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 __FBSDID("$FreeBSD$"); 27 28 #include <string.h> 29 #include <stdint.h> 30 31 static char *twobyte_memmem(const unsigned char *h, size_t k, const unsigned char *n) 32 { 33 uint16_t nw = n[0]<<8 | n[1], hw = h[0]<<8 | h[1]; 34 for (h+=2, k-=2; k; k--, hw = hw<<8 | *h++) 35 if (hw == nw) return (char *)h-2; 36 return hw == nw ? (char *)h-2 : 0; 37 } 38 39 static char *threebyte_memmem(const unsigned char *h, size_t k, const unsigned char *n) 40 { 41 uint32_t nw = n[0]<<24 | n[1]<<16 | n[2]<<8; 42 uint32_t hw = h[0]<<24 | h[1]<<16 | h[2]<<8; 43 for (h+=3, k-=3; k; k--, hw = (hw|*h++)<<8) 44 if (hw == nw) return (char *)h-3; 45 return hw == nw ? (char *)h-3 : 0; 46 } 47 48 static char *fourbyte_memmem(const unsigned char *h, size_t k, const unsigned char *n) 49 { 50 uint32_t nw = n[0]<<24 | n[1]<<16 | n[2]<<8 | n[3]; 51 uint32_t hw = h[0]<<24 | h[1]<<16 | h[2]<<8 | h[3]; 52 for (h+=4, k-=4; k; k--, hw = hw<<8 | *h++) 53 if (hw == nw) return (char *)h-4; 54 return hw == nw ? (char *)h-4 : 0; 55 } 56 57 #define MAX(a,b) ((a)>(b)?(a):(b)) 58 #define MIN(a,b) ((a)<(b)?(a):(b)) 59 60 #define BITOP(a,b,op) \ 61 ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a)))) 62 63 /* 64 * Two Way string search algorithm, with a bad shift table applied to the last 65 * byte of the window. A bit array marks which entries in the shift table are 66 * initialized to avoid fully initializing a 1kb/2kb table. 67 * 68 * Reference: CROCHEMORE M., PERRIN D., 1991, Two-way string-matching, 69 * Journal of the ACM 38(3):651-675 70 */ 71 static char *twoway_memmem(const unsigned char *h, const unsigned char *z, const unsigned char *n, size_t l) 72 { 73 size_t i, ip, jp, k, p, ms, p0, mem, mem0; 74 size_t byteset[32 / sizeof(size_t)] = { 0 }; 75 size_t shift[256]; 76 77 /* Computing length of needle and fill shift table */ 78 for (i=0; i<l; i++) 79 BITOP(byteset, n[i], |=), shift[n[i]] = i+1; 80 81 /* Compute maximal suffix */ 82 ip = -1; jp = 0; k = p = 1; 83 while (jp+k<l) { 84 if (n[ip+k] == n[jp+k]) { 85 if (k == p) { 86 jp += p; 87 k = 1; 88 } else k++; 89 } else if (n[ip+k] > n[jp+k]) { 90 jp += k; 91 k = 1; 92 p = jp - ip; 93 } else { 94 ip = jp++; 95 k = p = 1; 96 } 97 } 98 ms = ip; 99 p0 = p; 100 101 /* And with the opposite comparison */ 102 ip = -1; jp = 0; k = p = 1; 103 while (jp+k<l) { 104 if (n[ip+k] == n[jp+k]) { 105 if (k == p) { 106 jp += p; 107 k = 1; 108 } else k++; 109 } else if (n[ip+k] < n[jp+k]) { 110 jp += k; 111 k = 1; 112 p = jp - ip; 113 } else { 114 ip = jp++; 115 k = p = 1; 116 } 117 } 118 if (ip+1 > ms+1) ms = ip; 119 else p = p0; 120 121 /* Periodic needle? */ 122 if (memcmp(n, n+p, ms+1)) { 123 mem0 = 0; 124 p = MAX(ms, l-ms-1) + 1; 125 } else mem0 = l-p; 126 mem = 0; 127 128 /* Search loop */ 129 for (;;) { 130 /* If remainder of haystack is shorter than needle, done */ 131 if (z-h < l) return 0; 132 133 /* Check last byte first; advance by shift on mismatch */ 134 if (BITOP(byteset, h[l-1], &)) { 135 k = l-shift[h[l-1]]; 136 if (k) { 137 if (mem0 && mem && k < p) k = l-p; 138 h += k; 139 mem = 0; 140 continue; 141 } 142 } else { 143 h += l; 144 mem = 0; 145 continue; 146 } 147 148 /* Compare right half */ 149 for (k=MAX(ms+1,mem); k<l && n[k] == h[k]; k++); 150 if (k < l) { 151 h += k-ms; 152 mem = 0; 153 continue; 154 } 155 /* Compare left half */ 156 for (k=ms+1; k>mem && n[k-1] == h[k-1]; k--); 157 if (k <= mem) return (char *)h; 158 h += p; 159 mem = mem0; 160 } 161 } 162 163 void *memmem(const void *h0, size_t k, const void *n0, size_t l) 164 { 165 const unsigned char *h = h0, *n = n0; 166 167 /* Return immediately on empty needle */ 168 if (!l) return (void *)h; 169 170 /* Return immediately when needle is longer than haystack */ 171 if (k<l) return 0; 172 173 /* Use faster algorithms for short needles */ 174 h = memchr(h0, *n, k); 175 if (!h || l==1) return (void *)h; 176 k -= h - (const unsigned char *)h0; 177 if (k<l) return 0; 178 if (l==2) return twobyte_memmem(h, k, n); 179 if (l==3) return threebyte_memmem(h, k, n); 180 if (l==4) return fourbyte_memmem(h, k, n); 181 182 return twoway_memmem(h, h+k, n, l); 183 } 184