src/uniq.c

   1 /* uniq -- remove duplicate lines from a sorted file
   2    Copyright (C) 1986, 1991, 1995 Free Software Foundation, Inc.
   3
   4    This program is free software; you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation; either version 2, or (at your option)
   7    any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program; if not, write to the Free Software
  16    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  17
  18 /* Written by Richard Stallman and David MacKenzie. */
  19 \f
  20 #include <config.h>
  21
  22 /* Get isblank from GNU libc.  */
  23 #define _GNU_SOURCE
  24
  25 #include <stdio.h>
  26 #include <getopt.h>
  27 #include <sys/types.h>
  28
  29 #if HAVE_LIMITS_H
  30 # include <limits.h>
  31 #endif
  32
  33 #ifndef UINT_MAX
  34 # define UINT_MAX ((unsigned int) ~(unsigned int) 0)
  35 #endif
  36
  37 #ifndef INT_MAX
  38 # define INT_MAX ((int) (UINT_MAX >> 1))
  39 #endif
  40
  41 #include "system.h"
  42 #include "linebuffer.h"
  43 #include "version.h"
  44 #include "error.h"
  45 #include "xstrtol.h"
  46
  47 /* Undefine, to avoid warning about redefinition on some systems.  */
  48 #undef min
  49 #define min(x, y) ((x) < (y) ? (x) : (y))
  50
  51 /* The name this program was run with. */
  52 char *program_name;
  53
  54 /* Number of fields to skip on each line when doing comparisons. */
  55 static int skip_fields;
  56
  57 /* Number of chars to skip after skipping any fields. */
  58 static int skip_chars;
  59
  60 /* Number of chars to compare; if 0, compare the whole lines. */
  61 static int check_chars;
  62
  63 enum countmode
  64 {
  65   count_occurrences,            /* -c Print count before output lines. */
  66   count_none                    /* Default.  Do not print counts. */
  67 };
  68
  69 /* Whether and how to precede the output lines with a count of the number of
  70    times they occurred in the input. */
  71 static enum countmode countmode;
  72
  73 enum output_mode
  74 {
  75   output_repeated,              /* -d Only lines that are repeated. */
  76   output_unique,                /* -u Only lines that are not repeated. */
  77   output_all                    /* Default.  Print first copy of each line. */
  78 };
  79
  80 /* Which lines to output. */
  81 static enum output_mode mode;
  82
  83 /* If nonzero, display usage information and exit.  */
  84 static int show_help;
  85
  86 /* If nonzero, print the version on standard output then exit.  */
  87 static int show_version;
  88
  89 static struct option const longopts[] =
  90 {
  91   {"count", no_argument, NULL, 'c'},
  92   {"repeated", no_argument, NULL, 'd'},
  93   {"unique", no_argument, NULL, 'u'},
  94   {"skip-fields", required_argument, NULL, 'f'},
  95   {"skip-chars", required_argument, NULL, 's'},
  96   {"check-chars", required_argument, NULL, 'w'},
  97   {"help", no_argument, &show_help, 1},
  98   {"version", no_argument, &show_version, 1},
  99   {NULL, 0, NULL, 0}
 100 };
 101
 102 static void
 103 usage (int status)
 104 {
 105   if (status != 0)
 106     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 107              program_name);
 108   else
 109     {
 110       printf (_("\
 111 Usage: %s [OPTION]... [INPUT [OUTPUT]]\n\
 112 "),
 113               program_name);
 114       printf (_("\
 115 Discard all but one of successive identical lines from INPUT (or\n\
 116 standard input), writing to OUTPUT (or standard output).\n\
 117 \n\
 118   -c, --count           prefix lines by the number of occurrences\n\
 119   -d, --repeated        only print duplicate lines\n\
 120   -f, --skip-fields=N   avoid comparing the first N fields\n\
 121   -s, --skip-chars=N    avoid comparing the first N characters\n\
 122   -u, --unique          only print unique lines\n\
 123   -w, --check-chars=N   compare no more than N characters in lines\n\
 124   -N                    same as -f N\n\
 125   +N                    same as -s N\n\
 126       --help            display this help and exit\n\
 127       --version         output version information and exit\n\
 128 \n\
 129 A field is a run of whitespace, than non-whitespace characters.\n\
 130 Fields are skipped before chars. \n\
 131 "));
 132     }
 133   exit (status);
 134 }
 135
 136 /* Given a linebuffer LINE,
 137    return a pointer to the beginning of the line's field to be compared. */
 138
 139 static char *
 140 find_field (const struct linebuffer *line)
 141 {
 142   register int count;
 143   register char *lp = line->buffer;
 144   register int size = line->length;
 145   register int i = 0;
 146
 147   for (count = 0; count < skip_fields && i < size; count++)
 148     {
 149       while (i < size && ISBLANK (lp[i]))
 150         i++;
 151       while (i < size && !ISBLANK (lp[i]))
 152         i++;
 153     }
 154
 155   for (count = 0; count < skip_chars && i < size; count++)
 156     i++;
 157
 158   return lp + i;
 159 }
 160
 161 /* Return zero if two strings OLD and NEW match, nonzero if not.
 162    OLD and NEW point not to the beginnings of the lines
 163    but rather to the beginnings of the fields to compare.
 164    OLDLEN and NEWLEN are their lengths. */
 165
 166 static int
 167 different (const char *old, const char *new, int oldlen, int newlen)
 168 {
 169   register int order;
 170
 171   if (check_chars)
 172     {
 173       if (oldlen > check_chars)
 174         oldlen = check_chars;
 175       if (newlen > check_chars)
 176         newlen = check_chars;
 177     }
 178   order = memcmp (old, new, min (oldlen, newlen));
 179   if (order == 0)
 180     return oldlen - newlen;
 181   return order;
 182 }
 183 \f
 184 /* Output the line in linebuffer LINE to stream STREAM
 185    provided that the switches say it should be output.
 186    If requested, print the number of times it occurred, as well;
 187    LINECOUNT + 1 is the number of times that the line occurred. */
 188
 189 static void
 190 writeline (const struct linebuffer *line, FILE *stream, int linecount)
 191 {
 192   if ((mode == output_unique && linecount != 0)
 193       || (mode == output_repeated && linecount == 0))
 194     return;
 195
 196   if (countmode == count_occurrences)
 197     fprintf (stream, "%7d\t", linecount + 1);
 198
 199   fwrite (line->buffer, sizeof (char), line->length, stream);
 200   putc ('\n', stream);
 201 }
 202
 203 /* Process input file INFILE with output to OUTFILE.
 204    If either is "-", use the standard I/O stream for it instead. */
 205
 206 static void
 207 check_file (const char *infile, const char *outfile)
 208 {
 209   FILE *istream;
 210   FILE *ostream;
 211   struct linebuffer lb1, lb2;
 212   struct linebuffer *thisline, *prevline, *exch;
 213   char *prevfield, *thisfield;
 214   int prevlen, thislen;
 215   int match_count = 0;
 216
 217   if (!strcmp (infile, "-"))
 218     istream = stdin;
 219   else
 220     istream = fopen (infile, "r");
 221   if (istream == NULL)
 222     error (1, errno, "%s", infile);
 223
 224   if (!strcmp (outfile, "-"))
 225     ostream = stdout;
 226   else
 227     ostream = fopen (outfile, "w");
 228   if (ostream == NULL)
 229     error (1, errno, "%s", outfile);
 230
 231   thisline = &lb1;
 232   prevline = &lb2;
 233
 234   initbuffer (thisline);
 235   initbuffer (prevline);
 236
 237   if (readline (prevline, istream) == 0)
 238     goto closefiles;
 239   prevfield = find_field (prevline);
 240   prevlen = prevline->length - (prevfield - prevline->buffer);
 241
 242   while (!feof (istream))
 243     {
 244       if (readline (thisline, istream) == 0)
 245         break;
 246       thisfield = find_field (thisline);
 247       thislen = thisline->length - (thisfield - thisline->buffer);
 248       if (!different (thisfield, prevfield, thislen, prevlen))
 249         match_count++;
 250       else
 251         {
 252           writeline (prevline, ostream, match_count);
 253           match_count = 0;
 254
 255           exch = prevline;
 256           prevline = thisline;
 257           thisline = exch;
 258           prevfield = thisfield;
 259           prevlen = thislen;
 260         }
 261     }
 262
 263   writeline (prevline, ostream, match_count);
 264
 265  closefiles:
 266   if (ferror (istream) || fclose (istream) == EOF)
 267     error (1, errno, _("error reading %s"), infile);
 268
 269   if (ferror (ostream) || fclose (ostream) == EOF)
 270     error (1, errno, _("error writing %s"), outfile);
 271
 272   free (lb1.buffer);
 273   free (lb2.buffer);
 274 }
 275
 276 void
 277 main (int argc, char **argv)
 278 {
 279   int optc;
 280   char *infile = "-", *outfile = "-";
 281
 282   program_name = argv[0];
 283   setlocale (LC_ALL, "");
 284   bindtextdomain (PACKAGE, LOCALEDIR);
 285   textdomain (PACKAGE);
 286
 287   skip_chars = 0;
 288   skip_fields = 0;
 289   check_chars = 0;
 290   mode = output_all;
 291   countmode = count_none;
 292
 293   while ((optc = getopt_long (argc, argv, "0123456789cdf:s:uw:", longopts,
 294                               (int *) 0)) != EOF)
 295     {
 296       switch (optc)
 297         {
 298         case 0:
 299           break;
 300
 301         case '0':
 302         case '1':
 303         case '2':
 304         case '3':
 305         case '4':
 306         case '5':
 307         case '6':
 308         case '7':
 309         case '8':
 310         case '9':
 311           skip_fields = skip_fields * 10 + optc - '0';
 312           break;
 313
 314         case 'c':
 315           countmode = count_occurrences;
 316           break;
 317
 318         case 'd':
 319           mode = output_repeated;
 320           break;
 321
 322         case 'f':               /* Like '-#'. */
 323           {
 324             long int tmp_long;
 325             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 326                 || tmp_long <= 0 || tmp_long > INT_MAX)
 327               error (1, 0, _("invalid number of fields to skip: `%s'"),
 328                      optarg);
 329             skip_fields = (int) tmp_long;
 330           }
 331           break;
 332
 333         case 's':               /* Like '+#'. */
 334           {
 335             long int tmp_long;
 336             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 337                 || tmp_long <= 0 || tmp_long > INT_MAX)
 338               error (1, 0, _("invalid number of bytes to skip: `%s'"),
 339                      optarg);
 340             skip_chars = (int) tmp_long;
 341           }
 342           break;
 343
 344         case 'u':
 345           mode = output_unique;
 346           break;
 347
 348         case 'w':
 349           {
 350             long int tmp_long;
 351             if (xstrtol (optarg, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 352                 || tmp_long <= 0 || tmp_long > INT_MAX)
 353               error (1, 0, _("invalid number of bytes to compare: `%s'"),
 354                      optarg);
 355             check_chars = (int) tmp_long;
 356           }
 357           break;
 358
 359         default:
 360           usage (1);
 361         }
 362     }
 363
 364   if (show_version)
 365     {
 366       printf ("uniq - %s\n", version_string);
 367       exit (0);
 368     }
 369
 370   if (show_help)
 371     usage (0);
 372
 373   if (optind >= 2 && strcmp (argv[optind - 1], "--") != 0)
 374     {
 375       /* Interpret non-option arguments with leading `+' only
 376          if we haven't seen `--'.  */
 377       while (optind < argc && argv[optind][0] == '+')
 378         {
 379           char *opt_str = argv[optind++];
 380           long int tmp_long;
 381           if (xstrtol (opt_str, NULL, 10, &tmp_long, NULL) != LONGINT_OK
 382               || tmp_long <= 0 || tmp_long > INT_MAX)
 383             error (1, 0, _("invalid number of bytes to compare: `%s'"),
 384                    opt_str);
 385           skip_chars = (int) tmp_long;
 386         }
 387     }
 388
 389   if (optind < argc)
 390     infile = argv[optind++];
 391
 392   if (optind < argc)
 393     outfile = argv[optind++];
 394
 395   if (optind < argc)
 396     usage (1);                  /* Extra arguments. */
 397
 398   check_file (infile, outfile);
 399
 400   exit (0);
 401 }