src/comm.c

   1 /* comm -- compare two sorted files line by line.
   2    Copyright (C) 86, 90, 91, 1995-2000 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 Foundation,
  16    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 #include <stdio.h>
  23 #include <getopt.h>
  24 #include <sys/types.h>
  25 #include "system.h"
  26 #include "closeout.h"
  27 #include "linebuffer.h"
  28 #include "error.h"
  29 #include "hard-locale.h"
  30 #include "memcoll.h"
  31
  32 /* The official name of this program (e.g., no `g' prefix).  */
  33 #define PROGRAM_NAME "comm"
  34
  35 #define AUTHORS "Richard Stallman and David MacKenzie"
  36
  37 /* Undefine, to avoid warning about redefinition on some systems.  */
  38 #undef min
  39 #define min(x, y) ((x) < (y) ? (x) : (y))
  40
  41 /* The name this program was run with. */
  42 char *program_name;
  43
  44 #ifdef ENABLE_NLS
  45 /* Nonzero if the LC_COLLATE locale is hard.  */
  46 static int hard_LC_COLLATE;
  47 #endif
  48
  49 /* If nonzero, print lines that are found only in file 1. */
  50 static int only_file_1;
  51
  52 /* If nonzero, print lines that are found only in file 2. */
  53 static int only_file_2;
  54
  55 /* If nonzero, print lines that are found in both files. */
  56 static int both;
  57
  58 static struct option const long_options[] =
  59 {
  60   {GETOPT_HELP_OPTION_DECL},
  61   {GETOPT_VERSION_OPTION_DECL},
  62   {0, 0, 0, 0}
  63 };
  64
  65 \f
  66
  67 void
  68 usage (int status)
  69 {
  70   if (status != 0)
  71     fprintf (stderr, _("Try `%s --help' for more information.\n"),
  72              program_name);
  73   else
  74     {
  75       printf (_("\
  76 Usage: %s [OPTION]... LEFT_FILE RIGHT_FILE\n\
  77 "),
  78               program_name);
  79       printf (_("\
  80 Compare sorted files LEFT_FILE and RIGHT_FILE line by line.\n\
  81 \n\
  82   -1              suppress lines unique to left file\n\
  83   -2              suppress lines unique to right file\n\
  84   -3              suppress lines unique to both files\n\
  85       --help      display this help and exit\n\
  86       --version   output version information and exit\n\
  87 "));
  88       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
  89     }
  90   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
  91 }
  92
  93 /* Output the line in linebuffer LINE to stream STREAM
  94    provided the switches say it should be output.
  95    CLASS is 1 for a line found only in file 1,
  96    2 for a line only in file 2, 3 for a line in both. */
  97
  98 static void
  99 writeline (const struct linebuffer *line, FILE *stream, int class)
 100 {
 101   switch (class)
 102     {
 103     case 1:
 104       if (!only_file_1)
 105         return;
 106       break;
 107
 108     case 2:
 109       if (!only_file_2)
 110         return;
 111       /* Print a TAB if we are printing lines from file 1.  */
 112       if (only_file_1)
 113         putc ('\t', stream);
 114       break;
 115
 116     case 3:
 117       if (!both)
 118         return;
 119       /* Print a TAB if we are printing lines from file 1.  */
 120       if (only_file_1)
 121         putc ('\t', stream);
 122       /* Print a TAB if we are printing lines from file 2.  */
 123       if (only_file_2)
 124         putc ('\t', stream);
 125       break;
 126     }
 127
 128   fwrite (line->buffer, sizeof (char), line->length, stream);
 129 }
 130
 131 /* Compare INFILES[0] and INFILES[1].
 132    If either is "-", use the standard input for that file.
 133    Assume that each input file is sorted;
 134    merge them and output the result.
 135    Return 0 if successful, 1 if any errors occur. */
 136
 137 static int
 138 compare_files (char **infiles)
 139 {
 140   /* For each file, we have one linebuffer in lb1.  */
 141   struct linebuffer lb1[2];
 142
 143   /* thisline[i] points to the linebuffer holding the next available line
 144      in file i, or is NULL if there are no lines left in that file.  */
 145   struct linebuffer *thisline[2];
 146
 147   /* streams[i] holds the input stream for file i.  */
 148   FILE *streams[2];
 149
 150   int i, ret = 0;
 151
 152   /* Initialize the storage. */
 153   for (i = 0; i < 2; i++)
 154     {
 155       initbuffer (&lb1[i]);
 156       thisline[i] = &lb1[i];
 157       streams[i] = (STREQ (infiles[i], "-") ? stdin : fopen (infiles[i], "r"));
 158       if (!streams[i])
 159         {
 160           error (0, errno, "%s", infiles[i]);
 161           return 1;
 162         }
 163
 164       thisline[i] = readline (thisline[i], streams[i]);
 165     }
 166
 167   while (thisline[0] || thisline[1])
 168     {
 169       int order;
 170
 171       /* Compare the next available lines of the two files.  */
 172
 173       if (!thisline[0])
 174         order = 1;
 175       else if (!thisline[1])
 176         order = -1;
 177       else
 178         {
 179 #ifdef ENABLE_NLS
 180           if (hard_LC_COLLATE)
 181             order = memcoll (thisline[0]->buffer, thisline[0]->length - 1,
 182                              thisline[1]->buffer, thisline[1]->length - 1);
 183           else
 184 #endif
 185             {
 186               size_t len = min (thisline[0]->length, thisline[1]->length) - 1;
 187               order = memcmp (thisline[0]->buffer, thisline[1]->buffer, len);
 188               if (order == 0)
 189                 order = thisline[0]->length - thisline[1]->length;
 190             }
 191         }
 192
 193       /* Output the line that is lesser. */
 194       if (order == 0)
 195         writeline (thisline[1], stdout, 3);
 196       else if (order > 0)
 197         writeline (thisline[1], stdout, 2);
 198       else
 199         writeline (thisline[0], stdout, 1);
 200
 201       /* Step the file the line came from.
 202          If the files match, step both files.  */
 203       if (order >= 0)
 204         thisline[1] = readline (thisline[1], streams[1]);
 205       if (order <= 0)
 206         thisline[0] = readline (thisline[0], streams[0]);
 207     }
 208
 209   /* Free all storage and close all input streams. */
 210   for (i = 0; i < 2; i++)
 211     {
 212       free (lb1[i].buffer);
 213       if (ferror (streams[i]) || fclose (streams[i]) == EOF)
 214         {
 215           error (0, errno, "%s", infiles[i]);
 216           ret = 1;
 217         }
 218     }
 219   return ret;
 220 }
 221
 222 int
 223 main (int argc, char **argv)
 224 {
 225   int c;
 226
 227   program_name = argv[0];
 228   setlocale (LC_ALL, "");
 229   bindtextdomain (PACKAGE, LOCALEDIR);
 230   textdomain (PACKAGE);
 231
 232   atexit (close_stdout);
 233
 234 #ifdef ENABLE_NLS
 235   hard_LC_COLLATE = hard_locale (LC_COLLATE);
 236 #endif
 237
 238   only_file_1 = 1;
 239   only_file_2 = 1;
 240   both = 1;
 241
 242   while ((c = getopt_long (argc, argv, "123", long_options, NULL)) != -1)
 243     switch (c)
 244       {
 245       case 0:
 246         break;
 247
 248       case '1':
 249         only_file_1 = 0;
 250         break;
 251
 252       case '2':
 253         only_file_2 = 0;
 254         break;
 255
 256       case '3':
 257         both = 0;
 258         break;
 259
 260       case_GETOPT_HELP_CHAR;
 261
 262       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 263
 264       default:
 265         usage (1);
 266       }
 267
 268   if (optind + 2 != argc)
 269     usage (1);
 270
 271   exit (compare_files (argv + optind) == 0
 272         ? EXIT_SUCCESS : EXIT_FAILURE);
 273 }