usr.bin/uniq/uniq.c

   1 /*
   2  * Copyright (c) 1989, 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  * Case Larsen.
   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  * 4. 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
  33 #ifndef lint
  34 static const char copyright[] =
  35 "@(#) Copyright (c) 1989, 1993\n\
  36         The Regents of the University of California.  All rights reserved.\n";
  37 #endif /* not lint */
  38
  39 #ifndef lint
  40 #if 0
  41 static char sccsid[] = "@(#)uniq.c      8.3 (Berkeley) 5/4/95";
  42 #endif
  43 static const char rcsid[] =
  44   "$FreeBSD$";
  45 #endif /* not lint */
  46
  47 #include <sys/capsicum.h>
  48
  49 #include <ctype.h>
  50 #include <err.h>
  51 #include <errno.h>
  52 #include <limits.h>
  53 #include <locale.h>
  54 #include <nl_types.h>
  55 #include <stdint.h>
  56 #define _WITH_GETLINE
  57 #include <stdio.h>
  58 #include <stdlib.h>
  59 #include <string.h>
  60 #include <termios.h>
  61 #include <unistd.h>
  62 #include <wchar.h>
  63 #include <wctype.h>
  64
  65 static int cflag, dflag, uflag, iflag;
  66 static int numchars, numfields, repeats;
  67
  68 static FILE     *file(const char *, const char *);
  69 static wchar_t  *convert(const char *);
  70 static int       inlcmp(const char *, const char *);
  71 static void      show(FILE *, const char *);
  72 static wchar_t  *skip(wchar_t *);
  73 static void      obsolete(char *[]);
  74 static void      usage(void);
  75
  76 static void
  77 strerror_init(void)
  78 {
  79
  80         /*
  81          * Cache NLS data before entering capability mode.
  82          * XXXPJD: There should be strerror_init() and strsignal_init() in libc.
  83          */
  84         (void)catopen("libc", NL_CAT_LOCALE);
  85 }
  86
  87 int
  88 main (int argc, char *argv[])
  89 {
  90         wchar_t *tprev, *tthis;
  91         FILE *ifp, *ofp;
  92         int ch, comp;
  93         size_t prevbuflen, thisbuflen, b1;
  94         char *prevline, *thisline, *p;
  95         const char *ifn;
  96         cap_rights_t rights;
  97
  98         (void) setlocale(LC_ALL, "");
  99
 100         obsolete(argv);
 101         while ((ch = getopt(argc, argv, "cdif:s:u")) != -1)
 102                 switch (ch) {
 103                 case 'c':
 104                         cflag = 1;
 105                         break;
 106                 case 'd':
 107                         dflag = 1;
 108                         break;
 109                 case 'i':
 110                         iflag = 1;
 111                         break;
 112                 case 'f':
 113                         numfields = strtol(optarg, &p, 10);
 114                         if (numfields < 0 || *p)
 115                                 errx(1, "illegal field skip value: %s", optarg);
 116                         break;
 117                 case 's':
 118                         numchars = strtol(optarg, &p, 10);
 119                         if (numchars < 0 || *p)
 120                                 errx(1, "illegal character skip value: %s", optarg);
 121                         break;
 122                 case 'u':
 123                         uflag = 1;
 124                         break;
 125                 case '?':
 126                 default:
 127                         usage();
 128                 }
 129
 130         argc -= optind;
 131         argv += optind;
 132
 133         /* If no flags are set, default is -d -u. */
 134         if (cflag) {
 135                 if (dflag || uflag)
 136                         usage();
 137         } else if (!dflag && !uflag)
 138                 dflag = uflag = 1;
 139
 140         if (argc > 2)
 141                 usage();
 142
 143         ifp = stdin;
 144         ifn = "stdin";
 145         ofp = stdout;
 146         if (argc > 0 && strcmp(argv[0], "-") != 0)
 147                 ifp = file(ifn = argv[0], "r");
 148         cap_rights_init(&rights, CAP_FSTAT, CAP_READ);
 149         if (cap_rights_limit(fileno(ifp), &rights) < 0 && errno != ENOSYS)
 150                 err(1, "unable to limit rights for %s", ifn);
 151         cap_rights_init(&rights, CAP_FSTAT, CAP_WRITE);
 152         if (argc > 1)
 153                 ofp = file(argv[1], "w");
 154         else
 155                 cap_rights_set(&rights, CAP_IOCTL);
 156         if (cap_rights_limit(fileno(ofp), &rights) < 0 && errno != ENOSYS) {
 157                 err(1, "unable to limit rights for %s",
 158                     argc > 1 ? argv[1] : "stdout");
 159         }
 160         if (cap_rights_is_set(&rights, CAP_IOCTL)) {
 161                 unsigned long cmd;
 162
 163                 cmd = TIOCGETA; /* required by isatty(3) in printf(3) */
 164
 165                 if (cap_ioctls_limit(fileno(ofp), &cmd, 1) < 0 &&
 166                     errno != ENOSYS) {
 167                         err(1, "unable to limit ioctls for %s",
 168                             argc > 1 ? argv[1] : "stdout");
 169                 }
 170         }
 171
 172         strerror_init();
 173         if (cap_enter() < 0 && errno != ENOSYS)
 174                 err(1, "unable to enter capability mode");
 175
 176         prevbuflen = thisbuflen = 0;
 177         prevline = thisline = NULL;
 178
 179         if (getline(&prevline, &prevbuflen, ifp) < 0) {
 180                 if (ferror(ifp))
 181                         err(1, "%s", ifn);
 182                 exit(0);
 183         }
 184         tprev = convert(prevline);
 185
 186         if (!cflag && uflag && dflag)
 187                 show(ofp, prevline);
 188
 189         tthis = NULL;
 190         while (getline(&thisline, &thisbuflen, ifp) >= 0) {
 191                 if (tthis != NULL)
 192                         free(tthis);
 193                 tthis = convert(thisline);
 194
 195                 if (tthis == NULL && tprev == NULL)
 196                         comp = inlcmp(thisline, prevline);
 197                 else if (tthis == NULL || tprev == NULL)
 198                         comp = 1;
 199                 else
 200                         comp = wcscoll(tthis, tprev);
 201
 202                 if (comp) {
 203                         /* If different, print; set previous to new value. */
 204                         if (cflag || !dflag || !uflag)
 205                                 show(ofp, prevline);
 206                         p = prevline;
 207                         b1 = prevbuflen;
 208                         prevline = thisline;
 209                         prevbuflen = thisbuflen;
 210                         if (tprev != NULL)
 211                                 free(tprev);
 212                         tprev = tthis;
 213                         if (!cflag && uflag && dflag)
 214                                 show(ofp, prevline);
 215                         thisline = p;
 216                         thisbuflen = b1;
 217                         tthis = NULL;
 218                         repeats = 0;
 219                 } else
 220                         ++repeats;
 221         }
 222         if (ferror(ifp))
 223                 err(1, "%s", ifn);
 224         if (cflag || !dflag || !uflag)
 225                 show(ofp, prevline);
 226         exit(0);
 227 }
 228
 229 static wchar_t *
 230 convert(const char *str)
 231 {
 232         size_t n;
 233         wchar_t *buf, *ret, *p;
 234
 235         if ((n = mbstowcs(NULL, str, 0)) == (size_t)-1)
 236                 return (NULL);
 237         if (SIZE_MAX / sizeof(*buf) < n + 1)
 238                 errx(1, "conversion buffer length overflow");
 239         if ((buf = malloc((n + 1) * sizeof(*buf))) == NULL)
 240                 err(1, "malloc");
 241         if (mbstowcs(buf, str, n + 1) != n)
 242                 errx(1, "internal mbstowcs() error");
 243         /* The last line may not end with \n. */
 244         if (n > 0 && buf[n - 1] == L'\n')
 245                 buf[n - 1] = L'\0';
 246
 247         /* If requested get the chosen fields + character offsets. */
 248         if (numfields || numchars) {
 249                 if ((ret = wcsdup(skip(buf))) == NULL)
 250                         err(1, "wcsdup");
 251                 free(buf);
 252         } else
 253                 ret = buf;
 254
 255         if (iflag) {
 256                 for (p = ret; *p != L'\0'; p++)
 257                         *p = towlower(*p);
 258         }
 259
 260         return (ret);
 261 }
 262
 263 static int
 264 inlcmp(const char *s1, const char *s2)
 265 {
 266         int c1, c2;
 267
 268         while (*s1 == *s2++)
 269                 if (*s1++ == '\0')
 270                         return (0);
 271         c1 = (unsigned char)*s1;
 272         c2 = (unsigned char)*(s2 - 1);
 273         /* The last line may not end with \n. */
 274         if (c1 == '\n')
 275                 c1 = '\0';
 276         if (c2 == '\n')
 277                 c2 = '\0';
 278         return (c1 - c2);
 279 }
 280
 281 /*
 282  * show --
 283  *      Output a line depending on the flags and number of repetitions
 284  *      of the line.
 285  */
 286 static void
 287 show(FILE *ofp, const char *str)
 288 {
 289
 290         if (cflag)
 291                 (void)fprintf(ofp, "%4d %s", repeats + 1, str);
 292         if ((dflag && repeats) || (uflag && !repeats))
 293                 (void)fprintf(ofp, "%s", str);
 294 }
 295
 296 static wchar_t *
 297 skip(wchar_t *str)
 298 {
 299         int nchars, nfields;
 300
 301         for (nfields = 0; *str != L'\0' && nfields++ != numfields; ) {
 302                 while (iswblank(*str))
 303                         str++;
 304                 while (*str != L'\0' && !iswblank(*str))
 305                         str++;
 306         }
 307         for (nchars = numchars; nchars-- && *str != L'\0'; ++str)
 308                 ;
 309         return(str);
 310 }
 311
 312 static FILE *
 313 file(const char *name, const char *mode)
 314 {
 315         FILE *fp;
 316
 317         if ((fp = fopen(name, mode)) == NULL)
 318                 err(1, "%s", name);
 319         return(fp);
 320 }
 321
 322 static void
 323 obsolete(char *argv[])
 324 {
 325         int len;
 326         char *ap, *p, *start;
 327
 328         while ((ap = *++argv)) {
 329                 /* Return if "--" or not an option of any form. */
 330                 if (ap[0] != '-') {
 331                         if (ap[0] != '+')
 332                                 return;
 333                 } else if (ap[1] == '-')
 334                         return;
 335                 if (!isdigit((unsigned char)ap[1]))
 336                         continue;
 337                 /*
 338                  * Digit signifies an old-style option.  Malloc space for dash,
 339                  * new option and argument.
 340                  */
 341                 len = strlen(ap);
 342                 if ((start = p = malloc(len + 3)) == NULL)
 343                         err(1, "malloc");
 344                 *p++ = '-';
 345                 *p++ = ap[0] == '+' ? 's' : 'f';
 346                 (void)strcpy(p, ap + 1);
 347                 *argv = start;
 348         }
 349 }
 350
 351 static void
 352 usage(void)
 353 {
 354         (void)fprintf(stderr,
 355 "usage: uniq [-c | -d | -u] [-i] [-f fields] [-s chars] [input [output]]\n");
 356         exit(1);
 357 }