1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
|
// -*- C++ -*-
// Copyright (C) 2007, 2008, 2009 Free Software Foundation, Inc.
//
// This file is part of the GNU ISO C++ Library. This library is free
// software; you can redistribute it and/or modify it under the terms
// of the GNU General Public License as published by the Free Software
// Foundation; either version 3, or (at your option) any later
// version.
// This library is distributed in the hope that it will be useful, but
// WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
// General Public License for more details.
// Under Section 7 of GPL version 3, you are granted additional
// permissions described in the GCC Runtime Library Exception, version
// 3.1, as published by the Free Software Foundation.
// You should have received a copy of the GNU General Public License and
// a copy of the GCC Runtime Library Exception along with this program;
// see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
// <http://www.gnu.org/licenses/>.
/** @file parallel/unique_copy.h
* @brief Parallel implementations of std::unique_copy().
* This file is a GNU parallel extension to the Standard C++ Library.
*/
// Written by Robert Geisberger and Robin Dapp.
#ifndef _GLIBCXX_PARALLEL_UNIQUE_COPY_H
#define _GLIBCXX_PARALLEL_UNIQUE_COPY_H 1
#include <parallel/parallel.h>
#include <parallel/multiseq_selection.h>
namespace __gnu_parallel
{
/** @brief Parallel std::unique_copy(), w/__o explicit equality predicate.
* @param __first Begin iterator of input sequence.
* @param __last End iterator of input sequence.
* @param __result Begin iterator of result __sequence.
* @param __binary_pred Equality predicate.
* @return End iterator of result __sequence. */
template<typename _IIter,
class _OutputIterator,
class _BinaryPredicate>
_OutputIterator
__parallel_unique_copy(_IIter __first, _IIter __last,
_OutputIterator __result, _BinaryPredicate __binary_pred)
{
_GLIBCXX_CALL(__last - __first)
typedef std::iterator_traits<_IIter> _TraitsType;
typedef typename _TraitsType::value_type _ValueType;
typedef typename _TraitsType::difference_type _DifferenceType;
_DifferenceType size = __last - __first;
if (size == 0)
return __result;
// Let the first thread process two parts.
_DifferenceType *__counter;
_DifferenceType *__borders;
_ThreadIndex __num_threads = __get_max_threads();
// First part contains at least one element.
# pragma omp parallel num_threads(__num_threads)
{
# pragma omp single
{
__num_threads = omp_get_num_threads();
__borders = new _DifferenceType[__num_threads + 2];
equally_split(size, __num_threads + 1, __borders);
__counter = new _DifferenceType[__num_threads + 1];
}
_ThreadIndex __iam = omp_get_thread_num();
_DifferenceType __begin, __end;
// Check for length without duplicates
// Needed for position in output
_DifferenceType __i = 0;
_OutputIterator __out = __result;
if (__iam == 0)
{
__begin = __borders[0] + 1; // == 1
__end = __borders[__iam + 1];
++__i;
*__out++ = *__first;
for (_IIter iter = __first + __begin; iter < __first + __end; ++iter)
{
if (!__binary_pred(*iter, *(iter-1)))
{
++__i;
*__out++ = *iter;
}
}
}
else
{
__begin = __borders[__iam]; //one part
__end = __borders[__iam + 1];
for (_IIter iter = __first + __begin; iter < __first + __end; ++iter)
{
if (!__binary_pred(*iter, *(iter - 1)))
++__i;
}
}
__counter[__iam] = __i;
// Last part still untouched.
_DifferenceType __begin_output;
# pragma omp barrier
// Store result in output on calculated positions.
__begin_output = 0;
if (__iam == 0)
{
for (int __t = 0; __t < __num_threads; ++__t)
__begin_output += __counter[__t];
__i = 0;
_OutputIterator __iter_out = __result + __begin_output;
__begin = __borders[__num_threads];
__end = size;
for (_IIter iter = __first + __begin; iter < __first + __end; ++iter)
{
if (iter == __first || !__binary_pred(*iter, *(iter - 1)))
{
++__i;
*__iter_out++ = *iter;
}
}
__counter[__num_threads] = __i;
}
else
{
for (int __t = 0; __t < __iam; __t++)
__begin_output += __counter[__t];
_OutputIterator __iter_out = __result + __begin_output;
for (_IIter iter = __first + __begin; iter < __first + __end; ++iter)
{
if (!__binary_pred(*iter, *(iter-1)))
*__iter_out++ = *iter;
}
}
}
_DifferenceType __end_output = 0;
for (int __t = 0; __t < __num_threads + 1; __t++)
__end_output += __counter[__t];
delete[] __borders;
return __result + __end_output;
}
/** @brief Parallel std::unique_copy(), without explicit equality predicate
* @param __first Begin iterator of input sequence.
* @param __last End iterator of input sequence.
* @param __result Begin iterator of result __sequence.
* @return End iterator of result __sequence. */
template<typename _IIter, class _OutputIterator>
inline _OutputIterator
__parallel_unique_copy(_IIter __first, _IIter __last,
_OutputIterator __result)
{
typedef typename std::iterator_traits<_IIter>::value_type
_ValueType;
return __parallel_unique_copy(__first, __last, __result,
std::equal_to<_ValueType>());
}
}//namespace __gnu_parallel
#endif /* _GLIBCXX_PARALLEL_UNIQUE_COPY_H */
|
