1 /*-
2 * SPDX-License-Identifier: MIT
3 *
4 * Copyright (c) 2005-2014 Rich Felker, et al.
5 *
6 * Permission is hereby granted, free of charge, to any person obtaining
7 * a copy of this software and associated documentation files (the
8 * "Software"), to deal in the Software without restriction, including
9 * without limitation the rights to use, copy, modify, merge, publish,
10 * distribute, sublicense, and/or sell copies of the Software, and to
11 * permit persons to whom the Software is furnished to do so, subject to
12 * the following conditions:
13 *
14 * The above copyright notice and this permission notice shall be
15 * included in all copies or substantial portions of the Software.
16 *
17 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
18 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
19 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
20 * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
21 * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
22 * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
23 * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
24 */
25 #include <sys/cdefs.h>
26 #include <stdint.h>
27 #include <string.h>
28
29 static char *
twobyte_strstr(const unsigned char * h,const unsigned char * n)30 twobyte_strstr(const unsigned char *h, const unsigned char *n)
31 {
32 uint16_t nw = n[0] << 8 | n[1], hw = h[0] << 8 | h[1];
33 for (h++; *h && hw != nw; hw = hw << 8 | *++h)
34 ;
35 return *h ? (char *)h - 1 : 0;
36 }
37
38 static char *
threebyte_strstr(const unsigned char * h,const unsigned char * n)39 threebyte_strstr(const unsigned char *h, const unsigned char *n)
40 {
41 uint32_t nw = (uint32_t)n[0] << 24 | n[1] << 16 | n[2] << 8;
42 uint32_t hw = (uint32_t)h[0] << 24 | h[1] << 16 | h[2] << 8;
43 for (h += 2; *h && hw != nw; hw = (hw | *++h) << 8)
44 ;
45 return *h ? (char *)h - 2 : 0;
46 }
47
48 static char *
fourbyte_strstr(const unsigned char * h,const unsigned char * n)49 fourbyte_strstr(const unsigned char *h, const unsigned char *n)
50 {
51 uint32_t nw = (uint32_t)n[0] << 24 | n[1] << 16 | n[2] << 8 | n[3];
52 uint32_t hw = (uint32_t)h[0] << 24 | h[1] << 16 | h[2] << 8 | h[3];
53 for (h += 3; *h && hw != nw; hw = hw << 8 | *++h)
54 ;
55 return *h ? (char *)h - 3 : 0;
56 }
57
58 #define MAX(a, b) ((a) > (b) ? (a) : (b))
59 #define MIN(a, b) ((a) < (b) ? (a) : (b))
60
61 #define BITOP(a, b, op) \
62 ((a)[(size_t)(b) / (8 * sizeof *(a))] op \
63 (size_t)1 << ((size_t)(b) % (8 * sizeof *(a))))
64
65 /*
66 * Two Way string search algorithm, with a bad shift table applied to the last
67 * byte of the window. A bit array marks which entries in the shift table are
68 * initialized to avoid fully initializing a 1kb/2kb table.
69 *
70 * Reference: CROCHEMORE M., PERRIN D., 1991, Two-way string-matching,
71 * Journal of the ACM 38(3):651-675
72 */
73 static char *
twoway_strstr(const unsigned char * h,const unsigned char * n)74 twoway_strstr(const unsigned char *h, const unsigned char *n)
75 {
76 const unsigned char *z;
77 size_t l, ip, jp, k, p, ms, p0, mem, mem0;
78 size_t byteset[32 / sizeof(size_t)] = { 0 };
79 size_t shift[256];
80
81 /* Computing length of needle and fill shift table */
82 for (l = 0; n[l] && h[l]; l++)
83 BITOP(byteset, n[l], |=), shift[n[l]] = l + 1;
84 if (n[l])
85 return 0; /* hit the end of h */
86
87 /* Compute maximal suffix */
88 ip = -1;
89 jp = 0;
90 k = p = 1;
91 while (jp + k < l) {
92 if (n[ip + k] == n[jp + k]) {
93 if (k == p) {
94 jp += p;
95 k = 1;
96 } else
97 k++;
98 } else if (n[ip + k] > n[jp + k]) {
99 jp += k;
100 k = 1;
101 p = jp - ip;
102 } else {
103 ip = jp++;
104 k = p = 1;
105 }
106 }
107 ms = ip;
108 p0 = p;
109
110 /* And with the opposite comparison */
111 ip = -1;
112 jp = 0;
113 k = p = 1;
114 while (jp + k < l) {
115 if (n[ip + k] == n[jp + k]) {
116 if (k == p) {
117 jp += p;
118 k = 1;
119 } else
120 k++;
121 } else if (n[ip + k] < n[jp + k]) {
122 jp += k;
123 k = 1;
124 p = jp - ip;
125 } else {
126 ip = jp++;
127 k = p = 1;
128 }
129 }
130 if (ip + 1 > ms + 1)
131 ms = ip;
132 else
133 p = p0;
134
135 /* Periodic needle? */
136 if (memcmp(n, n + p, ms + 1)) {
137 mem0 = 0;
138 p = MAX(ms, l - ms - 1) + 1;
139 } else
140 mem0 = l - p;
141 mem = 0;
142
143 /* Initialize incremental end-of-haystack pointer */
144 z = h;
145
146 /* Search loop */
147 for (;;) {
148 /* Update incremental end-of-haystack pointer */
149 if (z - h < l) {
150 /* Fast estimate for MAX(l,63) */
151 size_t grow = l | 63;
152 const unsigned char *z2 = memchr(z, 0, grow);
153 if (z2) {
154 z = z2;
155 if (z - h < l)
156 return 0;
157 } else
158 z += grow;
159 }
160
161 /* Check last byte first; advance by shift on mismatch */
162 if (BITOP(byteset, h[l - 1], &)) {
163 k = l - shift[h[l - 1]];
164 if (k) {
165 if (k < mem)
166 k = mem;
167 h += k;
168 mem = 0;
169 continue;
170 }
171 } else {
172 h += l;
173 mem = 0;
174 continue;
175 }
176
177 /* Compare right half */
178 for (k = MAX(ms + 1, mem); n[k] && n[k] == h[k]; k++)
179 ;
180 if (n[k]) {
181 h += k - ms;
182 mem = 0;
183 continue;
184 }
185 /* Compare left half */
186 for (k = ms + 1; k > mem && n[k - 1] == h[k - 1]; k--)
187 ;
188 if (k <= mem)
189 return (char *)h;
190 h += p;
191 mem = mem0;
192 }
193 }
194
195 char *
strstr(const char * h,const char * n)196 strstr(const char *h, const char *n)
197 {
198 /* Return immediately on empty needle */
199 if (!n[0])
200 return (char *)h;
201
202 /* Use faster algorithms for short needles */
203 h = strchr(h, *n);
204 if (!h || !n[1])
205 return (char *)h;
206 if (!h[1])
207 return 0;
208 if (!n[2])
209 return twobyte_strstr((void *)h, (void *)n);
210 if (!h[2])
211 return 0;
212 if (!n[3])
213 return threebyte_strstr((void *)h, (void *)n);
214 if (!h[3])
215 return 0;
216 if (!n[4])
217 return fourbyte_strstr((void *)h, (void *)n);
218
219 return twoway_strstr((void *)h, (void *)n);
220 }
221