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