src/comm.c

   1 /* comm -- compare two sorted files line by line.
   2    Copyright (C) 86, 90, 91, 1995-2002 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 "xmemcoll.h"
  31
  32 /* The official name of this program (e.g., no `g' prefix).  */
  33 #define PROGRAM_NAME "comm"
  34
  35 #define AUTHORS N_ ("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 /* Nonzero if the LC_COLLATE locale is hard.  */
  45 static int hard_LC_COLLATE;
  46
  47 /* If nonzero, print lines that are found only in file 1. */
  48 static int only_file_1;
  49
  50 /* If nonzero, print lines that are found only in file 2. */
  51 static int only_file_2;
  52
  53 /* If nonzero, print lines that are found in both files. */
  54 static int both;
  55
  56 static struct option const long_options[] =
  57 {
  58   {GETOPT_HELP_OPTION_DECL},
  59   {GETOPT_VERSION_OPTION_DECL},
  60   {0, 0, 0, 0}
  61 };
  62
  63 \f
  64
  65 void
  66 usage (int status)
  67 {
  68   if (status != 0)
  69     fprintf (stderr, _("Try `%s --help' for more information.\n"),
  70              program_name);
  71   else
  72     {
  73       printf (_("\
  74 Usage: %s [OPTION]... LEFT_FILE RIGHT_FILE\n\
  75 "),
  76               program_name);
  77       fputs (_("\
  78 Compare sorted files LEFT_FILE and RIGHT_FILE line by line.\n\
  79 \n\
  80   -1              suppress lines unique to left file\n\
  81   -2              suppress lines unique to right file\n\
  82   -3              suppress lines that appear in both files\n\
  83 "), stdout);
  84       fputs (HELP_OPTION_DESCRIPTION, stdout);
  85       fputs (VERSION_OPTION_DESCRIPTION, stdout);
  86       printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
  87     }
  88   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
  89 }
  90
  91 /* Output the line in linebuffer LINE to stream STREAM
  92    provided the switches say it should be output.
  93    CLASS is 1 for a line found only in file 1,
  94    2 for a line only in file 2, 3 for a line in both. */
  95
  96 static void
  97 writeline (const struct linebuffer *line, FILE *stream, int class)
  98 {
  99   switch (class)
 100     {
 101     case 1:
 102       if (!only_file_1)
 103         return;
 104       break;
 105
 106     case 2:
 107       if (!only_file_2)
 108         return;
 109       /* Print a TAB if we are printing lines from file 1.  */
 110       if (only_file_1)
 111         putc ('\t', stream);
 112       break;
 113
 114     case 3:
 115       if (!both)
 116         return;
 117       /* Print a TAB if we are printing lines from file 1.  */
 118       if (only_file_1)
 119         putc ('\t', stream);
 120       /* Print a TAB if we are printing lines from file 2.  */
 121       if (only_file_2)
 122         putc ('\t', stream);
 123       break;
 124     }
 125
 126   fwrite (line->buffer, sizeof (char), line->length, stream);
 127 }
 128
 129 /* Compare INFILES[0] and INFILES[1].
 130    If either is "-", use the standard input for that file.
 131    Assume that each input file is sorted;
 132    merge them and output the result.
 133    Return 0 if successful, 1 if any errors occur. */
 134
 135 static int
 136 compare_files (char **infiles)
 137 {
 138   /* For each file, we have one linebuffer in lb1.  */
 139   struct linebuffer lb1[2];
 140
 141   /* thisline[i] points to the linebuffer holding the next available line
 142      in file i, or is NULL if there are no lines left in that file.  */
 143   struct linebuffer *thisline[2];
 144
 145   /* streams[i] holds the input stream for file i.  */
 146   FILE *streams[2];
 147
 148   int i, ret = 0;
 149
 150   /* Initialize the storage. */
 151   for (i = 0; i < 2; i++)
 152     {
 153       initbuffer (&lb1[i]);
 154       thisline[i] = &lb1[i];
 155       streams[i] = (STREQ (infiles[i], "-") ? stdin : fopen (infiles[i], "r"));
 156       if (!streams[i])
 157         {
 158           error (0, errno, "%s", infiles[i]);
 159           return 1;
 160         }
 161
 162       thisline[i] = readline (thisline[i], streams[i]);
 163     }
 164
 165   while (thisline[0] || thisline[1])
 166     {
 167       int order;
 168
 169       /* Compare the next available lines of the two files.  */
 170
 171       if (!thisline[0])
 172         order = 1;
 173       else if (!thisline[1])
 174         order = -1;
 175       else
 176         {
 177           if (HAVE_SETLOCALE && hard_LC_COLLATE)
 178             order = xmemcoll (thisline[0]->buffer, thisline[0]->length - 1,
 179                               thisline[1]->buffer, thisline[1]->length - 1);
 180           else
 181             {
 182               size_t len = min (thisline[0]->length, thisline[1]->length) - 1;
 183               order = memcmp (thisline[0]->buffer, thisline[1]->buffer, len);
 184               if (order == 0)
 185                 order = (thisline[0]->length < thisline[1]->length
 186                          ? -1
 187                          : thisline[0]->length != thisline[1]->length);
 188             }
 189         }
 190
 191       /* Output the line that is lesser. */
 192       if (order == 0)
 193         writeline (thisline[1], stdout, 3);
 194       else if (order > 0)
 195         writeline (thisline[1], stdout, 2);
 196       else
 197         writeline (thisline[0], stdout, 1);
 198
 199       /* Step the file the line came from.
 200          If the files match, step both files.  */
 201       if (order >= 0)
 202         thisline[1] = readline (thisline[1], streams[1]);
 203       if (order <= 0)
 204         thisline[0] = readline (thisline[0], streams[0]);
 205     }
 206
 207   /* Free all storage and close all input streams. */
 208   for (i = 0; i < 2; i++)
 209     {
 210       free (lb1[i].buffer);
 211       if (ferror (streams[i]) || fclose (streams[i]) == EOF)
 212         {
 213           error (0, errno, "%s", infiles[i]);
 214           ret = 1;
 215         }
 216     }
 217   return ret;
 218 }
 219
 220 int
 221 main (int argc, char **argv)
 222 {
 223   int c;
 224
 225   program_name = argv[0];
 226   setlocale (LC_ALL, "");
 227   bindtextdomain (PACKAGE, LOCALEDIR);
 228   textdomain (PACKAGE);
 229   hard_LC_COLLATE = hard_locale (LC_COLLATE);
 230
 231   atexit (close_stdout);
 232
 233   only_file_1 = 1;
 234   only_file_2 = 1;
 235   both = 1;
 236
 237   while ((c = getopt_long (argc, argv, "123", long_options, NULL)) != -1)
 238     switch (c)
 239       {
 240       case 0:
 241         break;
 242
 243       case '1':
 244         only_file_1 = 0;
 245         break;
 246
 247       case '2':
 248         only_file_2 = 0;
 249         break;
 250
 251       case '3':
 252         both = 0;
 253         break;
 254
 255       case_GETOPT_HELP_CHAR;
 256
 257       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 258
 259       default:
 260         usage (1);
 261       }
 262
 263   if (optind + 2 != argc)
 264     usage (1);
 265
 266   exit (compare_files (argv + optind) == 0
 267         ? EXIT_SUCCESS : EXIT_FAILURE);
 268 }