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