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 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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.  */
  17
  18 /* Written by Richard Stallman and David MacKenzie. */
  19 \f
  20 #include <config.h>
  21
  22 #include <getopt.h>
  23 #include <sys/types.h>
  24 #include "system.h"
  25 #include "linebuffer.h"
  26 #include "error.h"
  27 #include "hard-locale.h"
  28 #include "quote.h"
  29 #include "stdio--.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   {NULL, 0, NULL, 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
 142 static void
 143 compare_files (char **infiles)
 144 {
 145   /* For each file, we have one linebuffer in lb1.  */
 146   struct linebuffer lb1[2];
 147
 148   /* thisline[i] points to the linebuffer holding the next available line
 149      in file i, or is NULL if there are no lines left in that file.  */
 150   struct linebuffer *thisline[2];
 151
 152   /* streams[i] holds the input stream for file i.  */
 153   FILE *streams[2];
 154
 155   int i;
 156
 157   /* Initialize the storage. */
 158   for (i = 0; i < 2; i++)
 159     {
 160       initbuffer (&lb1[i]);
 161       thisline[i] = &lb1[i];
 162       streams[i] = (STREQ (infiles[i], "-") ? stdin : fopen (infiles[i], "r"));
 163       if (!streams[i])
 164         error (EXIT_FAILURE, errno, "%s", infiles[i]);
 165
 166       thisline[i] = readlinebuffer (thisline[i], streams[i]);
 167       if (ferror (streams[i]))
 168         error (EXIT_FAILURE, errno, "%s", infiles[i]);
 169     }
 170
 171   while (thisline[0] || thisline[1])
 172     {
 173       int order;
 174
 175       /* Compare the next available lines of the two files.  */
 176
 177       if (!thisline[0])
 178         order = 1;
 179       else if (!thisline[1])
 180         order = -1;
 181       else
 182         {
 183           if (hard_LC_COLLATE)
 184             order = xmemcoll (thisline[0]->buffer, thisline[0]->length - 1,
 185                               thisline[1]->buffer, thisline[1]->length - 1);
 186           else
 187             {
 188               size_t len = min (thisline[0]->length, thisline[1]->length) - 1;
 189               order = memcmp (thisline[0]->buffer, thisline[1]->buffer, len);
 190               if (order == 0)
 191                 order = (thisline[0]->length < thisline[1]->length
 192                          ? -1
 193                          : thisline[0]->length != thisline[1]->length);
 194             }
 195         }
 196
 197       /* Output the line that is lesser. */
 198       if (order == 0)
 199         writeline (thisline[1], stdout, 3);
 200       else if (order > 0)
 201         writeline (thisline[1], stdout, 2);
 202       else
 203         writeline (thisline[0], stdout, 1);
 204
 205       /* Step the file the line came from.
 206          If the files match, step both files.  */
 207       if (order >= 0)
 208         {
 209           thisline[1] = readlinebuffer (thisline[1], streams[1]);
 210           if (ferror (streams[1]))
 211             error (EXIT_FAILURE, errno, "%s", infiles[1]);
 212         }
 213       if (order <= 0)
 214         {
 215           thisline[0] = readlinebuffer (thisline[0], streams[0]);
 216           if (ferror (streams[0]))
 217             error (EXIT_FAILURE, errno, "%s", infiles[0]);
 218         }
 219     }
 220
 221   for (i = 0; i < 2; i++)
 222     if (fclose (streams[i]) != 0)
 223       error (EXIT_FAILURE, errno, "%s", infiles[i]);
 224 }
 225
 226 int
 227 main (int argc, char **argv)
 228 {
 229   int c;
 230
 231   initialize_main (&argc, &argv);
 232   program_name = argv[0];
 233   setlocale (LC_ALL, "");
 234   bindtextdomain (PACKAGE, LOCALEDIR);
 235   textdomain (PACKAGE);
 236   hard_LC_COLLATE = hard_locale (LC_COLLATE);
 237
 238   atexit (close_stdout);
 239
 240   only_file_1 = true;
 241   only_file_2 = true;
 242   both = true;
 243
 244   while ((c = getopt_long (argc, argv, "123", long_options, NULL)) != -1)
 245     switch (c)
 246       {
 247       case '1':
 248         only_file_1 = false;
 249         break;
 250
 251       case '2':
 252         only_file_2 = false;
 253         break;
 254
 255       case '3':
 256         both = false;
 257         break;
 258
 259       case_GETOPT_HELP_CHAR;
 260
 261       case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
 262
 263       default:
 264         usage (EXIT_FAILURE);
 265       }
 266
 267   if (argc - optind < 2)
 268     {
 269       if (argc <= optind)
 270         error (0, 0, _("missing operand"));
 271       else
 272         error (0, 0, _("missing operand after %s"), quote (argv[argc - 1]));
 273       usage (EXIT_FAILURE);
 274     }
 275
 276   if (2 < argc - optind)
 277     {
 278       error (0, 0, _("extra operand %s"), quote (argv[optind + 2]));
 279       usage (EXIT_FAILURE);
 280     }
 281
 282   compare_files (argv + optind);
 283
 284   exit (EXIT_SUCCESS);
 285 }