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_strstr(const unsigned char *h, const unsigned char *n) 30 { 31 uint16_t nw = n[0]<<8 | n[1], hw = h[0]<<8 | h[1]; 32 for (h++; *h && hw != nw; hw = hw<<8 | *++h); 33 return *h ? (char *)h-1 : 0; 34 } 35 36 static char *threebyte_strstr(const unsigned char *h, const unsigned char *n) 37 { 38 uint32_t nw = n[0]<<24 | n[1]<<16 | n[2]<<8; 39 uint32_t hw = h[0]<<24 | h[1]<<16 | h[2]<<8; 40 for (h+=2; *h && hw != nw; hw = (hw|*++h)<<8); 41 return *h ? (char *)h-2 : 0; 42 } 43 44 static char *fourbyte_strstr(const unsigned char *h, const unsigned char *n) 45 { 46 uint32_t nw = n[0]<<24 | n[1]<<16 | n[2]<<8 | n[3]; 47 uint32_t hw = h[0]<<24 | h[1]<<16 | h[2]<<8 | h[3]; 48 for (h+=3; *h && hw != nw; hw = hw<<8 | *++h); 49 return *h ? (char *)h-3 : 0; 50 } 51 52 #define MAX(a,b) ((a)>(b)?(a):(b)) 53 #define MIN(a,b) ((a)<(b)?(a):(b)) 54 55 #define BITOP(a,b,op) \ 56 ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a)))) 57 58 /* 59 * Two Way string search algorithm, with a bad shift table applied to the last 60 * byte of the window. A bit array marks which entries in the shift table are 61 * initialized to avoid fully initializing a 1kb/2kb table. 62 * 63 * Reference: CROCHEMORE M., PERRIN D., 1991, Two-way string-matching, 64 * Journal of the ACM 38(3):651-675 65 */ 66 static char *twoway_strstr(const unsigned char *h, const unsigned char *n) 67 { 68 const unsigned char *z; 69 size_t l, ip, jp, k, p, ms, p0, mem, mem0; 70 size_t byteset[32 / sizeof(size_t)] = { 0 }; 71 size_t shift[256]; 72 73 /* Computing length of needle and fill shift table */ 74 for (l=0; n[l] && h[l]; l++) 75 BITOP(byteset, n[l], |=), shift[n[l]] = l+1; 76 if (n[l]) return 0; /* hit the end of h */ 77 78 /* Compute maximal suffix */ 79 ip = -1; jp = 0; k = p = 1; 80 while (jp+k<l) { 81 if (n[ip+k] == n[jp+k]) { 82 if (k == p) { 83 jp += p; 84 k = 1; 85 } else k++; 86 } else if (n[ip+k] > n[jp+k]) { 87 jp += k; 88 k = 1; 89 p = jp - ip; 90 } else { 91 ip = jp++; 92 k = p = 1; 93 } 94 } 95 ms = ip; 96 p0 = p; 97 98 /* And with the opposite comparison */ 99 ip = -1; jp = 0; k = p = 1; 100 while (jp+k<l) { 101 if (n[ip+k] == n[jp+k]) { 102 if (k == p) { 103 jp += p; 104 k = 1; 105 } else k++; 106 } else if (n[ip+k] < n[jp+k]) { 107 jp += k; 108 k = 1; 109 p = jp - ip; 110 } else { 111 ip = jp++; 112 k = p = 1; 113 } 114 } 115 if (ip+1 > ms+1) ms = ip; 116 else p = p0; 117 118 /* Periodic needle? */ 119 if (memcmp(n, n+p, ms+1)) { 120 mem0 = 0; 121 p = MAX(ms, l-ms-1) + 1; 122 } else mem0 = l-p; 123 mem = 0; 124 125 /* Initialize incremental end-of-haystack pointer */ 126 z = h; 127 128 /* Search loop */ 129 for (;;) { 130 /* Update incremental end-of-haystack pointer */ 131 if (z-h < l) { 132 /* Fast estimate for MIN(l,63) */ 133 size_t grow = l | 63; 134 const unsigned char *z2 = memchr(z, 0, grow); 135 if (z2) { 136 z = z2; 137 if (z-h < l) return 0; 138 } else z += grow; 139 } 140 141 /* Check last byte first; advance by shift on mismatch */ 142 if (BITOP(byteset, h[l-1], &)) { 143 k = l-shift[h[l-1]]; 144 //printf("adv by %zu (on %c) at [%s] (%zu;l=%zu)\n", k, h[l-1], h, shift[h[l-1]], l); 145 if (k) { 146 if (mem0 && mem && k < p) k = l-p; 147 h += k; 148 mem = 0; 149 continue; 150 } 151 } else { 152 h += l; 153 mem = 0; 154 continue; 155 } 156 157 /* Compare right half */ 158 for (k=MAX(ms+1,mem); n[k] && n[k] == h[k]; k++); 159 if (n[k]) { 160 h += k-ms; 161 mem = 0; 162 continue; 163 } 164 /* Compare left half */ 165 for (k=ms+1; k>mem && n[k-1] == h[k-1]; k--); 166 if (k <= mem) return (char *)h; 167 h += p; 168 mem = mem0; 169 } 170 } 171 172 char *strstr(const char *h, const char *n) 173 { 174 /* Return immediately on empty needle */ 175 if (!n[0]) return (char *)h; 176 177 /* Use faster algorithms for short needles */ 178 h = strchr(h, *n); 179 if (!h || !n[1]) return (char *)h; 180 if (!h[1]) return 0; 181 if (!n[2]) return twobyte_strstr((void *)h, (void *)n); 182 if (!h[2]) return 0; 183 if (!n[3]) return threebyte_strstr((void *)h, (void *)n); 184 if (!h[3]) return 0; 185 if (!n[4]) return fourbyte_strstr((void *)h, (void *)n); 186 187 return twoway_strstr((void *)h, (void *)n); 188 } 189