usr.bin/look/look.c

   1 /*      $NetBSD: look.c,v 1.13 2009/04/12 14:01:20 lukem Exp $  */
   2
   3 /*-
   4  * Copyright (c) 1991, 1993
   5  *      The Regents of the University of California.  All rights reserved.
   6  *
   7  * This code is derived from software contributed to Berkeley by
   8  * David Hitz of Auspex Systems, Inc.
   9  *
  10  * Redistribution and use in source and binary forms, with or without
  11  * modification, are permitted provided that the following conditions
  12  * are met:
  13  * 1. Redistributions of source code must retain the above copyright
  14  *    notice, this list of conditions and the following disclaimer.
  15  * 2. Redistributions in binary form must reproduce the above copyright
  16  *    notice, this list of conditions and the following disclaimer in the
  17  *    documentation and/or other materials provided with the distribution.
  18  * 3. Neither the name of the University nor the names of its contributors
  19  *    may be used to endorse or promote products derived from this software
  20  *    without specific prior written permission.
  21  *
  22  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  23  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  24  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  25  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  26  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  27  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  28  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  29  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  30  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  31  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  32  * SUCH DAMAGE.
  33  */
  34
  35 #include <sys/cdefs.h>
  36 #ifndef lint
  37 __COPYRIGHT("@(#) Copyright (c) 1991, 1993\
  38  The Regents of the University of California.  All rights reserved.");
  39 #endif /* not lint */
  40
  41 #ifndef lint
  42 #if 0
  43 static char sccsid[] = "@(#)look.c      8.2 (Berkeley) 5/4/95";
  44 #endif
  45 __RCSID("$NetBSD: look.c,v 1.13 2009/04/12 14:01:20 lukem Exp $");
  46 #endif /* not lint */
  47
  48 /*
  49  * look -- find lines in a sorted list.
  50  *
  51  * The man page said that TABs and SPACEs participate in -d comparisons.
  52  * In fact, they were ignored.  This implements historic practice, not
  53  * the manual page.
  54  */
  55
  56 #include <sys/types.h>
  57 #include <sys/mman.h>
  58 #include <sys/stat.h>
  59
  60 #include <ctype.h>
  61 #include <errno.h>
  62 #include <fcntl.h>
  63 #include <limits.h>
  64 #include <stdio.h>
  65 #include <stdlib.h>
  66 #include <string.h>
  67 #include <unistd.h>
  68 #include <err.h>
  69
  70 #include "pathnames.h"
  71
  72 /*
  73  * FOLD and DICT convert characters to a normal form for comparison,
  74  * according to the user specified flags.
  75  *
  76  * DICT expects integers because it uses a non-character value to
  77  * indicate a character which should not participate in comparisons.
  78  */
  79 #define EQUAL           0
  80 #define GREATER         1
  81 #define LESS            (-1)
  82 #define NO_COMPARE      (-2)
  83
  84 #define FOLD(c) (isascii(c) && isupper(c) ? tolower(c) : (c))
  85 #define DICT(c) (isascii(c) && isalnum(c) ? (c) : NO_COMPARE)
  86
  87 int dflag, fflag;
  88
  89 char    *binary_search __P((char *, char *, char *));
  90 int      compare __P((char *, char *, char *));
  91 char    *linear_search __P((char *, char *, char *));
  92 int      look __P((char *, char *, char *));
  93 int      main __P((int, char **));
  94 void     print_from __P((char *, char *, char *));
  95 void     usage __P((void));
  96
  97 int
  98 main(argc, argv)
  99         int argc;
 100         char *argv[];
 101 {
 102         struct stat sb;
 103         int ch, fd, termchar;
 104         char *back, *front, *string, *p;
 105         const char *file;
 106         size_t len;
 107
 108         string = NULL;
 109         file = _PATH_WORDS;
 110         termchar = '\0';
 111         while ((ch = getopt(argc, argv, "dft:")) != -1)
 112                 switch(ch) {
 113                 case 'd':
 114                         dflag = 1;
 115                         break;
 116                 case 'f':
 117                         fflag = 1;
 118                         break;
 119                 case 't':
 120                         termchar = *optarg;
 121                         break;
 122                 case '?':
 123                 default:
 124                         usage();
 125                 }
 126         argc -= optind;
 127         argv += optind;
 128
 129         switch (argc) {
 130         case 2:                         /* Don't set -df for user. */
 131                 string = *argv++;
 132                 file = *argv;
 133                 break;
 134         case 1:                         /* But set -df by default. */
 135                 dflag = fflag = 1;
 136                 string = *argv;
 137                 break;
 138         default:
 139                 usage();
 140         }
 141
 142         if (termchar != '\0' && (p = strchr(string, termchar)) != NULL)
 143                 *++p = '\0';
 144
 145         if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
 146                 err(2, "%s", file);
 147         len = (size_t)sb.st_size;
 148         if ((off_t)len != sb.st_size) {
 149                 errno = EFBIG;
 150                 err(2, "%s", file);
 151         }
 152         if ((front = mmap(NULL, len,
 153             PROT_READ, MAP_FILE|MAP_SHARED, fd, (off_t)0)) == NULL)
 154                 err(2, "%s", file);
 155         back = front + len;
 156         exit(look(string, front, back));
 157 }
 158
 159 int
 160 look(string, front, back)
 161         char *string, *front, *back;
 162 {
 163         int ch;
 164         char *readp, *writep;
 165
 166         /* Reformat string string to avoid doing it multiple times later. */
 167         for (readp = writep = string; (ch = *readp++) != 0; ) {
 168                 if (fflag)
 169                         ch = FOLD(ch);
 170                 if (dflag)
 171                         ch = DICT(ch);
 172                 if (ch != NO_COMPARE)
 173                         *(writep++) = ch;
 174         }
 175         *writep = '\0';
 176
 177         front = binary_search(string, front, back);
 178         front = linear_search(string, front, back);
 179
 180         if (front)
 181                 print_from(string, front, back);
 182         return (front ? 0 : 1);
 183 }
 184
 185
 186 /*
 187  * Binary search for "string" in memory between "front" and "back".
 188  *
 189  * This routine is expected to return a pointer to the start of a line at
 190  * *or before* the first word matching "string".  Relaxing the constraint
 191  * this way simplifies the algorithm.
 192  *
 193  * Invariants:
 194  *      front points to the beginning of a line at or before the first
 195  *      matching string.
 196  *
 197  *      back points to the beginning of a line at or after the first
 198  *      matching line.
 199  *
 200  * Base of the Invariants.
 201  *      front = NULL;
 202  *      back = EOF;
 203  *
 204  * Advancing the Invariants:
 205  *
 206  *      p = first newline after halfway point from front to back.
 207  *
 208  *      If the string at "p" is not greater than the string to match,
 209  *      p is the new front.  Otherwise it is the new back.
 210  *
 211  * Termination:
 212  *
 213  *      The definition of the routine allows it return at any point,
 214  *      since front is always at or before the line to print.
 215  *
 216  *      In fact, it returns when the chosen "p" equals "back".  This
 217  *      implies that there exists a string is least half as long as
 218  *      (back - front), which in turn implies that a linear search will
 219  *      be no more expensive than the cost of simply printing a string or two.
 220  *
 221  *      Trying to continue with binary search at this point would be
 222  *      more trouble than it's worth.
 223  */
 224 #define SKIP_PAST_NEWLINE(p, back) \
 225         while (p < back && *p++ != '\n');
 226
 227 char *
 228 binary_search(string, front, back)
 229         char *string, *front, *back;
 230 {
 231         char *p;
 232
 233         p = front + (back - front) / 2;
 234         SKIP_PAST_NEWLINE(p, back);
 235
 236         /*
 237          * If the file changes underneath us, make sure we don't
 238          * infinitely loop.
 239          */
 240         while (p < back && back > front) {
 241                 if (compare(string, p, back) == GREATER)
 242                         front = p;
 243                 else
 244                         back = p;
 245                 p = front + (back - front) / 2;
 246                 SKIP_PAST_NEWLINE(p, back);
 247         }
 248         return (front);
 249 }
 250
 251 /*
 252  * Find the first line that starts with string, linearly searching from front
 253  * to back.
 254  *
 255  * Return NULL for no such line.
 256  *
 257  * This routine assumes:
 258  *
 259  *      o front points at the first character in a line.
 260  *      o front is before or at the first line to be printed.
 261  */
 262 char *
 263 linear_search(string, front, back)
 264         char *string, *front, *back;
 265 {
 266         while (front < back) {
 267                 switch (compare(string, front, back)) {
 268                 case EQUAL:             /* Found it. */
 269                         return (front);
 270                         break;
 271                 case LESS:              /* No such string. */
 272                         return (NULL);
 273                         break;
 274                 case GREATER:           /* Keep going. */
 275                         break;
 276                 }
 277                 SKIP_PAST_NEWLINE(front, back);
 278         }
 279         return (NULL);
 280 }
 281
 282 /*
 283  * Print as many lines as match string, starting at front.
 284  */
 285 void
 286 print_from(string, front, back)
 287         char *string, *front, *back;
 288 {
 289         for (; front < back && compare(string, front, back) == EQUAL; ++front) {
 290                 for (; front < back && *front != '\n'; ++front)
 291                         if (putchar(*front) == EOF)
 292                                 err(2, "stdout");
 293                 if (putchar('\n') == EOF)
 294                         err(2, "stdout");
 295         }
 296 }
 297
 298 /*
 299  * Return LESS, GREATER, or EQUAL depending on how the string1 compares with
 300  * string2 (s1 ??? s2).
 301  *
 302  *      o Matches up to len(s1) are EQUAL.
 303  *      o Matches up to len(s2) are GREATER.
 304  *
 305  * Compare understands about the -f and -d flags, and treats comparisons
 306  * appropriately.
 307  *
 308  * The string "s1" is null terminated.  The string s2 is '\n' terminated (or
 309  * "back" terminated).
 310  */
 311 int
 312 compare(s1, s2, back)
 313         char *s1, *s2, *back;
 314 {
 315         int ch;
 316
 317         for (; *s1 && s2 < back && *s2 != '\n'; ++s1, ++s2) {
 318                 ch = *s2;
 319                 if (fflag)
 320                         ch = FOLD(ch);
 321                 if (dflag)
 322                         ch = DICT(ch);
 323
 324                 if (ch == NO_COMPARE) {
 325                         ++s2;           /* Ignore character in comparison. */
 326                         continue;
 327                 }
 328                 if (*s1 != ch)
 329                         return (*s1 < ch ? LESS : GREATER);
 330         }
 331         return (*s1 ? GREATER : EQUAL);
 332 }
 333
 334 void
 335 usage()
 336 {
 337         (void)fprintf(stderr, "usage: look [-df] [-t char] string [file]\n");
 338         exit(2);
 339 }