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 #ifndef lint 36 static const char copyright[] = 37 "@(#) Copyright (c) 1991, 1993\n\ 38 The Regents of the University of California. All rights reserved.\n"; 39 #endif /* not lint */ 40 41 #ifndef lint 42 #endif /* not lint */ 43 #include <sys/cdefs.h> 44 /* 45 * look -- find lines in a sorted list. 46 * 47 * The man page said that TABs and SPACEs participate in -d comparisons. 48 * In fact, they were ignored. This implements historic practice, not 49 * the manual page. 50 */ 51 52 #include <sys/types.h> 53 #include <sys/mman.h> 54 #include <sys/stat.h> 55 56 #include <err.h> 57 #include <errno.h> 58 #include <fcntl.h> 59 #include <getopt.h> 60 #include <limits.h> 61 #include <locale.h> 62 #include <stdint.h> 63 #include <stdio.h> 64 #include <stdlib.h> 65 #include <string.h> 66 #include <unistd.h> 67 #include <wchar.h> 68 #include <wctype.h> 69 70 #include "pathnames.h" 71 72 static char _path_words[] = _PATH_WORDS; 73 74 #define EQUAL 0 75 #define GREATER 1 76 #define LESS (-1) 77 78 static int dflag, fflag; 79 80 static char *binary_search(wchar_t *, unsigned char *, unsigned char *); 81 static int compare(wchar_t *, unsigned char *, unsigned char *); 82 static char *linear_search(wchar_t *, unsigned char *, unsigned char *); 83 static int look(wchar_t *, unsigned char *, unsigned char *); 84 static wchar_t *prepkey(const char *, wchar_t); 85 static void print_from(wchar_t *, unsigned char *, unsigned char *); 86 87 static void usage(void) __dead2; 88 89 static struct option longopts[] = { 90 { "alternative",no_argument, NULL, 'a' }, 91 { "alphanum", no_argument, NULL, 'd' }, 92 { "ignore-case",no_argument, NULL, 'i' }, 93 { "terminate", required_argument, NULL, 't'}, 94 { NULL, 0, NULL, 0 }, 95 }; 96 97 int 98 main(int argc, char *argv[]) 99 { 100 struct stat sb; 101 int ch, fd, match; 102 wchar_t termchar; 103 unsigned char *back, *front; 104 unsigned const char *file; 105 wchar_t *key; 106 107 (void) setlocale(LC_CTYPE, ""); 108 109 file = _path_words; 110 termchar = L'\0'; 111 while ((ch = getopt_long(argc, argv, "+adft:", longopts, NULL)) != -1) 112 switch(ch) { 113 case 'a': 114 /* COMPATIBILITY */ 115 break; 116 case 'd': 117 dflag = 1; 118 break; 119 case 'f': 120 fflag = 1; 121 break; 122 case 't': 123 if (mbrtowc(&termchar, optarg, MB_LEN_MAX, NULL) != 124 strlen(optarg)) 125 errx(2, "invalid termination character"); 126 break; 127 case '?': 128 default: 129 usage(); 130 } 131 argc -= optind; 132 argv += optind; 133 134 if (argc == 0) 135 usage(); 136 if (argc == 1) /* But set -df by default. */ 137 dflag = fflag = 1; 138 key = prepkey(*argv++, termchar); 139 if (argc >= 2) 140 file = *argv++; 141 142 match = 1; 143 144 do { 145 if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb)) 146 err(2, "%s", file); 147 if ((uintmax_t)sb.st_size > (uintmax_t)SIZE_T_MAX) 148 errx(2, "%s: %s", file, strerror(EFBIG)); 149 if (sb.st_size == 0) { 150 close(fd); 151 continue; 152 } 153 if ((front = mmap(NULL, (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED) 154 err(2, "%s", file); 155 back = front + sb.st_size; 156 match *= (look(key, front, back)); 157 close(fd); 158 } while (argc-- > 2 && (file = *argv++)); 159 160 exit(match); 161 } 162 163 static wchar_t * 164 prepkey(const char *string, wchar_t termchar) 165 { 166 const char *readp; 167 wchar_t *key, *writep; 168 wchar_t ch; 169 size_t clen; 170 171 /* 172 * Reformat search string and convert to wide character representation 173 * to avoid doing it multiple times later. 174 */ 175 if ((key = malloc(sizeof(wchar_t) * (strlen(string) + 1))) == NULL) 176 err(2, NULL); 177 readp = string; 178 writep = key; 179 while ((clen = mbrtowc(&ch, readp, MB_LEN_MAX, NULL)) != 0) { 180 if (clen == (size_t)-1 || clen == (size_t)-2) 181 errc(2, EILSEQ, NULL); 182 if (fflag) 183 ch = towlower(ch); 184 if (!dflag || iswalnum(ch)) 185 *writep++ = ch; 186 readp += clen; 187 } 188 *writep = L'\0'; 189 if (termchar != L'\0' && (writep = wcschr(key, termchar)) != NULL) 190 *++writep = L'\0'; 191 return (key); 192 } 193 194 static int 195 look(wchar_t *string, unsigned char *front, unsigned char *back) 196 { 197 198 front = binary_search(string, front, back); 199 front = linear_search(string, front, back); 200 201 if (front) 202 print_from(string, front, back); 203 return (front ? 0 : 1); 204 } 205 206 207 /* 208 * Binary search for "string" in memory between "front" and "back". 209 * 210 * This routine is expected to return a pointer to the start of a line at 211 * *or before* the first word matching "string". Relaxing the constraint 212 * this way simplifies the algorithm. 213 * 214 * Invariants: 215 * front points to the beginning of a line at or before the first 216 * matching string. 217 * 218 * back points to the beginning of a line at or after the first 219 * matching line. 220 * 221 * Base of the Invariants. 222 * front = NULL; 223 * back = EOF; 224 * 225 * Advancing the Invariants: 226 * 227 * p = first newline after halfway point from front to back. 228 * 229 * If the string at "p" is not greater than the string to match, 230 * p is the new front. Otherwise it is the new back. 231 * 232 * Termination: 233 * 234 * The definition of the routine allows it return at any point, 235 * since front is always at or before the line to print. 236 * 237 * In fact, it returns when the chosen "p" equals "back". This 238 * implies that there exists a string is least half as long as 239 * (back - front), which in turn implies that a linear search will 240 * be no more expensive than the cost of simply printing a string or two. 241 * 242 * Trying to continue with binary search at this point would be 243 * more trouble than it's worth. 244 */ 245 #define SKIP_PAST_NEWLINE(p, back) \ 246 while (p < back && *p++ != '\n'); 247 248 static char * 249 binary_search(wchar_t *string, unsigned char *front, unsigned char *back) 250 { 251 unsigned char *p; 252 253 p = front + (back - front) / 2; 254 SKIP_PAST_NEWLINE(p, back); 255 256 /* 257 * If the file changes underneath us, make sure we don't 258 * infinitely loop. 259 */ 260 while (p < back && back > front) { 261 if (compare(string, p, back) == GREATER) 262 front = p; 263 else 264 back = p; 265 p = front + (back - front) / 2; 266 SKIP_PAST_NEWLINE(p, back); 267 } 268 return (front); 269 } 270 271 /* 272 * Find the first line that starts with string, linearly searching from front 273 * to back. 274 * 275 * Return NULL for no such line. 276 * 277 * This routine assumes: 278 * 279 * o front points at the first character in a line. 280 * o front is before or at the first line to be printed. 281 */ 282 static char * 283 linear_search(wchar_t *string, unsigned char *front, unsigned char *back) 284 { 285 while (front < back) { 286 switch (compare(string, front, back)) { 287 case EQUAL: /* Found it. */ 288 return (front); 289 case LESS: /* No such string. */ 290 return (NULL); 291 case GREATER: /* Keep going. */ 292 break; 293 } 294 SKIP_PAST_NEWLINE(front, back); 295 } 296 return (NULL); 297 } 298 299 /* 300 * Print as many lines as match string, starting at front. 301 */ 302 static void 303 print_from(wchar_t *string, unsigned char *front, unsigned char *back) 304 { 305 for (; front < back && compare(string, front, back) == EQUAL; ++front) { 306 for (; front < back && *front != '\n'; ++front) 307 if (putchar(*front) == EOF) 308 err(2, "stdout"); 309 if (putchar('\n') == EOF) 310 err(2, "stdout"); 311 } 312 } 313 314 /* 315 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with 316 * string2 (s1 ??? s2). 317 * 318 * o Matches up to len(s1) are EQUAL. 319 * o Matches up to len(s2) are GREATER. 320 * 321 * Compare understands about the -f and -d flags, and treats comparisons 322 * appropriately. 323 * 324 * The string "s1" is null terminated. The string s2 is '\n' terminated (or 325 * "back" terminated). 326 */ 327 static int 328 compare(wchar_t *s1, unsigned char *s2, unsigned char *back) 329 { 330 wchar_t ch1, ch2; 331 size_t len2; 332 333 for (; *s1 && s2 < back && *s2 != '\n'; ++s1, s2 += len2) { 334 ch1 = *s1; 335 len2 = mbrtowc(&ch2, s2, back - s2, NULL); 336 if (len2 == (size_t)-1 || len2 == (size_t)-2) { 337 ch2 = *s2; 338 len2 = 1; 339 } 340 if (fflag) 341 ch2 = towlower(ch2); 342 if (dflag && !iswalnum(ch2)) { 343 /* Ignore character in comparison. */ 344 --s1; 345 continue; 346 } 347 if (ch1 != ch2) 348 return (ch1 < ch2 ? LESS : GREATER); 349 } 350 return (*s1 ? GREATER : EQUAL); 351 } 352 353 static void 354 usage(void) 355 { 356 (void)fprintf(stderr, "usage: look [-df] [-t char] string [file ...]\n"); 357 exit(2); 358 } 359