1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 3 * 4 * Copyright (c) 1991, 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * This code is derived from software contributed to Berkeley by 8 * David Hitz of Auspex Systems, Inc. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. Neither the name of the University nor the names of its contributors 19 * may be used to endorse or promote products derived from this software 20 * without specific prior written permission. 21 * 22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 32 * SUCH DAMAGE. 33 */ 34 35 /* 36 * look -- find lines in a sorted list. 37 * 38 * The man page said that TABs and SPACEs participate in -d comparisons. 39 * In fact, they were ignored. This implements historic practice, not 40 * the manual page. 41 */ 42 43 #include <sys/types.h> 44 #include <sys/mman.h> 45 #include <sys/stat.h> 46 47 #include <err.h> 48 #include <errno.h> 49 #include <fcntl.h> 50 #include <getopt.h> 51 #include <limits.h> 52 #include <locale.h> 53 #include <stdint.h> 54 #include <stdio.h> 55 #include <stdlib.h> 56 #include <string.h> 57 #include <unistd.h> 58 #include <wchar.h> 59 #include <wctype.h> 60 61 #include "pathnames.h" 62 63 static char _path_words[] = _PATH_WORDS; 64 65 #define EQUAL 0 66 #define GREATER 1 67 #define LESS (-1) 68 69 static int dflag, fflag; 70 71 static char *binary_search(wchar_t *, unsigned char *, unsigned char *); 72 static int compare(wchar_t *, unsigned char *, unsigned char *); 73 static char *linear_search(wchar_t *, unsigned char *, unsigned char *); 74 static int look(wchar_t *, unsigned char *, unsigned char *); 75 static wchar_t *prepkey(const char *, wchar_t); 76 static void print_from(wchar_t *, unsigned char *, unsigned char *); 77 78 static void usage(void) __dead2; 79 80 static struct option longopts[] = { 81 { "alternative",no_argument, NULL, 'a' }, 82 { "alphanum", no_argument, NULL, 'd' }, 83 { "ignore-case",no_argument, NULL, 'i' }, 84 { "terminate", required_argument, NULL, 't'}, 85 { NULL, 0, NULL, 0 }, 86 }; 87 88 int 89 main(int argc, char *argv[]) 90 { 91 struct stat sb; 92 int ch, fd, match; 93 wchar_t termchar; 94 unsigned char *back, *front; 95 unsigned const char *file; 96 wchar_t *key; 97 98 (void) setlocale(LC_CTYPE, ""); 99 100 file = _path_words; 101 termchar = L'\0'; 102 while ((ch = getopt_long(argc, argv, "+adft:", longopts, NULL)) != -1) 103 switch(ch) { 104 case 'a': 105 /* COMPATIBILITY */ 106 break; 107 case 'd': 108 dflag = 1; 109 break; 110 case 'f': 111 fflag = 1; 112 break; 113 case 't': 114 if (mbrtowc(&termchar, optarg, MB_LEN_MAX, NULL) != 115 strlen(optarg)) 116 errx(2, "invalid termination character"); 117 break; 118 case '?': 119 default: 120 usage(); 121 } 122 argc -= optind; 123 argv += optind; 124 125 if (argc == 0) 126 usage(); 127 if (argc == 1) /* But set -df by default. */ 128 dflag = fflag = 1; 129 key = prepkey(*argv++, termchar); 130 if (argc >= 2) 131 file = *argv++; 132 133 match = 1; 134 135 do { 136 if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb)) 137 err(2, "%s", file); 138 if ((uintmax_t)sb.st_size > (uintmax_t)SIZE_T_MAX) 139 errx(2, "%s: %s", file, strerror(EFBIG)); 140 if (sb.st_size == 0) { 141 close(fd); 142 continue; 143 } 144 if ((front = mmap(NULL, (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED) 145 err(2, "%s", file); 146 back = front + sb.st_size; 147 match *= (look(key, front, back)); 148 close(fd); 149 } while (argc-- > 2 && (file = *argv++)); 150 151 exit(match); 152 } 153 154 static wchar_t * 155 prepkey(const char *string, wchar_t termchar) 156 { 157 const char *readp; 158 wchar_t *key, *writep; 159 wchar_t ch; 160 size_t clen; 161 162 /* 163 * Reformat search string and convert to wide character representation 164 * to avoid doing it multiple times later. 165 */ 166 if ((key = malloc(sizeof(wchar_t) * (strlen(string) + 1))) == NULL) 167 err(2, NULL); 168 readp = string; 169 writep = key; 170 while ((clen = mbrtowc(&ch, readp, MB_LEN_MAX, NULL)) != 0) { 171 if (clen == (size_t)-1 || clen == (size_t)-2) 172 errc(2, EILSEQ, NULL); 173 if (fflag) 174 ch = towlower(ch); 175 if (!dflag || iswalnum(ch)) 176 *writep++ = ch; 177 readp += clen; 178 } 179 *writep = L'\0'; 180 if (termchar != L'\0' && (writep = wcschr(key, termchar)) != NULL) 181 *++writep = L'\0'; 182 return (key); 183 } 184 185 static int 186 look(wchar_t *string, unsigned char *front, unsigned char *back) 187 { 188 189 front = binary_search(string, front, back); 190 front = linear_search(string, front, back); 191 192 if (front) 193 print_from(string, front, back); 194 return (front ? 0 : 1); 195 } 196 197 198 /* 199 * Binary search for "string" in memory between "front" and "back". 200 * 201 * This routine is expected to return a pointer to the start of a line at 202 * *or before* the first word matching "string". Relaxing the constraint 203 * this way simplifies the algorithm. 204 * 205 * Invariants: 206 * front points to the beginning of a line at or before the first 207 * matching string. 208 * 209 * back points to the beginning of a line at or after the first 210 * matching line. 211 * 212 * Base of the Invariants. 213 * front = NULL; 214 * back = EOF; 215 * 216 * Advancing the Invariants: 217 * 218 * p = first newline after halfway point from front to back. 219 * 220 * If the string at "p" is not greater than the string to match, 221 * p is the new front. Otherwise it is the new back. 222 * 223 * Termination: 224 * 225 * The definition of the routine allows it return at any point, 226 * since front is always at or before the line to print. 227 * 228 * In fact, it returns when the chosen "p" equals "back". This 229 * implies that there exists a string is least half as long as 230 * (back - front), which in turn implies that a linear search will 231 * be no more expensive than the cost of simply printing a string or two. 232 * 233 * Trying to continue with binary search at this point would be 234 * more trouble than it's worth. 235 */ 236 #define SKIP_PAST_NEWLINE(p, back) \ 237 while (p < back && *p++ != '\n'); 238 239 static char * 240 binary_search(wchar_t *string, unsigned char *front, unsigned char *back) 241 { 242 unsigned char *p; 243 244 p = front + (back - front) / 2; 245 SKIP_PAST_NEWLINE(p, back); 246 247 /* 248 * If the file changes underneath us, make sure we don't 249 * infinitely loop. 250 */ 251 while (p < back && back > front) { 252 if (compare(string, p, back) == GREATER) 253 front = p; 254 else 255 back = p; 256 p = front + (back - front) / 2; 257 SKIP_PAST_NEWLINE(p, back); 258 } 259 return (front); 260 } 261 262 /* 263 * Find the first line that starts with string, linearly searching from front 264 * to back. 265 * 266 * Return NULL for no such line. 267 * 268 * This routine assumes: 269 * 270 * o front points at the first character in a line. 271 * o front is before or at the first line to be printed. 272 */ 273 static char * 274 linear_search(wchar_t *string, unsigned char *front, unsigned char *back) 275 { 276 while (front < back) { 277 switch (compare(string, front, back)) { 278 case EQUAL: /* Found it. */ 279 return (front); 280 case LESS: /* No such string. */ 281 return (NULL); 282 case GREATER: /* Keep going. */ 283 break; 284 } 285 SKIP_PAST_NEWLINE(front, back); 286 } 287 return (NULL); 288 } 289 290 /* 291 * Print as many lines as match string, starting at front. 292 */ 293 static void 294 print_from(wchar_t *string, unsigned char *front, unsigned char *back) 295 { 296 for (; front < back && compare(string, front, back) == EQUAL; ++front) { 297 for (; front < back && *front != '\n'; ++front) 298 if (putchar(*front) == EOF) 299 err(2, "stdout"); 300 if (putchar('\n') == EOF) 301 err(2, "stdout"); 302 } 303 } 304 305 /* 306 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with 307 * string2 (s1 ??? s2). 308 * 309 * o Matches up to len(s1) are EQUAL. 310 * o Matches up to len(s2) are GREATER. 311 * 312 * Compare understands about the -f and -d flags, and treats comparisons 313 * appropriately. 314 * 315 * The string "s1" is null terminated. The string s2 is '\n' terminated (or 316 * "back" terminated). 317 */ 318 static int 319 compare(wchar_t *s1, unsigned char *s2, unsigned char *back) 320 { 321 wchar_t ch1, ch2; 322 size_t len2; 323 324 for (; *s1 && s2 < back && *s2 != '\n'; ++s1, s2 += len2) { 325 ch1 = *s1; 326 len2 = mbrtowc(&ch2, s2, back - s2, NULL); 327 if (len2 == (size_t)-1 || len2 == (size_t)-2) { 328 ch2 = *s2; 329 len2 = 1; 330 } 331 if (fflag) 332 ch2 = towlower(ch2); 333 if (dflag && !iswalnum(ch2)) { 334 /* Ignore character in comparison. */ 335 --s1; 336 continue; 337 } 338 if (ch1 != ch2) 339 return (ch1 < ch2 ? LESS : GREATER); 340 } 341 return (*s1 ? GREATER : EQUAL); 342 } 343 344 static void 345 usage(void) 346 { 347 (void)fprintf(stderr, "usage: look [-df] [-t char] string [file ...]\n"); 348 exit(2); 349 } 350