src/comm.c

   1 /* comm -- compare two sorted files line by line.
   2    Copyright (C) 86, 90, 91, 1995-2005 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 3 of the License, or
   7    (at your option) 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, see <http://www.gnu.org/licenses/>.  */
  16
  17 /* Written by Richard Stallman and David MacKenzie. */
  18 \f
  19 #include <config.h>
  20
  21 #include <getopt.h>
  22 #include <sys/types.h>
  23 #include "system.h"
  24 #include "linebuffer.h"
  25 #include "error.h"
  26 #include "hard-locale.h"
  27 #include "quote.h"
  28 #include "stdio--.h"
  29 #include "xmemcoll.h"
  30
  31 /* The official name of this program (e.g., no `g' prefix).  */
  32 #define PROGRAM_NAME "comm"
  33
  34 #define AUTHORS \
  35   proper_name ("Richard Stallman"), \
  36   proper_name ("David MacKenzie")
  37
  38 /* Undefine, to avoid warning about redefinition on some systems.  */
  39 #undef min
  40 #define min(x, y) ((x) < (y) ? (x) : (y))
  41
  42 /* The name this program was run with. */
  43 char *program_name;
  44
  45 /* True if the LC_COLLATE locale is hard.  */
  46 static bool hard_LC_COLLATE;
  47
  48 /* If true, print lines that are found only in file 1. */
  49 static bool only_file_1;
  50
  51 /* If true, print lines that are found only in file 2. */
  52 static bool only_file_2;
  53
  54 /* If true, print lines that are found in both files. */
  55 static bool both;
  56
  57 static struct option const long_options[] =
  58 {
  59   {GETOPT_HELP_OPTION_DECL},
  60   {GETOPT_VERSION_OPTION_DECL},
  61   {NULL, 0, NULL, 0}
  62 };
  63
  64 \f
  65
  66 void
  67 usage (int status)
  68 {
  69   if (status != EXIT_SUCCESS)
  70     fprintf (stderr, _("Try `%s --help' for more information.\n"),
  71              program_name);
  72   else
  73     {
  74       printf (_("\
  75 Usage: %s [OPTION]... FILE1 FILE2\n\
  76 "),
  77               program_name);
  78       fputs (_("\
  79 Compare sorted files FILE1 and FILE2 line by line.\n\
  80 "), stdout);
  81       fputs (_("\
  82 \n\
  83 With no options, produce three-column output.  Column one contains\n\
  84 lines unique to FILE1, column two contains lines unique to FILE2,\n\
  85 and column three contains lines common to both files.\n\
  86 "), stdout);
  87       fputs (_("\
  88 \n\
  89   -1              suppress lines unique to FILE1\n\
  90   -2              suppress lines unique to FILE2\n\
  91   -3              suppress lines that appear in both files\n\
  92 "), stdout);
  93       fputs (HELP_OPTION_DESCRIPTION, stdout);
  94       fputs (VERSION_OPTION_DESCRIPTION, stdout);
  95       emit_bug_reporting_address ();
  96     }
  97   exit (status);
  98 }
  99
 100 /* Output the line in linebuffer LINE to stream STREAM
 101    provided the switches say it should be output.
 102    CLASS is 1 for a line found only in file 1,
 103    2 for a line only in file 2, 3 for a line in both. */
 104
 105 static void
 106 writeline (const struct linebuffer *line, FILE *stream, int class)
 107 {
 108   switch (class)
 109     {
 110     case 1:
 111       if (!only_file_1)
 112         return;
 113       break;
 114
 115     case 2:
 116       if (!only_file_2)
 117         return;
 118       /* Print a TAB if we are printing lines from file 1.  */
 119       if (only_file_1)
 120         putc ('\t', stream);
 121       break;
 122
 123     case 3:
 124       if (!both)
 125         return;
 126       /* Print a TAB if we are printing lines from file 1.  */
 127       if (only_file_1)
 128         putc ('\t', stream);
 129       /* Print a TAB if we are printing lines from file 2.  */
 130       if (only_file_2)
 131         putc ('\t', stream);
 132       break;
 133     }
 134
 135   fwrite (line->buffer, sizeof (char), line->length, stream);
 136 }
 137
 138 /* Compare INFILES[0] and INFILES[1].
 139    If either is "-", use the standard input for that file.
 140    Assume that each input file is sorted;
 141    merge them and output the result.  */
 142
 143 static void
 144 compare_files (char **infiles)
 145 {
 146   /* For each file, we have one linebuffer in lb1.  */
 147   struct linebuffer lb1[2];
 148
 149   /* thisline[i] points to the linebuffer holding the next available line
 150      in file i, or is NULL if there are no lines left in that file.  */
 151   struct linebuffer *thisline[2];
 152
 153   /* streams[i] holds the input stream for file i.  */
 154   FILE *streams[2];
 155
 156   int i;
 157
 158   /* Initialize the storage. */
 159   for (i = 0; i < 2; i++)
 160     {
 161       initbuffer (&lb1[i]);
 162       thisline[i] = &lb1[i];
 163       streams[i] = (STREQ (infiles[i], "-") ? stdin : fopen (infiles[i], "r"));
 164       if (!streams[i])
 165         error (EXIT_FAILURE, errno, "%s", infiles[i]);
 166
 167       thisline[i] = readlinebuffer (thisline[i], streams[i]);
 168       if (ferror (streams[i]))
 169         error (EXIT_FAILURE, errno, "%s", infiles[i]);
 170     }
 171
 172   while (thisline[0] || thisline[1])
 173     {
 174       int order;
 175
 176       /* Compare the next available lines of the two files.  */
 177
 178       if (!thisline[0])
 179         order = 1;
 180       else if (!thisline[1])
 181         order = -1;
 182       else
 183         {
 184           if (hard_LC_COLLATE)
 185             order = xmemcoll (thisline[0]->buffer, thisline[0]->length - 1,
 186                               thisline[1]->buffer, thisline[1]->length - 1);
 187           else
 188             {
 189               size_t len = min (thisline[0]->length, thisline[1]->length) - 1;
 190               order = memcmp (thisline[0]->buffer, thisline[1]->buffer, len);
 191               if (order == 0)
 192                 order = (thisline[0]->length < thisline[1]->length
 193                          ? -1
 194                          : thisline[0]->length != thisline[1]->length);
 195             }
 196         }
 197
 198       /* Output the line that is lesser. */
 199       if (order == 0)
 200         writeline (thisline[1], stdout, 3);
 201       else if (order > 0)
 202         writeline (thisline[1], stdout, 2);
 203       else
 204         writeline (thisline[0], stdout, 1);
 205
 206       /* Step the file the line came from.
 207          If the files match, step both files.  */
 208       if (order >= 0)
 209         {
 210           thisline[1] = readlinebuffer (thisline[1], streams[1]);
 211           if (ferror (streams[1]))
 212             error (EXIT_FAILURE, errno, "%s", infiles[1]);
 213         }
 214       if (order <= 0)
 215         {
 216           thisline[0] = readlinebuffer (thisline[0], streams[0]);
 217           if (ferror (streams[0]))
 218             error (EXIT_FAILURE, errno, "%s", infiles[0]);
 219         }
 220     }
 221
 222   for (i = 0; i < 2; i++)
 223     if (fclose (streams[i]) != 0)
 224       error (EXIT_FAILURE, errno, "%s", infiles[i]);
 225 }
 226
 227 int
 228 main (int argc, char **argv)
 229 {
 230   int c;
 231
 232   initialize_main (&argc, &argv);
 233   program_name = argv[0];
 234   setlocale (LC_ALL, "");
 235   bindtextdomain (PACKAGE, LOCALEDIR);
 236   textdomain (PACKAGE);
 237   hard_LC_COLLATE = hard_locale (LC_COLLATE);
 238
 239   atexit (close_stdout);
 240
 241   only_file_1 = true;
 242   only_file_2 = true;
 243   both = true;
 244
 245   while ((c = getopt_long (argc, argv, "123", long_options, NULL)) != -1)
 246     switch (c)
 247       {
 248       case '1':
 249         only_file_1 = false;
 250         break;
 251
 252       case '2':
 253         only_file_2 = false;
 254         break;
 255
 256       case '3':
 257         both = false;
 258         break;
 259
 260       case_GETOPT_HELP_CHAR;
 261
 262       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 263
 264       default:
 265         usage (EXIT_FAILURE);
 266       }
 267
 268   if (argc - optind < 2)
 269     {
 270       if (argc <= optind)
 271         error (0, 0, _("missing operand"));
 272       else
 273         error (0, 0, _("missing operand after %s"), quote (argv[argc - 1]));
 274       usage (EXIT_FAILURE);
 275     }
 276
 277   if (2 < argc - optind)
 278     {
 279       error (0, 0, _("extra operand %s"), quote (argv[optind + 2]));
 280       usage (EXIT_FAILURE);
 281     }
 282
 283   compare_files (argv + optind);
 284
 285   exit (EXIT_SUCCESS);
 286 }