1 /*- 2 * Copyright (c) 2005-2014 Rich Felker, et al. 3 * 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: 11 * 12 * The above copyright notice and this permission notice shall be 13 * included in all copies or substantial portions of the Software. 14 * 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. 22 */ 23 #include <sys/cdefs.h> 24 __FBSDID("$FreeBSD$"); 25 26 #include <string.h> 27 #include <stdint.h> 28 29 static char *twobyte_memmem(const unsigned char *h, size_t k, const unsigned char *n) 30 { 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; 34 return 0; 35 } 36 37 static char *threebyte_memmem(const unsigned char *h, size_t k, const unsigned char *n) 38 { 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; 43 return 0; 44 } 45 46 static char *fourbyte_memmem(const unsigned char *h, size_t k, const unsigned char *n) 47 { 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; 52 return 0; 53 } 54 55 #define MAX(a,b) ((a)>(b)?(a):(b)) 56 #define MIN(a,b) ((a)<(b)?(a):(b)) 57 58 #define BITOP(a,b,op) \ 59 ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a)))) 60 61 /* 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. 65 * 66 * Reference: CROCHEMORE M., PERRIN D., 1991, Two-way string-matching, 67 * Journal of the ACM 38(3):651-675 68 */ 69 static char *twoway_memmem(const unsigned char *h, const unsigned char *z, const unsigned char *n, size_t l) 70 { 71 size_t i, ip, jp, k, p, ms, p0, mem, mem0; 72 size_t byteset[32 / sizeof(size_t)] = { 0 }; 73 size_t shift[256]; 74 75 /* Computing length of needle and fill shift table */ 76 for (i=0; i<l; i++) 77 BITOP(byteset, n[i], |=), shift[n[i]] = i+1; 78 79 /* Compute maximal suffix */ 80 ip = -1; jp = 0; k = p = 1; 81 while (jp+k<l) { 82 if (n[ip+k] == n[jp+k]) { 83 if (k == p) { 84 jp += p; 85 k = 1; 86 } else k++; 87 } else if (n[ip+k] > n[jp+k]) { 88 jp += k; 89 k = 1; 90 p = jp - ip; 91 } else { 92 ip = jp++; 93 k = p = 1; 94 } 95 } 96 ms = ip; 97 p0 = p; 98 99 /* And with the opposite comparison */ 100 ip = -1; jp = 0; k = p = 1; 101 while (jp+k<l) { 102 if (n[ip+k] == n[jp+k]) { 103 if (k == p) { 104 jp += p; 105 k = 1; 106 } else k++; 107 } else if (n[ip+k] < n[jp+k]) { 108 jp += k; 109 k = 1; 110 p = jp - ip; 111 } else { 112 ip = jp++; 113 k = p = 1; 114 } 115 } 116 if (ip+1 > ms+1) ms = ip; 117 else p = p0; 118 119 /* Periodic needle? */ 120 if (memcmp(n, n+p, ms+1)) { 121 mem0 = 0; 122 p = MAX(ms, l-ms-1) + 1; 123 } else mem0 = l-p; 124 mem = 0; 125 126 /* Search loop */ 127 for (;;) { 128 /* If remainder of haystack is shorter than needle, done */ 129 if (z-h < l) return 0; 130 131 /* Check last byte first; advance by shift on mismatch */ 132 if (BITOP(byteset, h[l-1], &)) { 133 k = l-shift[h[l-1]]; 134 if (k) { 135 if (mem0 && mem && k < p) k = l-p; 136 h += k; 137 mem = 0; 138 continue; 139 } 140 } else { 141 h += l; 142 mem = 0; 143 continue; 144 } 145 146 /* Compare right half */ 147 for (k=MAX(ms+1,mem); k<l && n[k] == h[k]; k++); 148 if (k < l) { 149 h += k-ms; 150 mem = 0; 151 continue; 152 } 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; 156 h += p; 157 mem = mem0; 158 } 159 } 160 161 void *memmem(const void *h0, size_t k, const void *n0, size_t l) 162 { 163 const unsigned char *h = h0, *n = n0; 164 165 /* Return immediately on empty needle */ 166 if (!l) return (void *)h; 167 168 /* Return immediately when needle is longer than haystack */ 169 if (k<l) return 0; 170 171 /* 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; 175 if (k<l) return 0; 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); 179 180 return twoway_memmem(h, h+k, n, l); 181 } 182