src/comm.c

   1 /* comm -- compare two sorted files line by line.
   2    Copyright (C) 86, 90, 91, 1995-2004 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 "quote.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 "Richard Stallman", "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 /* True if the LC_COLLATE locale is hard.  */
  45 static bool hard_LC_COLLATE;
  46
  47 /* If true, print lines that are found only in file 1. */
  48 static bool only_file_1;
  49
  50 /* If true, print lines that are found only in file 2. */
  51 static bool only_file_2;
  52
  53 /* If true, print lines that are found in both files. */
  54 static bool 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 != EXIT_SUCCESS)
  69     fprintf (stderr, _("Try `%s --help' for more information.\n"),
  70              program_name);
  71   else
  72     {
  73       printf (_("\
  74 Usage: %s [OPTION]... FILE1 FILE2\n\
  75 "),
  76               program_name);
  77       fputs (_("\
  78 Compare sorted files FILE1 and FILE2 line by line.\n\
  79 "), stdout);
  80       fputs (_("\
  81 \n\
  82 With no options, produce three-column output.  Column one contains\n\
  83 lines unique to FILE1, column two contains lines unique to FILE2,\n\
  84 and column three contains lines common to both files.\n\
  85 "), stdout);
  86       fputs (_("\
  87 \n\
  88   -1              suppress lines unique to FILE1\n\
  89   -2              suppress lines unique to FILE2\n\
  90   -3              suppress lines that appear in both files\n\
  91 "), stdout);
  92       fputs (HELP_OPTION_DESCRIPTION, stdout);
  93       fputs (VERSION_OPTION_DESCRIPTION, stdout);
  94       printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
  95     }
  96   exit (status);
  97 }
  98
  99 /* Output the line in linebuffer LINE to stream STREAM
 100    provided the switches say it should be output.
 101    CLASS is 1 for a line found only in file 1,
 102    2 for a line only in file 2, 3 for a line in both. */
 103
 104 static void
 105 writeline (const struct linebuffer *line, FILE *stream, int class)
 106 {
 107   switch (class)
 108     {
 109     case 1:
 110       if (!only_file_1)
 111         return;
 112       break;
 113
 114     case 2:
 115       if (!only_file_2)
 116         return;
 117       /* Print a TAB if we are printing lines from file 1.  */
 118       if (only_file_1)
 119         putc ('\t', stream);
 120       break;
 121
 122     case 3:
 123       if (!both)
 124         return;
 125       /* Print a TAB if we are printing lines from file 1.  */
 126       if (only_file_1)
 127         putc ('\t', stream);
 128       /* Print a TAB if we are printing lines from file 2.  */
 129       if (only_file_2)
 130         putc ('\t', stream);
 131       break;
 132     }
 133
 134   fwrite (line->buffer, sizeof (char), line->length, stream);
 135 }
 136
 137 /* Compare INFILES[0] and INFILES[1].
 138    If either is "-", use the standard input for that file.
 139    Assume that each input file is sorted;
 140    merge them and output the result.
 141    Return true if successful.  */
 142
 143 static bool
 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   /* errno values for each stream.  */
 157   int saved_errno[2];
 158
 159   int i;
 160   bool ret = true;
 161
 162   /* Initialize the storage. */
 163   for (i = 0; i < 2; i++)
 164     {
 165       initbuffer (&lb1[i]);
 166       thisline[i] = &lb1[i];
 167       streams[i] = (STREQ (infiles[i], "-") ? stdin : fopen (infiles[i], "r"));
 168       if (!streams[i])
 169         {
 170           error (0, errno, "%s", infiles[i]);
 171           return false;
 172         }
 173
 174       thisline[i] = readlinebuffer (thisline[i], streams[i]);
 175       saved_errno[i] = errno;
 176     }
 177
 178   while (thisline[0] || thisline[1])
 179     {
 180       int order;
 181
 182       /* Compare the next available lines of the two files.  */
 183
 184       if (!thisline[0])
 185         order = 1;
 186       else if (!thisline[1])
 187         order = -1;
 188       else
 189         {
 190           if (hard_LC_COLLATE)
 191             order = xmemcoll (thisline[0]->buffer, thisline[0]->length - 1,
 192                               thisline[1]->buffer, thisline[1]->length - 1);
 193           else
 194             {
 195               size_t len = min (thisline[0]->length, thisline[1]->length) - 1;
 196               order = memcmp (thisline[0]->buffer, thisline[1]->buffer, len);
 197               if (order == 0)
 198                 order = (thisline[0]->length < thisline[1]->length
 199                          ? -1
 200                          : thisline[0]->length != thisline[1]->length);
 201             }
 202         }
 203
 204       /* Output the line that is lesser. */
 205       if (order == 0)
 206         writeline (thisline[1], stdout, 3);
 207       else if (order > 0)
 208         writeline (thisline[1], stdout, 2);
 209       else
 210         writeline (thisline[0], stdout, 1);
 211
 212       /* Step the file the line came from.
 213          If the files match, step both files.  */
 214       if (order >= 0)
 215         {
 216           thisline[1] = readlinebuffer (thisline[1], streams[1]);
 217           saved_errno[1] = errno;
 218         }
 219       if (order <= 0)
 220         {
 221           thisline[0] = readlinebuffer (thisline[0], streams[0]);
 222           saved_errno[0] = errno;
 223         }
 224     }
 225
 226   /* Free all storage and close all input streams. */
 227   for (i = 0; i < 2; i++)
 228     {
 229       free (lb1[i].buffer);
 230       if (ferror (streams[i]))
 231         {
 232           error (0, saved_errno[i], "%s", infiles[i]);
 233           ret = false;
 234         }
 235       if (fclose (streams[i]) != 0)
 236         {
 237           error (0, errno, "%s", infiles[i]);
 238           ret = false;
 239         }
 240     }
 241   return ret;
 242 }
 243
 244 int
 245 main (int argc, char **argv)
 246 {
 247   int c;
 248
 249   initialize_main (&argc, &argv);
 250   program_name = argv[0];
 251   setlocale (LC_ALL, "");
 252   bindtextdomain (PACKAGE, LOCALEDIR);
 253   textdomain (PACKAGE);
 254   hard_LC_COLLATE = hard_locale (LC_COLLATE);
 255
 256   atexit (close_stdout);
 257
 258   only_file_1 = true;
 259   only_file_2 = true;
 260   both = true;
 261
 262   while ((c = getopt_long (argc, argv, "123", long_options, NULL)) != -1)
 263     switch (c)
 264       {
 265       case '1':
 266         only_file_1 = false;
 267         break;
 268
 269       case '2':
 270         only_file_2 = false;
 271         break;
 272
 273       case '3':
 274         both = false;
 275         break;
 276
 277       case_GETOPT_HELP_CHAR;
 278
 279       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 280
 281       default:
 282         usage (EXIT_FAILURE);
 283       }
 284
 285   if (argc - optind < 2)
 286     {
 287       if (argc <= optind)
 288         error (0, 0, _("missing operand"));
 289       else
 290         error (0, 0, _("missing operand after %s"), quote (argv[argc - 1]));
 291       usage (EXIT_FAILURE);
 292     }
 293
 294   if (2 < argc - optind)
 295     {
 296       error (0, 0, _("extra operand %s"), quote (argv[optind + 2]));
 297       usage (EXIT_FAILURE);
 298     }
 299
 300   exit (compare_files (argv + optind)
 301         ? EXIT_SUCCESS : EXIT_FAILURE);
 302 }