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
193
194
195
196
197
198
199
200
201
202
|
#include "sdsl/construct_sa_se.hpp"
namespace sdsl
{
void _construct_sa_IS(int_vector<> &text, int_vector<> &sa, std::string& filename_sa, size_t n, size_t text_offset, size_t sigma, uint64_t recursion)
{
uint64_t buffersize = 1024*1024/8;
size_t name = 0;
size_t number_of_lms_strings = 0;
std::string filename_c_array = tmp_file(filename_sa, "_c_array"+util::to_string(recursion));
// Phase 1
{
std::vector<uint64_t> bkt(sigma, 0);
// Step 1 - Count characters into c array
// TODO: better create this in higher recursion-level
for (size_t i=0; i<n; ++i) {
++bkt[text[text_offset+i]];
}
// Step 1.5 save them into cached_external_array
int_vector_buffer<> c_array(filename_c_array, std::ios::out, buffersize, 64);
for (size_t c=0; c<sigma; ++c) {
c_array[c] = bkt[c];
}
// Step 2 Calculate End-Pointer of Buckets
bkt[0] = 0;
for (size_t c=1; c<sigma; ++c) {
bkt[c] = bkt[c-1]+bkt[c];
}
// Step 3 - Insert S*-positions into correct bucket of SA but not in correct order inside the buckets
for (size_t i=n-2, was_s_typ = 1; i<n; --i) {
if (text[text_offset+i]>text[text_offset+i+1]) {
if (was_s_typ) {
sa[bkt[text[text_offset+i+1]]--] = i+1;
++number_of_lms_strings;
was_s_typ = 0;
}
} else if (text[text_offset+i]<text[text_offset+i+1]) {
was_s_typ = 1;
}
}
// Step 4 - Calculate Begin-Pointer of Buckets
bkt[0] = 0;
for (size_t c=1; c<sigma; ++c) {
bkt[c] = bkt[c-1]+c_array[c-1];
}
// Step 5 - Scan from Left-To-Right to induce L-Types
for (size_t i=0; i<n; ++i) {
if (sa[i] > 0 and text[text_offset+ sa[i] ] <= text[text_offset+ sa[i]-1 ]) { // faster than if(sa[i]>0 and bkt_beg[text[ sa[i]-1 ]] > i)
sa[bkt[text[text_offset+ sa[i]-1 ]]++] = sa[i]-1;
sa[i] = 0;
}
}
// Step 6 - Scan from Right-To-Left to induce S-Types
bkt[0] = 0;
for (size_t c=1; c<sigma; ++c) {
bkt[c] = bkt[c-1]+c_array[c];
}
c_array.close();
c_array.buffersize(0);
for (size_t i=n-1, endpointer=n; i<n; --i) {
if (sa[i]>0) {
if (text[text_offset+ sa[i]-1 ] <= text[text_offset+ sa[i] ]) { // faster than if(bkt_end[text[ sa[i]-1 ]] < i)
sa[bkt[text[text_offset+ sa[i]-1 ]]--] = sa[i]-1;
} else {
sa[--endpointer] = sa[i];
}
sa[i] = 0;
}
}
// Step 7 - Determine length of LMS-Strings
for (size_t i=n-2, end=n-2, was_s_typ = 1; i<n; --i) {
if (text[text_offset+i]>text[text_offset+i+1]) {
if (was_s_typ) {
sa[(i+1)>>1] = end-i;
end = i+1;
was_s_typ = 0;
}
} else if (text[text_offset+i]<text[text_offset+i+1]) {
was_s_typ = 1;
}
}
// Step 8 - Rename
for (size_t i=n-number_of_lms_strings+1, cur_pos=0, cur_len=0, last_pos=n-1, last_len=1; i<n; ++i) {
cur_pos = sa[i];
cur_len = sa[(cur_pos>>1)];
if (cur_len == last_len) {
size_t l = 0;
while (l < cur_len and text[text_offset+cur_pos+l] == text[text_offset+last_pos+l]) {
++l;
}
if (l >= cur_len) {
--name;
}
}
sa[(cur_pos>>1)] = ++name;
last_pos = cur_pos;
last_len = cur_len;
}
}
// Step 9 - Calculate SA of new string - in most cases recursive
if (name+1 < number_of_lms_strings) {
// Move Names to the end
for (size_t i=0, t=n-number_of_lms_strings; i<(n>>1); ++i) {
if (sa[i] > 0) {
sa[t++] = sa[i];
sa[i] = 0;
}
}
sa[n-1] = 0;
// Recursive call
std::string filename_sa_rec = tmp_file(filename_sa, "_sa_rec"+util::to_string(recursion+1));
_construct_sa_IS(sa, sa, filename_sa_rec, number_of_lms_strings, n-number_of_lms_strings, name+1, recursion+1);
for (size_t i=n-2, endpointer = n-1, was_s_typ = 1; i<n; --i) {
if (text[text_offset+i]>text[text_offset+i+1]) {
if (was_s_typ) {
sa[endpointer--] = i+1;
was_s_typ = 0;
}
} else if (text[text_offset+i]<text[text_offset+i+1]) {
was_s_typ = 1;
}
}
// Sort S*-positions in correct order into SA
for (size_t i=0; i<number_of_lms_strings; ++i) {
size_t pos = sa[i];
sa[i] = sa[n-number_of_lms_strings+pos];
sa[n-number_of_lms_strings+pos] = 0;
}
} else {
// Move s*-Positions to front
sa[0] = n-1;
for (size_t i=1; i<number_of_lms_strings; ++i) {
sa[i] = sa[n-number_of_lms_strings+i];
sa[n-number_of_lms_strings+i] = 0;
}
// Clear lex. names
for (size_t i=number_of_lms_strings; i<(n>>1); ++i) {
sa[i] = 0;
}
}
// Phase 3
{
// Step 10 - Count characters into c array
// Step 11 - Calculate End-Pointer of Buckets
int_vector_buffer<> c_array(filename_c_array, std::ios::in, buffersize, 64);
std::vector<uint64_t> bkt(sigma, 0);
for (size_t c=1; c<sigma; ++c) {
bkt[c] = bkt[c-1]+c_array[c];
}
// Step 12 - Move S*-positions in correct order into SA
for (size_t i=number_of_lms_strings-1; i<n; --i) {
size_t pos = sa[i];
sa[i] = 0;
sa[ bkt[text[text_offset+pos]]-- ] = pos;
}
// Step 13 - Calculate Begin-Pointer of Buckets
bkt[0] = 0;
for (size_t c=1; c<sigma; ++c) {
bkt[c] = bkt[c-1]+c_array[c-1];
}
// Step 14 - Scan from Left-To-Right to induce L-Types
for (size_t i=0; i<n; ++i) {
if (sa[i] > 0 and text[text_offset+ sa[i] ] <= text[text_offset+ sa[i]-1 ]) { // faster than if(sa[i]>0 and bkt_beg[text[ sa[i]-1 ]] > i)
sa[bkt[text[text_offset+ sa[i]-1 ]]++] = sa[i]-1;
}
}
// Step 15 - Scan from Right-To-Left to induce S-Types
bkt[0] = 0;
for (size_t c=1; c<sigma; ++c) {
bkt[c] = bkt[c-1]+c_array[c];
}
for (size_t i=n-1; i<n; --i) {
if (sa[i] > 0 and text[text_offset+sa[i]-1] <= text[text_offset+sa[i]]) {
sa[bkt[text[text_offset+ sa[i]-1 ]]--] = sa[i]-1;
}
}
c_array.close(true);
}
}
}
|
