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/capsicum.h> 45 #include <sys/mman.h> 46 #include <sys/stat.h> 47 48 #include <capsicum_helpers.h> 49 #include <err.h> 50 #include <errno.h> 51 #include <fcntl.h> 52 #include <getopt.h> 53 #include <limits.h> 54 #include <locale.h> 55 #include <stdint.h> 56 #include <stdio.h> 57 #include <stdlib.h> 58 #include <string.h> 59 #include <unistd.h> 60 #include <wchar.h> 61 #include <wctype.h> 62 63 #include "pathnames.h" 64 65 static char _path_words[] = _PATH_WORDS; 66 67 #define EQUAL 0 68 #define GREATER 1 69 #define LESS (-1) 70 71 static int dflag, fflag; 72 73 static char *binary_search(wchar_t *, unsigned char *, unsigned char *); 74 static int compare(wchar_t *, unsigned char *, unsigned char *); 75 static char *linear_search(wchar_t *, unsigned char *, unsigned char *); 76 static int look(wchar_t *, unsigned char *, unsigned char *); 77 static wchar_t *prepkey(const char *, wchar_t); 78 static void print_from(wchar_t *, unsigned char *, unsigned char *); 79 80 static void usage(void) __dead2; 81 82 static struct option longopts[] = { 83 { "alternative",no_argument, NULL, 'a' }, 84 { "alphanum", no_argument, NULL, 'd' }, 85 { "ignore-case",no_argument, NULL, 'i' }, 86 { "terminate", required_argument, NULL, 't'}, 87 { NULL, 0, NULL, 0 }, 88 }; 89 90 struct files { 91 int fd; 92 int err; 93 const char *path; 94 }; 95 96 int 97 main(int argc, char *argv[]) 98 { 99 struct stat sb; 100 int ch, match; 101 size_t nfiles; 102 wchar_t termchar; 103 cap_rights_t rights; 104 struct files *files; 105 unsigned char *back, *front; 106 wchar_t *key; 107 108 (void) setlocale(LC_CTYPE, ""); 109 110 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 argc--; 140 141 match = 1; 142 143 cap_rights_init(&rights, CAP_MMAP_R, CAP_READ, CAP_FSTAT); 144 nfiles = !argc ? 1 : argc; 145 if ((files = malloc(nfiles * sizeof(struct files))) == NULL) 146 err(2, NULL); 147 for (size_t idx = 0; idx < nfiles; idx++) { 148 files[idx].path = !argc ? _path_words : argv[idx]; 149 if ((files[idx].fd = open(files[idx].path, O_RDONLY, 0)) < 0) { 150 files[idx].err = errno; 151 continue; 152 } 153 files[idx].err = 0; 154 if (caph_rights_limit(files[idx].fd, &rights) != 0) 155 err(2, "unable to limit rights for %s", files[idx].path); 156 } 157 158 caph_cache_catpages(); 159 if (caph_enter() != 0) 160 err(EXIT_FAILURE, "failed to enter capability mode"); 161 162 for (size_t idx = 0; idx < nfiles; idx++) { 163 if (files[idx].err) 164 errc(2, files[idx].err, "%s", files[idx].path); 165 if (fstat(files[idx].fd, &sb)) 166 err(2, "%s", files[idx].path); 167 if ((uintmax_t)sb.st_size > (uintmax_t)SIZE_T_MAX) 168 errx(2, "%s: %s", files[idx].path, strerror(EFBIG)); 169 if (sb.st_size == 0) { 170 close(files[idx].fd); 171 continue; 172 } 173 if ((front = mmap(NULL, (size_t)sb.st_size, PROT_READ, 174 MAP_SHARED, files[idx].fd, (off_t)0)) == MAP_FAILED) 175 err(2, "%s", files[idx].path); 176 back = front + sb.st_size; 177 match *= (look(key, front, back)); 178 close(files[idx].fd); 179 } 180 181 free(files); 182 exit(match); 183 } 184 185 static wchar_t * 186 prepkey(const char *string, wchar_t termchar) 187 { 188 const char *readp; 189 wchar_t *key, *writep; 190 wchar_t ch; 191 size_t clen; 192 193 /* 194 * Reformat search string and convert to wide character representation 195 * to avoid doing it multiple times later. 196 */ 197 if ((key = malloc(sizeof(wchar_t) * (strlen(string) + 1))) == NULL) 198 err(2, NULL); 199 readp = string; 200 writep = key; 201 while ((clen = mbrtowc(&ch, readp, MB_LEN_MAX, NULL)) != 0) { 202 if (clen == (size_t)-1 || clen == (size_t)-2) 203 errc(2, EILSEQ, NULL); 204 if (fflag) 205 ch = towlower(ch); 206 if (!dflag || iswalnum(ch)) 207 *writep++ = ch; 208 readp += clen; 209 } 210 *writep = L'\0'; 211 if (termchar != L'\0' && (writep = wcschr(key, termchar)) != NULL) 212 *++writep = L'\0'; 213 return (key); 214 } 215 216 static int 217 look(wchar_t *string, unsigned char *front, unsigned char *back) 218 { 219 220 front = binary_search(string, front, back); 221 front = linear_search(string, front, back); 222 223 if (front) 224 print_from(string, front, back); 225 return (front ? 0 : 1); 226 } 227 228 229 /* 230 * Binary search for "string" in memory between "front" and "back". 231 * 232 * This routine is expected to return a pointer to the start of a line at 233 * *or before* the first word matching "string". Relaxing the constraint 234 * this way simplifies the algorithm. 235 * 236 * Invariants: 237 * front points to the beginning of a line at or before the first 238 * matching string. 239 * 240 * back points to the beginning of a line at or after the first 241 * matching line. 242 * 243 * Base of the Invariants. 244 * front = NULL; 245 * back = EOF; 246 * 247 * Advancing the Invariants: 248 * 249 * p = first newline after halfway point from front to back. 250 * 251 * If the string at "p" is not greater than the string to match, 252 * p is the new front. Otherwise it is the new back. 253 * 254 * Termination: 255 * 256 * The definition of the routine allows it return at any point, 257 * since front is always at or before the line to print. 258 * 259 * In fact, it returns when the chosen "p" equals "back". This 260 * implies that there exists a string is least half as long as 261 * (back - front), which in turn implies that a linear search will 262 * be no more expensive than the cost of simply printing a string or two. 263 * 264 * Trying to continue with binary search at this point would be 265 * more trouble than it's worth. 266 */ 267 #define SKIP_PAST_NEWLINE(p, back) \ 268 while (p < back && *p++ != '\n'); 269 270 static char * 271 binary_search(wchar_t *string, unsigned char *front, unsigned char *back) 272 { 273 unsigned char *p; 274 275 p = front + (back - front) / 2; 276 SKIP_PAST_NEWLINE(p, back); 277 278 /* 279 * If the file changes underneath us, make sure we don't 280 * infinitely loop. 281 */ 282 while (p < back && back > front) { 283 if (compare(string, p, back) == GREATER) 284 front = p; 285 else 286 back = p; 287 p = front + (back - front) / 2; 288 SKIP_PAST_NEWLINE(p, back); 289 } 290 return (front); 291 } 292 293 /* 294 * Find the first line that starts with string, linearly searching from front 295 * to back. 296 * 297 * Return NULL for no such line. 298 * 299 * This routine assumes: 300 * 301 * o front points at the first character in a line. 302 * o front is before or at the first line to be printed. 303 */ 304 static char * 305 linear_search(wchar_t *string, unsigned char *front, unsigned char *back) 306 { 307 while (front < back) { 308 switch (compare(string, front, back)) { 309 case EQUAL: /* Found it. */ 310 return (front); 311 case LESS: /* No such string. */ 312 return (NULL); 313 case GREATER: /* Keep going. */ 314 break; 315 } 316 SKIP_PAST_NEWLINE(front, back); 317 } 318 return (NULL); 319 } 320 321 /* 322 * Print as many lines as match string, starting at front. 323 */ 324 static void 325 print_from(wchar_t *string, unsigned char *front, unsigned char *back) 326 { 327 for (; front < back && compare(string, front, back) == EQUAL; ++front) { 328 for (; front < back && *front != '\n'; ++front) 329 if (putchar(*front) == EOF) 330 err(2, "stdout"); 331 if (putchar('\n') == EOF) 332 err(2, "stdout"); 333 } 334 } 335 336 /* 337 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with 338 * string2 (s1 ??? s2). 339 * 340 * o Matches up to len(s1) are EQUAL. 341 * o Matches up to len(s2) are GREATER. 342 * 343 * Compare understands about the -f and -d flags, and treats comparisons 344 * appropriately. 345 * 346 * The string "s1" is null terminated. The string s2 is '\n' terminated (or 347 * "back" terminated). 348 */ 349 static int 350 compare(wchar_t *s1, unsigned char *s2, unsigned char *back) 351 { 352 wchar_t ch1, ch2; 353 size_t len2; 354 355 for (; *s1 && s2 < back && *s2 != '\n'; ++s1, s2 += len2) { 356 ch1 = *s1; 357 len2 = mbrtowc(&ch2, s2, back - s2, NULL); 358 if (len2 == (size_t)-1 || len2 == (size_t)-2) { 359 ch2 = *s2; 360 len2 = 1; 361 } 362 if (fflag) 363 ch2 = towlower(ch2); 364 if (dflag && !iswalnum(ch2)) { 365 /* Ignore character in comparison. */ 366 --s1; 367 continue; 368 } 369 if (ch1 != ch2) 370 return (ch1 < ch2 ? LESS : GREATER); 371 } 372 return (*s1 ? GREATER : EQUAL); 373 } 374 375 static void 376 usage(void) 377 { 378 (void)fprintf(stderr, "usage: look [-df] [-t char] string [file ...]\n"); 379 exit(2); 380 } 381