src/comm.c

   1 /* comm -- compare two sorted files line by line.
   2    Copyright (C) 86, 90, 91, 1995-1999 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 "linebuffer.h"
  27 #include "error.h"
  28 #include "hard-locale.h"
  29 #include "memcoll.h"
  30
  31 /* The official name of this program (e.g., no `g' prefix).  */
  32 #define PROGRAM_NAME "comm"
  33
  34 #define AUTHORS "Richard Stallman and David MacKenzie"
  35
  36 /* Undefine, to avoid warning about redefinition on some systems.  */
  37 #undef min
  38 #define min(x, y) ((x) < (y) ? (x) : (y))
  39
  40 /* The name this program was run with. */
  41 char *program_name;
  42
  43 #ifdef ENABLE_NLS
  44 /* Nonzero if the LC_COLLATE locale is hard.  */
  45 static int hard_LC_COLLATE;
  46 #endif
  47
  48 /* If nonzero, print lines that are found only in file 1. */
  49 static int only_file_1;
  50
  51 /* If nonzero, print lines that are found only in file 2. */
  52 static int only_file_2;
  53
  54 /* If nonzero, print lines that are found in both files. */
  55 static int both;
  56
  57 static struct option const long_options[] =
  58 {
  59   {GETOPT_HELP_OPTION_DECL},
  60   {GETOPT_VERSION_OPTION_DECL},
  61   {0, 0, 0, 0}
  62 };
  63
  64 \f
  65
  66 void
  67 usage (int status)
  68 {
  69   if (status != 0)
  70     fprintf (stderr, _("Try `%s --help' for more information.\n"),
  71              program_name);
  72   else
  73     {
  74       printf (_("\
  75 Usage: %s [OPTION]... LEFT_FILE RIGHT_FILE\n\
  76 "),
  77               program_name);
  78       printf (_("\
  79 Compare sorted files LEFT_FILE and RIGHT_FILE line by line.\n\
  80 \n\
  81   -1              suppress lines unique to left file\n\
  82   -2              suppress lines unique to right file\n\
  83   -3              suppress lines unique to both files\n\
  84       --help      display this help and exit\n\
  85       --version   output version information and exit\n\
  86 "));
  87       puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
  88     }
  89   exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
  90 }
  91
  92 /* Output the line in linebuffer LINE to stream STREAM
  93    provided the switches say it should be output.
  94    CLASS is 1 for a line found only in file 1,
  95    2 for a line only in file 2, 3 for a line in both. */
  96
  97 static void
  98 writeline (const struct linebuffer *line, FILE *stream, int class)
  99 {
 100   switch (class)
 101     {
 102     case 1:
 103       if (!only_file_1)
 104         return;
 105       break;
 106
 107     case 2:
 108       if (!only_file_2)
 109         return;
 110       /* Skip the tab stop for case 1, if we are printing case 1.  */
 111       if (only_file_1)
 112         putc ('\t', stream);
 113       break;
 114
 115     case 3:
 116       if (!both)
 117         return;
 118       /* Skip the tab stop for case 1, if we are printing case 1.  */
 119       if (only_file_1)
 120         putc ('\t', stream);
 121       /* Skip the tab stop for case 2, if we are printing case 2.  */
 122       if (only_file_2)
 123         putc ('\t', stream);
 124       break;
 125     }
 126
 127   fwrite (line->buffer, sizeof (char), line->length, stream);
 128 }
 129
 130 /* Compare INFILES[0] and INFILES[1].
 131    If either is "-", use the standard input for that file.
 132    Assume that each input file is sorted;
 133    merge them and output the result.
 134    Return 0 if successful, 1 if any errors occur. */
 135
 136 static int
 137 compare_files (char **infiles)
 138 {
 139   /* For each file, we have one linebuffer in lb1.  */
 140   struct linebuffer lb1[2];
 141
 142   /* thisline[i] points to the linebuffer holding the next available line
 143      in file i, or is NULL if there are no lines left in that file.  */
 144   struct linebuffer *thisline[2];
 145
 146   /* streams[i] holds the input stream for file i.  */
 147   FILE *streams[2];
 148
 149   int i, ret = 0;
 150
 151   /* Initialize the storage. */
 152   for (i = 0; i < 2; i++)
 153     {
 154       initbuffer (&lb1[i]);
 155       thisline[i] = &lb1[i];
 156       streams[i] = (STREQ (infiles[i], "-") ? stdin : fopen (infiles[i], "r"));
 157       if (!streams[i])
 158         {
 159           error (0, errno, "%s", infiles[i]);
 160           return 1;
 161         }
 162
 163       thisline[i] = readline (thisline[i], streams[i]);
 164     }
 165
 166   while (thisline[0] || thisline[1])
 167     {
 168       int order;
 169
 170       /* Compare the next available lines of the two files.  */
 171
 172       if (!thisline[0])
 173         order = 1;
 174       else if (!thisline[1])
 175         order = -1;
 176       else
 177         {
 178 #ifdef ENABLE_NLS
 179           if (hard_LC_COLLATE)
 180             order = memcoll (thisline[0]->buffer, thisline[0]->length - 1,
 181                              thisline[1]->buffer, thisline[1]->length - 1);
 182           else
 183 #endif
 184             {
 185               size_t len = min (thisline[0]->length, thisline[1]->length) - 1;
 186               order = memcmp (thisline[0]->buffer, thisline[1]->buffer, len);
 187               if (order == 0)
 188                 order = thisline[0]->length - thisline[1]->length;
 189             }
 190         }
 191
 192       /* Output the line that is lesser. */
 193       if (order == 0)
 194         writeline (thisline[1], stdout, 3);
 195       else if (order > 0)
 196         writeline (thisline[1], stdout, 2);
 197       else
 198         writeline (thisline[0], stdout, 1);
 199
 200       /* Step the file the line came from.
 201          If the files match, step both files.  */
 202       if (order >= 0)
 203         thisline[1] = readline (thisline[1], streams[1]);
 204       if (order <= 0)
 205         thisline[0] = readline (thisline[0], streams[0]);
 206     }
 207
 208   /* Free all storage and close all input streams. */
 209   for (i = 0; i < 2; i++)
 210     {
 211       free (lb1[i].buffer);
 212       if (ferror (streams[i]) || fclose (streams[i]) == EOF)
 213         {
 214           error (0, errno, "%s", infiles[i]);
 215           ret = 1;
 216         }
 217     }
 218   if (ferror (stdout) || fclose (stdout) == EOF)
 219     {
 220       error (0, errno, _("write error"));
 221       ret = 1;
 222     }
 223   return ret;
 224 }
 225
 226 int
 227 main (int argc, char **argv)
 228 {
 229   int c;
 230
 231   program_name = argv[0];
 232   setlocale (LC_ALL, "");
 233   bindtextdomain (PACKAGE, LOCALEDIR);
 234   textdomain (PACKAGE);
 235
 236 #ifdef ENABLE_NLS
 237   hard_LC_COLLATE = hard_locale (LC_COLLATE);
 238 #endif
 239
 240   only_file_1 = 1;
 241   only_file_2 = 1;
 242   both = 1;
 243
 244   while ((c = getopt_long (argc, argv, "123", long_options, NULL)) != -1)
 245     switch (c)
 246       {
 247       case 0:
 248         break;
 249
 250       case '1':
 251         only_file_1 = 0;
 252         break;
 253
 254       case '2':
 255         only_file_2 = 0;
 256         break;
 257
 258       case '3':
 259         both = 0;
 260         break;
 261
 262       case_GETOPT_HELP_CHAR;
 263
 264       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 265
 266       default:
 267         usage (1);
 268       }
 269
 270   if (optind + 2 != argc)
 271     usage (1);
 272
 273   exit (compare_files (argv + optind) == 0
 274         ? EXIT_SUCCESS : EXIT_FAILURE);
 275 }