1 /*- 2 * Copyright (c) 1991, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * David Hitz of Auspex Systems, Inc. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 3. Neither the name of the University nor the names of its contributors 17 * may be used to endorse or promote products derived from this software 18 * without specific prior written permission. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 23 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 30 * SUCH DAMAGE. 31 */ 32 33 #ifndef lint 34 static const char copyright[] = 35 "@(#) Copyright (c) 1991, 1993\n\ 36 The Regents of the University of California. All rights reserved.\n"; 37 #endif /* not lint */ 38 39 #ifndef lint 40 #if 0 41 static char sccsid[] = "@(#)look.c 8.2 (Berkeley) 5/4/95"; 42 #endif 43 #endif /* not lint */ 44 #include <sys/cdefs.h> 45 __FBSDID("$FreeBSD$"); 46 47 /* 48 * look -- find lines in a sorted list. 49 * 50 * The man page said that TABs and SPACEs participate in -d comparisons. 51 * In fact, they were ignored. This implements historic practice, not 52 * the manual page. 53 */ 54 55 #include <sys/types.h> 56 #include <sys/mman.h> 57 #include <sys/stat.h> 58 59 #include <err.h> 60 #include <errno.h> 61 #include <fcntl.h> 62 #include <getopt.h> 63 #include <limits.h> 64 #include <locale.h> 65 #include <stdint.h> 66 #include <stdio.h> 67 #include <stdlib.h> 68 #include <string.h> 69 #include <unistd.h> 70 #include <wchar.h> 71 #include <wctype.h> 72 73 #include "pathnames.h" 74 75 static char _path_words[] = _PATH_WORDS; 76 77 #define EQUAL 0 78 #define GREATER 1 79 #define LESS (-1) 80 81 static int dflag, fflag; 82 83 static char *binary_search(wchar_t *, unsigned char *, unsigned char *); 84 static int compare(wchar_t *, unsigned char *, unsigned char *); 85 static char *linear_search(wchar_t *, unsigned char *, unsigned char *); 86 static int look(wchar_t *, unsigned char *, unsigned char *); 87 static wchar_t *prepkey(const char *, wchar_t); 88 static void print_from(wchar_t *, unsigned char *, unsigned char *); 89 90 static void usage(void); 91 92 static struct option longopts[] = { 93 { "alternative",no_argument, NULL, 'a' }, 94 { "alphanum", no_argument, NULL, 'd' }, 95 { "ignore-case",no_argument, NULL, 'i' }, 96 { "terminate", required_argument, NULL, 't'}, 97 { NULL, 0, NULL, 0 }, 98 }; 99 100 int 101 main(int argc, char *argv[]) 102 { 103 struct stat sb; 104 int ch, fd, match; 105 wchar_t termchar; 106 unsigned char *back, *front; 107 unsigned const char *file; 108 wchar_t *key; 109 110 (void) setlocale(LC_CTYPE, ""); 111 112 file = _path_words; 113 termchar = L'\0'; 114 while ((ch = getopt_long(argc, argv, "+adft:", longopts, NULL)) != -1) 115 switch(ch) { 116 case 'a': 117 /* COMPATIBILITY */ 118 break; 119 case 'd': 120 dflag = 1; 121 break; 122 case 'f': 123 fflag = 1; 124 break; 125 case 't': 126 if (mbrtowc(&termchar, optarg, MB_LEN_MAX, NULL) != 127 strlen(optarg)) 128 errx(2, "invalid termination character"); 129 break; 130 case '?': 131 default: 132 usage(); 133 } 134 argc -= optind; 135 argv += optind; 136 137 if (argc == 0) 138 usage(); 139 if (argc == 1) /* But set -df by default. */ 140 dflag = fflag = 1; 141 key = prepkey(*argv++, termchar); 142 if (argc >= 2) 143 file = *argv++; 144 145 match = 1; 146 147 do { 148 if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb)) 149 err(2, "%s", file); 150 if ((uintmax_t)sb.st_size > (uintmax_t)SIZE_T_MAX) 151 errx(2, "%s: %s", file, strerror(EFBIG)); 152 if (sb.st_size == 0) { 153 close(fd); 154 continue; 155 } 156 if ((front = mmap(NULL, (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED) 157 err(2, "%s", file); 158 back = front + sb.st_size; 159 match *= (look(key, front, back)); 160 close(fd); 161 } while (argc-- > 2 && (file = *argv++)); 162 163 exit(match); 164 } 165 166 static wchar_t * 167 prepkey(const char *string, wchar_t termchar) 168 { 169 const char *readp; 170 wchar_t *key, *writep; 171 wchar_t ch; 172 size_t clen; 173 174 /* 175 * Reformat search string and convert to wide character representation 176 * to avoid doing it multiple times later. 177 */ 178 if ((key = malloc(sizeof(wchar_t) * (strlen(string) + 1))) == NULL) 179 err(2, NULL); 180 readp = string; 181 writep = key; 182 while ((clen = mbrtowc(&ch, readp, MB_LEN_MAX, NULL)) != 0) { 183 if (clen == (size_t)-1 || clen == (size_t)-2) 184 errc(2, EILSEQ, NULL); 185 if (fflag) 186 ch = towlower(ch); 187 if (!dflag || iswalnum(ch)) 188 *writep++ = ch; 189 readp += clen; 190 } 191 *writep = L'\0'; 192 if (termchar != L'\0' && (writep = wcschr(key, termchar)) != NULL) 193 *++writep = L'\0'; 194 return (key); 195 } 196 197 static int 198 look(wchar_t *string, unsigned char *front, unsigned char *back) 199 { 200 201 front = binary_search(string, front, back); 202 front = linear_search(string, front, back); 203 204 if (front) 205 print_from(string, front, back); 206 return (front ? 0 : 1); 207 } 208 209 210 /* 211 * Binary search for "string" in memory between "front" and "back". 212 * 213 * This routine is expected to return a pointer to the start of a line at 214 * *or before* the first word matching "string". Relaxing the constraint 215 * this way simplifies the algorithm. 216 * 217 * Invariants: 218 * front points to the beginning of a line at or before the first 219 * matching string. 220 * 221 * back points to the beginning of a line at or after the first 222 * matching line. 223 * 224 * Base of the Invariants. 225 * front = NULL; 226 * back = EOF; 227 * 228 * Advancing the Invariants: 229 * 230 * p = first newline after halfway point from front to back. 231 * 232 * If the string at "p" is not greater than the string to match, 233 * p is the new front. Otherwise it is the new back. 234 * 235 * Termination: 236 * 237 * The definition of the routine allows it return at any point, 238 * since front is always at or before the line to print. 239 * 240 * In fact, it returns when the chosen "p" equals "back". This 241 * implies that there exists a string is least half as long as 242 * (back - front), which in turn implies that a linear search will 243 * be no more expensive than the cost of simply printing a string or two. 244 * 245 * Trying to continue with binary search at this point would be 246 * more trouble than it's worth. 247 */ 248 #define SKIP_PAST_NEWLINE(p, back) \ 249 while (p < back && *p++ != '\n'); 250 251 static char * 252 binary_search(wchar_t *string, unsigned char *front, unsigned char *back) 253 { 254 unsigned char *p; 255 256 p = front + (back - front) / 2; 257 SKIP_PAST_NEWLINE(p, back); 258 259 /* 260 * If the file changes underneath us, make sure we don't 261 * infinitely loop. 262 */ 263 while (p < back && back > front) { 264 if (compare(string, p, back) == GREATER) 265 front = p; 266 else 267 back = p; 268 p = front + (back - front) / 2; 269 SKIP_PAST_NEWLINE(p, back); 270 } 271 return (front); 272 } 273 274 /* 275 * Find the first line that starts with string, linearly searching from front 276 * to back. 277 * 278 * Return NULL for no such line. 279 * 280 * This routine assumes: 281 * 282 * o front points at the first character in a line. 283 * o front is before or at the first line to be printed. 284 */ 285 static char * 286 linear_search(wchar_t *string, unsigned char *front, unsigned char *back) 287 { 288 while (front < back) { 289 switch (compare(string, front, back)) { 290 case EQUAL: /* Found it. */ 291 return (front); 292 case LESS: /* No such string. */ 293 return (NULL); 294 case GREATER: /* Keep going. */ 295 break; 296 } 297 SKIP_PAST_NEWLINE(front, back); 298 } 299 return (NULL); 300 } 301 302 /* 303 * Print as many lines as match string, starting at front. 304 */ 305 static void 306 print_from(wchar_t *string, unsigned char *front, unsigned char *back) 307 { 308 for (; front < back && compare(string, front, back) == EQUAL; ++front) { 309 for (; front < back && *front != '\n'; ++front) 310 if (putchar(*front) == EOF) 311 err(2, "stdout"); 312 if (putchar('\n') == EOF) 313 err(2, "stdout"); 314 } 315 } 316 317 /* 318 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with 319 * string2 (s1 ??? s2). 320 * 321 * o Matches up to len(s1) are EQUAL. 322 * o Matches up to len(s2) are GREATER. 323 * 324 * Compare understands about the -f and -d flags, and treats comparisons 325 * appropriately. 326 * 327 * The string "s1" is null terminated. The string s2 is '\n' terminated (or 328 * "back" terminated). 329 */ 330 static int 331 compare(wchar_t *s1, unsigned char *s2, unsigned char *back) 332 { 333 wchar_t ch1, ch2; 334 size_t len2; 335 336 for (; *s1 && s2 < back && *s2 != '\n'; ++s1, s2 += len2) { 337 ch1 = *s1; 338 len2 = mbrtowc(&ch2, s2, back - s2, NULL); 339 if (len2 == (size_t)-1 || len2 == (size_t)-2) { 340 ch2 = *s2; 341 len2 = 1; 342 } 343 if (fflag) 344 ch2 = towlower(ch2); 345 if (dflag && !iswalnum(ch2)) { 346 /* Ignore character in comparison. */ 347 --s1; 348 continue; 349 } 350 if (ch1 != ch2) 351 return (ch1 < ch2 ? LESS : GREATER); 352 } 353 return (*s1 ? GREATER : EQUAL); 354 } 355 356 static void 357 usage(void) 358 { 359 (void)fprintf(stderr, "usage: look [-df] [-t char] string [file ...]\n"); 360 exit(2); 361 } 362