exact_nt<ET> > : public std::unary_function< Lazy_exact_nt<ET>, bool > { public : bool operator()( const Lazy_exact_nt<ET>& x ) const { return is_valid(x.approx()); } }; template < typename ET > struct NT_converter < Lazy_exact_nt<ET>, ET > { const ET& operator()(const Lazy_exact_nt<ET> &a) const { return a.exact(); } }; namespace internal { // Returns true if the value is representable by a double and to_double() // would return it. False means "don't know" (the exact number type is not // queried). template < typename ET > inline bool fit_in_double(const Lazy_exact_nt<ET>& l, double& r) { return fit_in_double(l.approx(), r); } } // namespace internal // We create a type of new node in Lazy_exact_nt's DAG // for the make_root_of_2() operation. template <typename ET > struct Lazy_exact_ro2 : public Lazy_exact_nt_rep< typename Root_of_traits<ET>::RootOf_2 > { typedef typename Root_of_traits<ET>::RootOf_2 RO2; typedef Lazy_exact_nt_rep<RO2> Base; typedef typename Base::AT::Protector P; mutable Lazy_exact_nt<ET> op1, op2, op3; bool smaller; bool old_rep;//if the rep=true then representation with polynomial coeff, else alpha, beta, gamma Lazy_exact_ro2 (const Lazy_exact_nt<ET> &a, const Lazy_exact_nt<ET> &b, const Lazy_exact_nt<ET> &c, bool s) : Base((P(), make_root_of_2(a.approx(), b.approx(), c.approx(), s))), op1(a), op2(b), op3(c), smaller(s), old_rep(true) {} Lazy_exact_ro2 (const Lazy_exact_nt<ET> &a, const Lazy_exact_nt<ET> &b, const Lazy_exact_nt<ET> &c) : Base((P(), make_root_of_2(a.approx(), b.approx(), c.approx()))), op1(a), op2(b), op3(c), smaller(true), old_rep(false) {} void update_exact() const { if (old_rep) this->et = new RO2(make_root_of_2(op1.exact(), op2.exact(), op3.exact(), smaller)); else this->et = new RO2(make_root_of_2(op1.exact(), op2.exact(), op3.exact())); if (!this->approx().is_point()) this->at = to_interval(*(this->et)); this->prune_dag(); } void prune_dag() const { op1 = op2 = op3 = Lazy_exact_nt<ET>(); } }; template <typename NT > struct Root_of_traits< Lazy_exact_nt < NT > > { private: typedef Root_of_traits<NT> T; public: typedef Root_of_traits< Lazy_exact_nt < NT > > Base; typedef Lazy_exact_nt< typename T::RootOf_1 > RootOf_1; typedef Lazy_exact_nt< typename T::RootOf_2 > RootOf_2; typedef RootOf_2 Root_of_2; typedef RootOf_1 Root_of_1; struct Make_root_of_2{ typedef RootOf_2 result_type; Root_of_2 operator()(const Lazy_exact_nt<NT>& a, const Lazy_exact_nt<NT>& b, const Lazy_exact_nt<NT>& c) const{ return new Lazy_exact_ro2<NT>(a, b, c); }; RootOf_2 operator()(const Lazy_exact_nt<NT>& a, const Lazy_exact_nt<NT>& b, const Lazy_exact_nt<NT>& c, bool smaller) const{ return new Lazy_exact_ro2<NT>(a, b, c, smaller); }; }; }; //these two functions for test suite requirement template < typename RT > typename CGAL::Root_of_traits<CGAL::Lazy_exact_nt<RT> >::RootOf_2 make_sqrt(const CGAL::Lazy_exact_nt< RT> & r) { typedef Lazy_exact_nt< RT> TT; CGAL_assertion(r >= 0); if(CGAL_NTS is_zero(r)) return make_root_of_2((TT) 1,(TT) 0,(TT) 0); return make_root_of_2((TT) 1,(TT) 0,-r,false); } template < typename RT > void print(std::ostream &os, const CGAL::Lazy_exact_nt< Root_of_2<RT> > &r) { print(os,r.exact()); } namespace INTERN_LAZY_EXACT_NT { template< typename ET , typename Tag> class Modular_traits_base{ public: typedef Lazy_exact_nt<ET> NT; typedef ::CGAL::Tag_false Is_modularizable; typedef ::CGAL::Null_functor Residue_type; typedef ::CGAL::Null_functor Modular_image; typedef ::CGAL::Null_functor Modular_image_representative; }; template< typename ET > class Modular_traits_base<ET, Tag_true>{ typedef Modular_traits<ET> MT_ET; public: typedef Lazy_exact_nt<ET> NT; typedef CGAL::Tag_true Is_modularizable; typedef typename MT_ET::Residue_type Residue_type; struct Modular_image{ Residue_type operator()(const NT& a){ typename MT_ET::Modular_image modular_image; return modular_image(a.exact()); } }; struct Modular_image_representative{ NT operator()(const Residue_type& x){ typename MT_ET::Modular_image_representative modular_image_representative; return NT(modular_image_representative(x)); } }; }; } // namespace INTERN_LAZY_EXACT_NT template < typename ET > class Modular_traits<Lazy_exact_nt<ET> > :public INTERN_LAZY_EXACT_NT::Modular_traits_base <ET,typename Modular_traits<ET>::Is_modularizable>{}; #undef CGAL_double #undef CGAL_int #undef CGAL_To_interval } //namespace CGAL #endif // CGAL_LAZY_EXACT_NT_H
- [cgal-discuss] Stack overflow in Alpha_shape_2, gisworx, 05/24/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Marc Glisse, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Francois Berenger, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Laurent Rineau (GeometryFactory), 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, GIS Worx, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Marc Glisse, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, GIS Worx, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Marc Glisse, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Sebastien Loriot (GeometryFactory), 05/26/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Sebastien Loriot (GeometryFactory), 05/31/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Marc Glisse, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, GIS Worx, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Marc Glisse, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Francois Berenger, 05/25/2011
- Re: [cgal-discuss] Stack overflow in Alpha_shape_2, Marc Glisse, 05/25/2011
Archive powered by MHonArc 2.6.16.