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. All advertising materials mentioning features or use of this software 17 * must display the following acknowledgement: 18 * This product includes software developed by the University of 19 * California, Berkeley and its contributors. 20 * 4. Neither the name of the University nor the names of its contributors 21 * may be used to endorse or promote products derived from this software 22 * without specific prior written permission. 23 * 24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 27 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 34 * SUCH DAMAGE. 35 * 36 * @(#) Copyright (c) 1991, 1993 The Regents of the University of California. All rights reserved. 37 * @(#)look.c 8.2 (Berkeley) 5/4/95 38 * $FreeBSD: src/usr.bin/look/look.c,v 1.11 1999/08/28 01:03:14 peter Exp $ 39 * $DragonFly: src/usr.bin/look/look.c,v 1.4 2005/01/07 02:43:41 cpressey Exp $ 40 */ 41 42 /* 43 * look -- find lines in a sorted list. 44 * 45 * The man page said that TABs and SPACEs participate in -d comparisons. 46 * In fact, they were ignored. This implements historic practice, not 47 * the manual page. 48 */ 49 50 #include <sys/types.h> 51 #include <sys/mman.h> 52 #include <sys/stat.h> 53 54 #include <ctype.h> 55 #include <err.h> 56 #include <errno.h> 57 #include <fcntl.h> 58 #include <limits.h> 59 #include <locale.h> 60 #include <stdio.h> 61 #include <stdlib.h> 62 #include <string.h> 63 #include <unistd.h> 64 65 #include "pathnames.h" 66 67 /* 68 * FOLD and DICT convert characters to a normal form for comparison, 69 * according to the user specified flags. 70 * 71 * DICT expects integers because it uses a non-character value to 72 * indicate a character which should not participate in comparisons. 73 */ 74 #define EQUAL 0 75 #define GREATER 1 76 #define LESS (-1) 77 #define NO_COMPARE (-2) 78 79 #define FOLD(c) (isupper(c) ? tolower(c) : (unsigned char) (c)) 80 #define DICT(c) (isalnum(c) ? (c) & 0xFF /* int */ : NO_COMPARE) 81 82 int dflag, fflag; 83 84 char *binary_search(unsigned char *, unsigned char *, unsigned char *); 85 int compare(unsigned char *, unsigned char *, unsigned char *); 86 char *linear_search(unsigned char *, unsigned char *, unsigned char *); 87 int look(unsigned char *, unsigned char *, unsigned char *); 88 void print_from(unsigned char *, unsigned char *, unsigned char *); 89 90 static void usage(void); 91 92 int 93 main(int argc, char **argv) 94 { 95 struct stat sb; 96 int ch, fd, termchar, match; 97 unsigned char *back, *front, *string, *p; 98 const unsigned char *file; 99 size_t len; 100 101 setlocale(LC_CTYPE, ""); 102 103 file = _PATH_WORDS; 104 termchar = '\0'; 105 while ((ch = getopt(argc, argv, "dft:")) != -1) 106 switch(ch) { 107 case 'd': 108 dflag = 1; 109 break; 110 case 'f': 111 fflag = 1; 112 break; 113 case 't': 114 termchar = *optarg; 115 break; 116 case '?': 117 default: 118 usage(); 119 } 120 argc -= optind; 121 argv += optind; 122 123 if (argc == 0) 124 usage(); 125 if (argc == 1) /* But set -df by default. */ 126 dflag = fflag = 1; 127 string = *argv++; 128 if (argc >= 2) 129 file = *argv++; 130 131 if (termchar != '\0' && (p = strchr(string, termchar)) != NULL) 132 *++p = '\0'; 133 match = 1; 134 135 do { 136 if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb)) 137 err(2, "%s", file); 138 len = (size_t)sb.st_size; 139 if ((off_t)len != sb.st_size) { 140 errno = EFBIG; 141 err(2, "%s", file); 142 } 143 if ((front = mmap(NULL, len, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED) 144 err(2, "%s", file); 145 back = front + sb.st_size; 146 match *= (look(string, front, back)); 147 close(fd); 148 } while (argc-- > 2 && (file = *argv++)); 149 150 exit(match); 151 } 152 153 int 154 look(unsigned char *string, unsigned char *front, unsigned char *back) 155 { 156 int ch; 157 unsigned char *readp, *writep; 158 159 /* Reformat string string to avoid doing it multiple times later. */ 160 for (readp = writep = string; (ch = *readp++) != '\0';) { 161 if (fflag) 162 ch = FOLD(ch); 163 if (dflag) 164 ch = DICT(ch); 165 if (ch != NO_COMPARE) 166 *(writep++) = ch; 167 } 168 *writep = '\0'; 169 170 front = binary_search(string, front, back); 171 front = linear_search(string, front, back); 172 173 if (front) 174 print_from(string, front, back); 175 return (front ? 0 : 1); 176 } 177 178 179 /* 180 * Binary search for "string" in memory between "front" and "back". 181 * 182 * This routine is expected to return a pointer to the start of a line at 183 * *or before* the first word matching "string". Relaxing the constraint 184 * this way simplifies the algorithm. 185 * 186 * Invariants: 187 * front points to the beginning of a line at or before the first 188 * matching string. 189 * 190 * back points to the beginning of a line at or after the first 191 * matching line. 192 * 193 * Base of the Invariants. 194 * front = NULL; 195 * back = EOF; 196 * 197 * Advancing the Invariants: 198 * 199 * p = first newline after halfway point from front to back. 200 * 201 * If the string at "p" is not greater than the string to match, 202 * p is the new front. Otherwise it is the new back. 203 * 204 * Termination: 205 * 206 * The definition of the routine allows it return at any point, 207 * since front is always at or before the line to print. 208 * 209 * In fact, it returns when the chosen "p" equals "back". This 210 * implies that there exists a string is least half as long as 211 * (back - front), which in turn implies that a linear search will 212 * be no more expensive than the cost of simply printing a string or two. 213 * 214 * Trying to continue with binary search at this point would be 215 * more trouble than it's worth. 216 */ 217 #define SKIP_PAST_NEWLINE(p, back) \ 218 while (p < back && *p++ != '\n'); 219 220 char * 221 binary_search(unsigned char *string, unsigned char *front, 222 unsigned char *back) 223 { 224 unsigned char *p; 225 226 p = front + (back - front) / 2; 227 SKIP_PAST_NEWLINE(p, back); 228 229 /* 230 * If the file changes underneath us, make sure we don't 231 * infinitely loop. 232 */ 233 while (p < back && back > front) { 234 if (compare(string, p, back) == GREATER) 235 front = p; 236 else 237 back = p; 238 p = front + (back - front) / 2; 239 SKIP_PAST_NEWLINE(p, back); 240 } 241 return (front); 242 } 243 244 /* 245 * Find the first line that starts with string, linearly searching from front 246 * to back. 247 * 248 * Return NULL for no such line. 249 * 250 * This routine assumes: 251 * 252 * o front points at the first character in a line. 253 * o front is before or at the first line to be printed. 254 */ 255 char * 256 linear_search(unsigned char *string, unsigned char *front, unsigned char *back) 257 { 258 while (front < back) { 259 switch (compare(string, front, back)) { 260 case EQUAL: /* Found it. */ 261 return (front); 262 break; 263 case LESS: /* No such string. */ 264 return (NULL); 265 break; 266 case GREATER: /* Keep going. */ 267 break; 268 } 269 SKIP_PAST_NEWLINE(front, back); 270 } 271 return (NULL); 272 } 273 274 /* 275 * Print as many lines as match string, starting at front. 276 */ 277 void 278 print_from(unsigned char *string, unsigned char *front, 279 unsigned char *back) 280 { 281 for (; front < back && compare(string, front, back) == EQUAL; ++front) { 282 for (; front < back && *front != '\n'; ++front) 283 if (putchar(*front) == EOF) 284 err(2, "stdout"); 285 if (putchar('\n') == EOF) 286 err(2, "stdout"); 287 } 288 } 289 290 /* 291 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with 292 * string2 (s1 ??? s2). 293 * 294 * o Matches up to len(s1) are EQUAL. 295 * o Matches up to len(s2) are GREATER. 296 * 297 * Compare understands about the -f and -d flags, and treats comparisons 298 * appropriately. 299 * 300 * The string "s1" is null terminated. The string s2 is '\n' terminated (or 301 * "back" terminated). 302 */ 303 int 304 compare(unsigned char *s1, unsigned char *s2, 305 unsigned char *back) 306 { 307 int ch; 308 309 for (; *s1 && s2 < back && *s2 != '\n'; ++s1, ++s2) { 310 ch = *s2; 311 if (fflag) 312 ch = FOLD(ch); 313 if (dflag) 314 ch = DICT(ch); 315 316 if (ch == NO_COMPARE) { 317 ++s2; /* Ignore character in comparison. */ 318 continue; 319 } 320 if (*s1 != ch) 321 return (*s1 < ch ? LESS : GREATER); 322 } 323 return (*s1 ? GREATER : EQUAL); 324 } 325 326 static void 327 usage(void) 328 { 329 fprintf(stderr, "usage: look [-df] [-t char] string [file ...]\n"); 330 exit(2); 331 } 332