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 static int dflag, fflag;
  70 /* uglified the source a bit with globals, so that we only need
  71    to allocate comparbuf once */
  72 static int stringlen;
  73 static char *string;
  74 static 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(void);
  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         if ((file = getenv("WORDLIST")) && !access(file, R_OK))
 108                 /* use the WORDLIST */;
 109         else
 110                 file = _PATH_WORDS;
 111
 112         termchar = '\0';
 113         string = NULL;          /* just for gcc */
 114
 115         while ((ch = getopt_long(argc, argv, "adft:Vh", longopts, NULL)) != -1)
 116                 switch(ch) {
 117                 case 'a':
 118                         file = _PATH_WORDS_ALT;
 119                         break;
 120                 case 'd':
 121                         dflag = 1;
 122                         break;
 123                 case 'f':
 124                         fflag = 1;
 125                         break;
 126                 case 't':
 127                         termchar = *optarg;
 128                         break;
 129                 case 'V':
 130                         printf(UTIL_LINUX_VERSION);
 131                         return EXIT_SUCCESS;
 132                 case 'h':
 133                         usage();
 134                 default:
 135                         errtryhelp(EXIT_FAILURE);
 136                 }
 137         argc -= optind;
 138         argv += optind;
 139
 140         switch (argc) {
 141         case 2:                         /* Don't set -df for user. */
 142                 string = *argv++;
 143                 file = *argv;
 144                 break;
 145         case 1:                         /* But set -df by default. */
 146                 dflag = fflag = 1;
 147                 string = *argv;
 148                 break;
 149         default:
 150                 warnx(_("bad usage"));
 151                 errtryhelp(EXIT_FAILURE);
 152         }
 153
 154         if (termchar != '\0' && (p = strchr(string, termchar)) != NULL)
 155                 *++p = '\0';
 156
 157         if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
 158                 err(EXIT_FAILURE, "%s", file);
 159         front = mmap(NULL, (size_t) sb.st_size, PROT_READ,
 160                      MAP_SHARED, fd, (off_t) 0);
 161         if
 162 #ifdef MAP_FAILED
 163                 (front == MAP_FAILED)
 164 #else
 165                 ((void *)(front) <= (void *)0)
 166 #endif
 167                         err(EXIT_FAILURE, "%s", file);
 168         back = front + sb.st_size;
 169         return look(front, back);
 170 }
 171
 172 static int
 173 look(char *front, char *back)
 174 {
 175         int ch;
 176         char *readp, *writep;
 177
 178         /* Reformat string to avoid doing it multiple times later. */
 179         if (dflag) {
 180                 for (readp = writep = string; (ch = *readp++) != 0;) {
 181                         if (isalnum(ch) || isblank(ch))
 182                                 *(writep++) = ch;
 183                 }
 184                 *writep = '\0';
 185                 stringlen = writep - string;
 186         } else
 187                 stringlen = strlen(string);
 188
 189         comparbuf = xmalloc(stringlen+1);
 190
 191         front = binary_search(front, back);
 192         front = linear_search(front, back);
 193
 194         if (front)
 195                 print_from(front, back);
 196
 197         free(comparbuf);
 198
 199         return (front ? 0 : 1);
 200 }
 201
 202
 203 /*
 204  * Binary search for "string" in memory between "front" and "back".
 205  *
 206  * This routine is expected to return a pointer to the start of a line at
 207  * *or before* the first word matching "string".  Relaxing the constraint
 208  * this way simplifies the algorithm.
 209  *
 210  * Invariants:
 211  *      front points to the beginning of a line at or before the first
 212  *      matching string.
 213  *
 214  *      back points to the beginning of a line at or after the first
 215  *      matching line.
 216  *
 217  * Advancing the Invariants:
 218  *
 219  *      p = first newline after halfway point from front to back.
 220  *
 221  *      If the string at "p" is not greater than the string to match,
 222  *      p is the new front.  Otherwise it is the new back.
 223  *
 224  * Termination:
 225  *
 226  *      The definition of the routine allows it return at any point,
 227  *      since front is always at or before the line to print.
 228  *
 229  *      In fact, it returns when the chosen "p" equals "back".  This
 230  *      implies that there exists a string is least half as long as
 231  *      (back - front), which in turn implies that a linear search will
 232  *      be no more expensive than the cost of simply printing a string or two.
 233  *
 234  *      Trying to continue with binary search at this point would be
 235  *      more trouble than it's worth.
 236  */
 237 #define SKIP_PAST_NEWLINE(p, back) \
 238         while (p < back && *p++ != '\n')
 239
 240 static char *
 241 binary_search(char *front, char *back)
 242 {
 243         char *p;
 244
 245         p = front + (back - front) / 2;
 246         SKIP_PAST_NEWLINE(p, back);
 247
 248         /*
 249          * If the file changes underneath us, make sure we don't
 250          * infinitely loop.
 251          */
 252         while (p < back && back > front) {
 253                 if (compare(p, back) == GREATER)
 254                         front = p;
 255                 else
 256                         back = p;
 257                 p = front + (back - front) / 2;
 258                 SKIP_PAST_NEWLINE(p, back);
 259         }
 260         return (front);
 261 }
 262
 263 /*
 264  * Find the first line that starts with string, linearly searching from front
 265  * to back.
 266  *
 267  * Return NULL for no such line.
 268  *
 269  * This routine assumes:
 270  *
 271  *      o front points at the first character in a line.
 272  *      o front is before or at the first line to be printed.
 273  */
 274 static char *
 275 linear_search(char *front, char *back)
 276 {
 277         while (front < back) {
 278                 switch (compare(front, back)) {
 279                 case EQUAL:             /* Found it. */
 280                         return (front);
 281                 case LESS:              /* No such string. */
 282                         return (NULL);
 283                 case GREATER:           /* Keep going. */
 284                         break;
 285                 }
 286                 SKIP_PAST_NEWLINE(front, back);
 287         }
 288         return (NULL);
 289 }
 290
 291 /*
 292  * Print as many lines as match string, starting at front.
 293  */
 294 static void
 295 print_from(char *front, char *back)
 296 {
 297         int eol;
 298
 299         while (front < back && compare(front, back) == EQUAL) {
 300                 if (compare(front, back) == EQUAL) {
 301                         eol = 0;
 302                         while (front < back && !eol) {
 303                                 if (putchar(*front) == EOF)
 304                                         err(EXIT_FAILURE, "stdout");
 305                                 if (*front++ == '\n')
 306                                         eol = 1;
 307                         }
 308                 } else
 309                         SKIP_PAST_NEWLINE(front, back);
 310         }
 311 }
 312
 313 /*
 314  * Return LESS, GREATER, or EQUAL depending on how  string  compares with
 315  * string2 (s1 ??? s2).
 316  *
 317  *      o Matches up to len(s1) are EQUAL.
 318  *      o Matches up to len(s2) are GREATER.
 319  *
 320  * Compare understands about the -f and -d flags, and treats comparisons
 321  * appropriately.
 322  *
 323  * The string "string" is null terminated.  The string "s2" is '\n' terminated
 324  * (or "s2end" terminated).
 325  *
 326  * We use strcasecmp etc, since it knows how to ignore case also
 327  * in other locales.
 328  */
 329 static int
 330 compare(char *s2, char *s2end) {
 331         int i;
 332         char *p;
 333
 334         /* copy, ignoring things that should be ignored */
 335         p = comparbuf;
 336         i = stringlen;
 337         while(s2 < s2end && *s2 != '\n' && i) {
 338                 if (!dflag || isalnum(*s2) || isblank(*s2))
 339                 {
 340                         *p++ = *s2;
 341                         i--;
 342                 }
 343                 s2++;
 344         }
 345         *p = 0;
 346
 347         /* and compare */
 348         if (fflag)
 349                 i = strncasecmp(comparbuf, string, stringlen);
 350         else
 351                 i = strncmp(comparbuf, string, stringlen);
 352
 353         return ((i > 0) ? LESS : (i < 0) ? GREATER : EQUAL);
 354 }
 355
 356 static void __attribute__((__noreturn__)) usage(void)
 357 {
 358         FILE *out = stdout;
 359         fputs(USAGE_HEADER, out);
 360         fprintf(out, _(" %s [options] <string> [<file>...]\n"), program_invocation_short_name);
 361
 362         fputs(USAGE_SEPARATOR, out);
 363         fputs(_("Display lines beginning with a specified string.\n"), out);
 364
 365         fputs(USAGE_OPTIONS, out);
 366         fputs(_(" -a, --alternative        use the alternative dictionary\n"), out);
 367         fputs(_(" -d, --alphanum           compare only blanks and alphanumeric characters\n"), out);
 368         fputs(_(" -f, --ignore-case        ignore case differences when comparing\n"), out);
 369         fputs(_(" -t, --terminate <char>   define the string-termination character\n"), out);
 370
 371         fputs(USAGE_SEPARATOR, out);
 372         printf(USAGE_HELP_OPTIONS(26));
 373         printf(USAGE_MAN_TAIL("look(1)"));
 374
 375         exit(EXIT_SUCCESS);
 376 }