2 * Copyright (c) 2005-2014 Rich Felker, et al.
4 * Permission is hereby granted, free of charge, to any person obtaining
5 * a copy of this software and associated documentation files (the
6 * "Software"), to deal in the Software without restriction, including
7 * without limitation the rights to use, copy, modify, merge, publish,
8 * distribute, sublicense, and/or sell copies of the Software, and to
9 * permit persons to whom the Software is furnished to do so, subject to
10 * the following conditions:
12 * The above copyright notice and this permission notice shall be
13 * included in all copies or substantial portions of the Software.
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
16 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
17 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
18 * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
19 * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
20 * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
21 * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
23 #include <sys/cdefs.h>
24 __FBSDID("$FreeBSD$");
29 static char *twobyte_memmem(const unsigned char *h, size_t k, const unsigned char *n)
31 uint16_t nw = n[0]<<8 | n[1], hw = h[0]<<8 | h[1];
32 for (h++, k--; k; k--, hw = hw<<8 | *++h)
33 if (hw == nw) return (char *)h-1;
37 static char *threebyte_memmem(const unsigned char *h, size_t k, const unsigned char *n)
39 uint32_t nw = n[0]<<24 | n[1]<<16 | n[2]<<8;
40 uint32_t hw = h[0]<<24 | h[1]<<16 | h[2]<<8;
41 for (h+=2, k-=2; k; k--, hw = (hw|*++h)<<8)
42 if (hw == nw) return (char *)h-2;
46 static char *fourbyte_memmem(const unsigned char *h, size_t k, const unsigned char *n)
48 uint32_t nw = n[0]<<24 | n[1]<<16 | n[2]<<8 | n[3];
49 uint32_t hw = h[0]<<24 | h[1]<<16 | h[2]<<8 | h[3];
50 for (h+=3, k-=3; k; k--, hw = hw<<8 | *++h)
51 if (hw == nw) return (char *)h-3;
55 #define MAX(a,b) ((a)>(b)?(a):(b))
56 #define MIN(a,b) ((a)<(b)?(a):(b))
58 #define BITOP(a,b,op) \
59 ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a))))
62 * Two Way string search algorithm, with a bad shift table applied to the last
63 * byte of the window. A bit array marks which entries in the shift table are
64 * initialized to avoid fully initializing a 1kb/2kb table.
66 * Reference: CROCHEMORE M., PERRIN D., 1991, Two-way string-matching,
67 * Journal of the ACM 38(3):651-675
69 static char *twoway_memmem(const unsigned char *h, const unsigned char *z, const unsigned char *n, size_t l)
71 size_t i, ip, jp, k, p, ms, p0, mem, mem0;
72 size_t byteset[32 / sizeof(size_t)] = { 0 };
75 /* Computing length of needle and fill shift table */
77 BITOP(byteset, n[i], |=), shift[n[i]] = i+1;
79 /* Compute maximal suffix */
80 ip = -1; jp = 0; k = p = 1;
82 if (n[ip+k] == n[jp+k]) {
87 } else if (n[ip+k] > n[jp+k]) {
99 /* And with the opposite comparison */
100 ip = -1; jp = 0; k = p = 1;
102 if (n[ip+k] == n[jp+k]) {
107 } else if (n[ip+k] < n[jp+k]) {
116 if (ip+1 > ms+1) ms = ip;
119 /* Periodic needle? */
120 if (memcmp(n, n+p, ms+1)) {
122 p = MAX(ms, l-ms-1) + 1;
128 /* If remainder of haystack is shorter than needle, done */
129 if (z-h < l) return 0;
131 /* Check last byte first; advance by shift on mismatch */
132 if (BITOP(byteset, h[l-1], &)) {
135 if (mem0 && mem && k < p) k = l-p;
146 /* Compare right half */
147 for (k=MAX(ms+1,mem); k<l && n[k] == h[k]; k++);
153 /* Compare left half */
154 for (k=ms+1; k>mem && n[k-1] == h[k-1]; k--);
155 if (k <= mem) return (char *)h;
161 void *memmem(const void *h0, size_t k, const void *n0, size_t l)
163 const unsigned char *h = h0, *n = n0;
165 /* Return immediately on empty needle */
166 if (!l) return (void *)h;
168 /* Return immediately when needle is longer than haystack */
171 /* Use faster algorithms for short needles */
172 h = memchr(h0, *n, k);
173 if (!h || l==1) return (void *)h;
174 k -= h - (const unsigned char *)h0;
176 if (l==2) return twobyte_memmem(h, k, n);
177 if (l==3) return threebyte_memmem(h, k, n);
178 if (l==4) return fourbyte_memmem(h, k, n);
180 return twoway_memmem(h, h+k, n, l);