usr.bin/look/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. Neither the name of the University nor the names of its contributors
  17  *    may be used to endorse or promote products derived from this software
  18  *    without specific prior written permission.
  19  *
  20  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  21  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  23  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  24  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  25  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  26  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  27  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  28  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  29  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  30  * SUCH DAMAGE.
  31  *
  32  * @(#) Copyright (c) 1991, 1993 The Regents of the University of California.  All rights reserved.
  33  * @(#)look.c   8.2 (Berkeley) 5/4/95
  34  * $FreeBSD: head/usr.bin/look/look.c 268242 2014-07-04 04:47:29Z eadler $
  35  */
  36
  37 /*
  38  * look -- find lines in a sorted list.
  39  *
  40  * The man page said that TABs and SPACEs participate in -d comparisons.
  41  * In fact, they were ignored.  This implements historic practice, not
  42  * the manual page.
  43  */
  44
  45 #include <sys/types.h>
  46 #include <sys/mman.h>
  47 #include <sys/stat.h>
  48
  49 #include <err.h>
  50 #include <errno.h>
  51 #include <fcntl.h>
  52 #include <getopt.h>
  53 #include <limits.h>
  54 #include <locale.h>
  55 #include <stdint.h>
  56 #include <stdio.h>
  57 #include <stdlib.h>
  58 #include <string.h>
  59 #include <unistd.h>
  60 #include <wchar.h>
  61 #include <wctype.h>
  62
  63 #include "pathnames.h"
  64
  65 static char _path_words[] = _PATH_WORDS;
  66
  67 #define EQUAL           0
  68 #define GREATER         1
  69 #define LESS            (-1)
  70
  71 static int dflag, fflag;
  72
  73 static char     *binary_search(wchar_t *, unsigned char *, unsigned char *);
  74 static int       compare(wchar_t *, unsigned char *, unsigned char *);
  75 static char     *linear_search(wchar_t *, unsigned char *, unsigned char *);
  76 static int       look(wchar_t *, unsigned char *, unsigned char *);
  77 static wchar_t  *prepkey(const char *, wchar_t);
  78 static void      print_from(wchar_t *, unsigned char *, unsigned char *);
  79
  80 static void usage(void);
  81
  82 static struct option longopts[] = {
  83         { "alternative",no_argument,    NULL, 'a' },
  84         { "alphanum",   no_argument,    NULL, 'd' },
  85         { "ignore-case",no_argument,    NULL, 'i' },
  86         { "terminate",  required_argument, NULL, 't'},
  87         { NULL,         0,              NULL, 0 },
  88 };
  89
  90 int
  91 main(int argc, char *argv[])
  92 {
  93         struct stat sb;
  94         int ch, fd, match;
  95         wchar_t termchar;
  96         unsigned char *back, *front;
  97         unsigned const char *file;
  98         wchar_t *key;
  99
 100         setlocale(LC_CTYPE, "");
 101
 102         file = _path_words;
 103         termchar = L'\0';
 104         while ((ch = getopt_long(argc, argv, "+adft:", longopts, NULL)) != -1)
 105                 switch(ch) {
 106                 case 'a':
 107                         /* COMPATIBILITY */
 108                         break;
 109                 case 'd':
 110                         dflag = 1;
 111                         break;
 112                 case 'f':
 113                         fflag = 1;
 114                         break;
 115                 case 't':
 116                         if (mbrtowc(&termchar, optarg, MB_LEN_MAX, NULL) !=
 117                             strlen(optarg))
 118                                 errx(2, "invalid termination character");
 119                         break;
 120                 case '?':
 121                 default:
 122                         usage();
 123                 }
 124         argc -= optind;
 125         argv += optind;
 126
 127         if (argc == 0)
 128                 usage();
 129         if (argc == 1)                  /* But set -df by default. */
 130                 dflag = fflag = 1;
 131         key = prepkey(*argv++, termchar);
 132         if (argc >= 2)
 133                 file = *argv++;
 134
 135         match = 1;
 136
 137         do {
 138                 if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
 139                         err(2, "%s", file);
 140                 if ((uintmax_t)sb.st_size > (uintmax_t)SIZE_T_MAX)
 141                         errx(2, "%s: %s", file, strerror(EFBIG));
 142                 if (sb.st_size == 0) {
 143                         close(fd);
 144                         continue;
 145                 }
 146                 if ((front = mmap(NULL, (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED)
 147                         err(2, "%s", file);
 148                 back = front + sb.st_size;
 149                 match *= (look(key, front, back));
 150                 close(fd);
 151         } while (argc-- > 2 && (file = *argv++));
 152
 153         exit(match);
 154 }
 155
 156 static wchar_t *
 157 prepkey(const char *string, wchar_t termchar)
 158 {
 159         const char *readp;
 160         wchar_t *key, *writep;
 161         wchar_t ch;
 162         size_t clen;
 163
 164         /*
 165          * Reformat search string and convert to wide character representation
 166          * to avoid doing it multiple times later.
 167          */
 168         if ((key = malloc(sizeof(wchar_t) * (strlen(string) + 1))) == NULL)
 169                 err(2, NULL);
 170         readp = string;
 171         writep = key;
 172         while ((clen = mbrtowc(&ch, readp, MB_LEN_MAX, NULL)) != 0) {
 173                 if (clen == (size_t)-1 || clen == (size_t)-2)
 174                         errc(2, EILSEQ, NULL);
 175                 if (fflag)
 176                         ch = towlower(ch);
 177                 if (!dflag || iswalnum(ch))
 178                         *writep++ = ch;
 179                 readp += clen;
 180         }
 181         *writep = L'\0';
 182         if (termchar != L'\0' && (writep = wcschr(key, termchar)) != NULL)
 183                 *++writep = L'\0';
 184         return (key);
 185 }
 186
 187 static int
 188 look(wchar_t *string, unsigned char *front, unsigned char *back)
 189 {
 190         front = binary_search(string, front, back);
 191         front = linear_search(string, front, back);
 192
 193         if (front)
 194                 print_from(string, front, back);
 195         return (front ? 0 : 1);
 196 }
 197
 198
 199 /*
 200  * Binary search for "string" in memory between "front" and "back".
 201  *
 202  * This routine is expected to return a pointer to the start of a line at
 203  * *or before* the first word matching "string".  Relaxing the constraint
 204  * this way simplifies the algorithm.
 205  *
 206  * Invariants:
 207  *      front points to the beginning of a line at or before the first
 208  *      matching string.
 209  *
 210  *      back points to the beginning of a line at or after the first
 211  *      matching line.
 212  *
 213  * Base of the Invariants.
 214  *      front = NULL;
 215  *      back = EOF;
 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(wchar_t *string, unsigned char *front, unsigned char *back)
 242 {
 243         unsigned 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(string, 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(wchar_t *string, unsigned char *front, unsigned char *back)
 276 {
 277         while (front < back) {
 278                 switch (compare(string, 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(wchar_t *string, unsigned char *front, unsigned char *back)
 296 {
 297         for (; front < back && compare(string, front, back) == EQUAL; ++front) {
 298                 for (; front < back && *front != '\n'; ++front)
 299                         if (putchar(*front) == EOF)
 300                                 err(2, "stdout");
 301                 if (putchar('\n') == EOF)
 302                         err(2, "stdout");
 303         }
 304 }
 305
 306 /*
 307  * Return LESS, GREATER, or EQUAL depending on how the string1 compares with
 308  * string2 (s1 ??? s2).
 309  *
 310  *      o Matches up to len(s1) are EQUAL.
 311  *      o Matches up to len(s2) are GREATER.
 312  *
 313  * Compare understands about the -f and -d flags, and treats comparisons
 314  * appropriately.
 315  *
 316  * The string "s1" is null terminated.  The string s2 is '\n' terminated (or
 317  * "back" terminated).
 318  */
 319 static int
 320 compare(wchar_t *s1, unsigned char *s2, unsigned char *back)
 321 {
 322         wchar_t ch1, ch2;
 323         size_t len2;
 324
 325         for (; *s1 && s2 < back && *s2 != '\n'; ++s1, s2 += len2) {
 326                 ch1 = *s1;
 327                 len2 = mbrtowc(&ch2, s2, back - s2, NULL);
 328                 if (len2 == (size_t)-1 || len2 == (size_t)-2) {
 329                         ch2 = *s2;
 330                         len2 = 1;
 331                 }
 332                 if (fflag)
 333                         ch2 = towlower(ch2);
 334                 if (dflag && !iswalnum(ch2)) {
 335                         /* Ignore character in comparison. */
 336                         --s1;
 337                         continue;
 338                 }
 339                 if (ch1 != ch2)
 340                         return (ch1 < ch2 ? LESS : GREATER);
 341         }
 342         return (*s1 ? GREATER : EQUAL);
 343 }
 344
 345 static void
 346 usage(void)
 347 {
 348         fprintf(stderr, "usage: look [-df] [-t char] string [file ...]\n");
 349         exit(2);
 350 }