misc-utils/look.c

   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
  37  /* 1999-02-22 Arkadiusz Miśkiewicz <misiek@pld.ORG.PL>
  38   * - added Native Language Support
  39   */
  40
  41 /*
  42  * look -- find lines in a sorted list.
  43  *
  44  * The man page said that TABs and SPACEs participate in -d comparisons.
  45  * In fact, they were ignored.  This implements historic practice, not
  46  * the manual page.
  47  */
  48
  49 #include <sys/mman.h>
  50 #include <sys/stat.h>
  51 #include <errno.h>
  52 #include <fcntl.h>
  53 #include <stdio.h>
  54 #include <stdlib.h>
  55 #include <stddef.h>
  56 #include <string.h>
  57 #include <ctype.h>
  58 #include <getopt.h>
  59
  60 #include "nls.h"
  61 #include "xalloc.h"
  62 #include "pathnames.h"
  63 #include "closestream.h"
  64
  65 #define EQUAL           0
  66 #define GREATER         1
  67 #define LESS            (-1)
  68
  69 int dflag, fflag;
  70 /* uglified the source a bit with globals, so that we only need
  71    to allocate comparbuf once */
  72 int stringlen;
  73 char *string;
  74 char *comparbuf;
  75
  76 static char *binary_search (char *, char *);
  77 static int compare (char *, char *);
  78 static char *linear_search (char *, char *);
  79 static int look (char *, char *);
  80 static void print_from (char *, char *);
  81 static void __attribute__ ((__noreturn__)) usage(FILE * out);
  82
  83 int
  84 main(int argc, char *argv[])
  85 {
  86         struct stat sb;
  87         int ch, fd, termchar;
  88         char *back, *file, *front, *p;
  89
  90         static const struct option longopts[] = {
  91                 {"alternative", no_argument, NULL, 'a'},
  92                 {"alphanum", no_argument, NULL, 'd'},
  93                 {"ignore-case", no_argument, NULL, 'f'},
  94                 {"terminate", required_argument, NULL, 't'},
  95                 {"version", no_argument, NULL, 'V'},
  96                 {"help", no_argument, NULL, 'h'},
  97                 {NULL, 0, NULL, 0}
  98         };
  99
 100         setlocale(LC_ALL, "");
 101         bindtextdomain(PACKAGE, LOCALEDIR);
 102         textdomain(PACKAGE);
 103         atexit(close_stdout);
 104
 105         setlocale(LC_ALL, "");
 106
 107         file = _PATH_WORDS;
 108         termchar = '\0';
 109         string = NULL;          /* just for gcc */
 110
 111         while ((ch = getopt_long(argc, argv, "adft:Vh", longopts, NULL)) != -1)
 112                 switch(ch) {
 113                 case 'a':
 114                         file = _PATH_WORDS_ALT;
 115                         break;
 116                 case 'd':
 117                         dflag = 1;
 118                         break;
 119                 case 'f':
 120                         fflag = 1;
 121                         break;
 122                 case 't':
 123                         termchar = *optarg;
 124                         break;
 125                 case 'V':
 126                         printf(UTIL_LINUX_VERSION);
 127                         return EXIT_SUCCESS;
 128                 case 'h':
 129                         usage(stdout);
 130                 case '?':
 131                 default:
 132                         usage(stderr);
 133                 }
 134         argc -= optind;
 135         argv += optind;
 136
 137         switch (argc) {
 138         case 2:                         /* Don't set -df for user. */
 139                 string = *argv++;
 140                 file = *argv;
 141                 break;
 142         case 1:                         /* But set -df by default. */
 143                 dflag = fflag = 1;
 144                 string = *argv;
 145                 break;
 146         default:
 147                 usage(stderr);
 148         }
 149
 150         if (termchar != '\0' && (p = strchr(string, termchar)) != NULL)
 151                 *++p = '\0';
 152
 153         if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
 154                 err(EXIT_FAILURE, "%s", file);
 155         front = mmap(NULL, (size_t) sb.st_size, PROT_READ,
 156 #ifdef MAP_FILE
 157                      MAP_FILE |
 158 #endif
 159                      MAP_SHARED, fd, (off_t) 0);
 160         if
 161 #ifdef MAP_FAILED
 162                 (front == MAP_FAILED)
 163 #else
 164                 ((void *)(front) <= (void *)0)
 165 #endif
 166                         err(EXIT_FAILURE, "%s", file);
 167
 168 #if 0
 169         /* workaround for mmap problem (rmiller@duskglow.com) */
 170         if (front == (void *)0)
 171                 return 1;
 172 #endif
 173
 174         back = front + sb.st_size;
 175         return look(front, back);
 176 }
 177
 178 int
 179 look(char *front, char *back)
 180 {
 181         int ch;
 182         char *readp, *writep;
 183
 184         /* Reformat string string to avoid doing it multiple times later. */
 185         if (dflag) {
 186                 for (readp = writep = string; (ch = *readp++) != 0;) {
 187                         if (isalnum(ch))
 188                                 *(writep++) = ch;
 189                 }
 190                 *writep = '\0';
 191                 stringlen = writep - string;
 192         } else
 193                 stringlen = strlen(string);
 194
 195         comparbuf = xmalloc(stringlen+1);
 196
 197         front = binary_search(front, back);
 198         front = linear_search(front, back);
 199
 200         if (front)
 201                 print_from(front, back);
 202
 203         free(comparbuf);
 204
 205         return (front ? 0 : 1);
 206 }
 207
 208
 209 /*
 210  * Binary search for "string" in memory between "front" and "back".
 211  *
 212  * This routine is expected to return a pointer to the start of a line at
 213  * *or before* the first word matching "string".  Relaxing the constraint
 214  * this way simplifies the algorithm.
 215  *
 216  * Invariants:
 217  *      front points to the beginning of a line at or before the first
 218  *      matching string.
 219  *
 220  *      back points to the beginning of a line at or after the first
 221  *      matching line.
 222  *
 223  * Advancing the Invariants:
 224  *
 225  *      p = first newline after halfway point from front to back.
 226  *
 227  *      If the string at "p" is not greater than the string to match,
 228  *      p is the new front.  Otherwise it is the new back.
 229  *
 230  * Termination:
 231  *
 232  *      The definition of the routine allows it return at any point,
 233  *      since front is always at or before the line to print.
 234  *
 235  *      In fact, it returns when the chosen "p" equals "back".  This
 236  *      implies that there exists a string is least half as long as
 237  *      (back - front), which in turn implies that a linear search will
 238  *      be no more expensive than the cost of simply printing a string or two.
 239  *
 240  *      Trying to continue with binary search at this point would be
 241  *      more trouble than it's worth.
 242  */
 243 #define SKIP_PAST_NEWLINE(p, back) \
 244         while (p < back && *p++ != '\n')
 245
 246 char *
 247 binary_search(char *front, char *back)
 248 {
 249         char *p;
 250
 251         p = front + (back - front) / 2;
 252         SKIP_PAST_NEWLINE(p, back);
 253
 254         /*
 255          * If the file changes underneath us, make sure we don't
 256          * infinitely loop.
 257          */
 258         while (p < back && back > front) {
 259                 if (compare(p, back) == GREATER)
 260                         front = p;
 261                 else
 262                         back = p;
 263                 p = front + (back - front) / 2;
 264                 SKIP_PAST_NEWLINE(p, back);
 265         }
 266         return (front);
 267 }
 268
 269 /*
 270  * Find the first line that starts with string, linearly searching from front
 271  * to back.
 272  *
 273  * Return NULL for no such line.
 274  *
 275  * This routine assumes:
 276  *
 277  *      o front points at the first character in a line.
 278  *      o front is before or at the first line to be printed.
 279  */
 280 char *
 281 linear_search(char *front, char *back)
 282 {
 283         while (front < back) {
 284                 switch (compare(front, back)) {
 285                 case EQUAL:             /* Found it. */
 286                         return (front);
 287                         break;
 288                 case LESS:              /* No such string. */
 289                         return (NULL);
 290                         break;
 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 void
 303 print_from(char *front, char *back)
 304 {
 305         int eol;
 306
 307         while (front < back && compare(front, back) == EQUAL) {
 308                 if (compare(front, back) == EQUAL) {
 309                         eol = 0;
 310                         while (front < back && !eol) {
 311                                 if (putchar(*front) == EOF)
 312                                         err(EXIT_FAILURE, "stdout");
 313                                 if (*front++ == '\n')
 314                                         eol = 1;
 315                         }
 316                 } else
 317                         SKIP_PAST_NEWLINE(front, back);
 318         }
 319 }
 320
 321 /*
 322  * Return LESS, GREATER, or EQUAL depending on how  string  compares with
 323  * string2 (s1 ??? s2).
 324  *
 325  *      o Matches up to len(s1) are EQUAL.
 326  *      o Matches up to len(s2) are GREATER.
 327  *
 328  * Compare understands about the -f and -d flags, and treats comparisons
 329  * appropriately.
 330  *
 331  * The string "string" is null terminated.  The string "s2" is '\n' terminated
 332  * (or "s2end" terminated).
 333  *
 334  * We use strcasecmp etc, since it knows how to ignore case also
 335  * in other locales.
 336  */
 337 int
 338 compare(char *s2, char *s2end) {
 339         int i;
 340         char *p;
 341
 342         /* copy, ignoring things that should be ignored */
 343         p = comparbuf;
 344         i = stringlen;
 345         while(s2 < s2end && *s2 != '\n' && i) {
 346                 if (!dflag || isalnum(*s2))
 347                 {
 348                         *p++ = *s2;
 349                         i--;
 350                 }
 351                 s2++;
 352         }
 353         *p = 0;
 354
 355         /* and compare */
 356         if (fflag)
 357                 i = strncasecmp(comparbuf, string, stringlen);
 358         else
 359                 i = strncmp(comparbuf, string, stringlen);
 360
 361         return ((i > 0) ? LESS : (i < 0) ? GREATER : EQUAL);
 362 }
 363
 364 static void __attribute__ ((__noreturn__)) usage(FILE * out)
 365 {
 366         fputs(_("\nUsage:\n"), out),
 367         fprintf(out,
 368               _(" %s [options] string [file]\n"), program_invocation_short_name);
 369
 370         fputs(_("\nOptions:\n"), out);
 371         fputs(_(" -a, --alternative      use alternative dictionary\n"
 372                 " -d, --alphanum         compare only alphanumeric characters\n"
 373                 " -f, --ignore-case      ignore case differences when comparing\n"
 374                 " -t, --terminate <char> define string termination character\n"
 375                 " -V, --version          output version information and exit\n"
 376                 " -h, --help             display this help and exit\n\n"), out);
 377
 378         fprintf(out, USAGE_MAN_TAIL("look(1)"));
 379         exit(out == stderr ? EXIT_FAILURE : EXIT_SUCCESS);
 380 }