gnu/dist/gettext/gettext-tools/src/msguniq.c

   1 /* Remove, select or merge duplicate translations.
   2    Copyright (C) 2001-2005 Free Software Foundation, Inc.
   3    Written by Bruno Haible <haible@clisp.cons.org>, 2001.
   4
   5    This program is free software; you can redistribute it and/or modify
   6    it under the terms of the GNU General Public License as published by
   7    the Free Software Foundation; either version 2, or (at your option)
   8    any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program; if not, write to the Free Software Foundation,
  17    Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  18
  19
  20 #ifdef HAVE_CONFIG_H
  21 # include "config.h"
  22 #endif
  23
  24 #include <getopt.h>
  25 #include <limits.h>
  26 #include <stdio.h>
  27 #include <stdlib.h>
  28 #include <locale.h>
  29
  30 #include "closeout.h"
  31 #include "dir-list.h"
  32 #include "str-list.h"
  33 #include "error.h"
  34 #include "error-progname.h"
  35 #include "progname.h"
  36 #include "relocatable.h"
  37 #include "basename.h"
  38 #include "message.h"
  39 #include "read-po.h"
  40 #include "write-po.h"
  41 #include "msgl-cat.h"
  42 #include "exit.h"
  43 #include "gettext.h"
  44
  45 #define _(str) gettext (str)
  46
  47
  48 /* Force output of PO file even if empty.  */
  49 static int force_po;
  50
  51 /* Target encoding.  */
  52 static const char *to_code;
  53
  54 /* Long options.  */
  55 static const struct option long_options[] =
  56 {
  57   { "add-location", no_argument, &line_comment, 1 },
  58   { "directory", required_argument, NULL, 'D' },
  59   { "escape", no_argument, NULL, 'E' },
  60   { "force-po", no_argument, &force_po, 1 },
  61   { "help", no_argument, NULL, 'h' },
  62   { "indent", no_argument, NULL, 'i' },
  63   { "no-escape", no_argument, NULL, 'e' },
  64   { "no-location", no_argument, &line_comment, 0 },
  65   { "no-wrap", no_argument, NULL, CHAR_MAX + 2 },
  66   { "output-file", required_argument, NULL, 'o' },
  67   { "properties-input", no_argument, NULL, 'P' },
  68   { "properties-output", no_argument, NULL, 'p' },
  69   { "repeated", no_argument, NULL, 'd' },
  70   { "sort-by-file", no_argument, NULL, 'F' },
  71   { "sort-output", no_argument, NULL, 's' },
  72   { "strict", no_argument, NULL, 'S' },
  73   { "stringtable-input", no_argument, NULL, CHAR_MAX + 3 },
  74   { "stringtable-output", no_argument, NULL, CHAR_MAX + 4 },
  75   { "to-code", required_argument, NULL, 't' },
  76   { "unique", no_argument, NULL, 'u' },
  77   { "use-first", no_argument, NULL, CHAR_MAX + 1 },
  78   { "version", no_argument, NULL, 'V' },
  79   { "width", required_argument, NULL, 'w', },
  80   { NULL, 0, NULL, 0 }
  81 };
  82
  83
  84 /* Forward declaration of local functions.  */
  85 static void usage (int status)
  86 #if defined __GNUC__ && ((__GNUC__ == 2 && __GNUC_MINOR__ >= 5) || __GNUC__ > 2)
  87         __attribute__ ((noreturn))
  88 #endif
  89 ;
  90
  91
  92 int
  93 main (int argc, char **argv)
  94 {
  95   int optchar;
  96   bool do_help;
  97   bool do_version;
  98   char *output_file;
  99   const char *input_file;
 100   string_list_ty *file_list;
 101   msgdomain_list_ty *result;
 102   bool sort_by_msgid = false;
 103   bool sort_by_filepos = false;
 104
 105   /* Set program name for messages.  */
 106   set_program_name (argv[0]);
 107   error_print_progname = maybe_print_progname;
 108
 109 #ifdef HAVE_SETLOCALE
 110   /* Set locale via LC_ALL.  */
 111   setlocale (LC_ALL, "");
 112 #endif
 113
 114   /* Set the text message domain.  */
 115   bindtextdomain (PACKAGE, relocate (LOCALEDIR));
 116   textdomain (PACKAGE);
 117
 118   /* Ensure that write errors on stdout are detected.  */
 119   atexit (close_stdout);
 120
 121   /* Set default values for variables.  */
 122   do_help = false;
 123   do_version = false;
 124   output_file = NULL;
 125   input_file = NULL;
 126   more_than = 0;
 127   less_than = INT_MAX;
 128   use_first = false;
 129
 130   while ((optchar = getopt_long (argc, argv, "dD:eEFhino:pPst:uVw:",
 131                                  long_options, NULL)) != EOF)
 132     switch (optchar)
 133       {
 134       case '\0':                /* Long option.  */
 135         break;
 136
 137       case 'd':
 138         more_than = 1;
 139         less_than = INT_MAX;
 140         break;
 141
 142       case 'D':
 143         dir_list_append (optarg);
 144         break;
 145
 146       case 'e':
 147         message_print_style_escape (false);
 148         break;
 149
 150       case 'E':
 151         message_print_style_escape (true);
 152         break;
 153
 154       case 'F':
 155         sort_by_filepos = true;
 156         break;
 157
 158       case 'h':
 159         do_help = true;
 160         break;
 161
 162       case 'i':
 163         message_print_style_indent ();
 164         break;
 165
 166       case 'n':
 167         line_comment = 1;
 168         break;
 169
 170       case 'o':
 171         output_file = optarg;
 172         break;
 173
 174       case 'p':
 175         message_print_syntax_properties ();
 176         break;
 177
 178       case 'P':
 179         input_syntax = syntax_properties;
 180         break;
 181
 182       case 's':
 183         sort_by_msgid = true;
 184         break;
 185
 186       case 'S':
 187         message_print_style_uniforum ();
 188         break;
 189
 190       case 't':
 191         to_code = optarg;
 192         break;
 193
 194       case 'u':
 195         more_than = 0;
 196         less_than = 2;
 197         break;
 198
 199       case 'V':
 200         do_version = true;
 201         break;
 202
 203       case 'w':
 204         {
 205           int value;
 206           char *endp;
 207           value = strtol (optarg, &endp, 10);
 208           if (endp != optarg)
 209             message_page_width_set (value);
 210         }
 211         break;
 212
 213       case CHAR_MAX + 1:
 214         use_first = true;
 215         break;
 216
 217       case CHAR_MAX + 2: /* --no-wrap */
 218         message_page_width_ignore ();
 219         break;
 220
 221       case CHAR_MAX + 3: /* --stringtable-input */
 222         input_syntax = syntax_stringtable;
 223         break;
 224
 225       case CHAR_MAX + 4: /* --stringtable-output */
 226         message_print_syntax_stringtable ();
 227         break;
 228
 229       default:
 230         usage (EXIT_FAILURE);
 231         /* NOTREACHED */
 232       }
 233
 234   /* Version information requested.  */
 235   if (do_version)
 236     {
 237       printf ("%s (GNU %s) %s\n", basename (program_name), PACKAGE, VERSION);
 238       /* xgettext: no-wrap */
 239       printf (_("Copyright (C) %s Free Software Foundation, Inc.\n\
 240 This is free software; see the source for copying conditions.  There is NO\n\
 241 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\
 242 "),
 243               "2001-2005");
 244       printf (_("Written by %s.\n"), "Bruno Haible");
 245       exit (EXIT_SUCCESS);
 246     }
 247
 248   /* Help is requested.  */
 249   if (do_help)
 250     usage (EXIT_SUCCESS);
 251
 252   /* Test whether we have an .po file name as argument.  */
 253   if (optind == argc)
 254     input_file = "-";
 255   else if (optind + 1 == argc)
 256     input_file = argv[optind];
 257   else
 258     {
 259       error (EXIT_SUCCESS, 0, _("at most one input file allowed"));
 260       usage (EXIT_FAILURE);
 261     }
 262
 263   /* Verify selected options.  */
 264   if (!line_comment && sort_by_filepos)
 265     error (EXIT_FAILURE, 0, _("%s and %s are mutually exclusive"),
 266            "--no-location", "--sort-by-file");
 267
 268   if (sort_by_msgid && sort_by_filepos)
 269     error (EXIT_FAILURE, 0, _("%s and %s are mutually exclusive"),
 270            "--sort-output", "--sort-by-file");
 271
 272   /* Determine list of files we have to process: a single file.  */
 273   file_list = string_list_alloc ();
 274   string_list_append (file_list, input_file);
 275
 276   /* Read input files, then filter, convert and merge messages.  */
 277   allow_duplicates = true;
 278   result = catenate_msgdomain_list (file_list, to_code);
 279
 280   string_list_free (file_list);
 281
 282   /* Sorting the list of messages.  */
 283   if (sort_by_filepos)
 284     msgdomain_list_sort_by_filepos (result);
 285   else if (sort_by_msgid)
 286     msgdomain_list_sort_by_msgid (result);
 287
 288   /* Write the PO file.  */
 289   msgdomain_list_print (result, output_file, force_po, false);
 290
 291   exit (EXIT_SUCCESS);
 292 }
 293
 294
 295 /* Display usage information and exit.  */
 296 static void
 297 usage (int status)
 298 {
 299   if (status != EXIT_SUCCESS)
 300     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 301              program_name);
 302   else
 303     {
 304       printf (_("\
 305 Usage: %s [OPTION] [INPUTFILE]\n\
 306 "), program_name);
 307       printf ("\n");
 308       /* xgettext: no-wrap */
 309       printf (_("\
 310 Unifies duplicate translations in a translation catalog.\n\
 311 Finds duplicate translations of the same message ID.  Such duplicates are\n\
 312 invalid input for other programs like msgfmt, msgmerge or msgcat.  By\n\
 313 default, duplicates are merged together.  When using the --repeated option,\n\
 314 only duplicates are output, and all other messages are discarded.  Comments\n\
 315 and extracted comments will be cumulated, except that if --use-first is\n\
 316 specified, they will be taken from the first translation.  File positions\n\
 317 will be cumulated.  When using the --unique option, duplicates are discarded.\n\
 318 "));
 319       printf ("\n");
 320       printf (_("\
 321 Mandatory arguments to long options are mandatory for short options too.\n"));
 322       printf ("\n");
 323       printf (_("\
 324 Input file location:\n"));
 325       printf (_("\
 326   INPUTFILE                   input PO file\n"));
 327       printf (_("\
 328   -D, --directory=DIRECTORY   add DIRECTORY to list for input files search\n"));
 329       printf (_("\
 330 If no input file is given or if it is -, standard input is read.\n"));
 331       printf ("\n");
 332       printf (_("\
 333 Output file location:\n"));
 334       printf (_("\
 335   -o, --output-file=FILE      write output to specified file\n"));
 336       printf (_("\
 337 The results are written to standard output if no output file is specified\n\
 338 or if it is -.\n"));
 339       printf ("\n");
 340       printf (_("\
 341 Message selection:\n"));
 342       printf (_("\
 343   -d, --repeated              print only duplicates\n"));
 344       printf (_("\
 345   -u, --unique                print only unique messages, discard duplicates\n"));
 346       printf ("\n");
 347       printf (_("\
 348 Input file syntax:\n"));
 349       printf (_("\
 350   -P, --properties-input      input file is in Java .properties syntax\n"));
 351       printf (_("\
 352       --stringtable-input     input file is in NeXTstep/GNUstep .strings syntax\n"));
 353       printf ("\n");
 354       printf (_("\
 355 Output details:\n"));
 356       printf (_("\
 357   -t, --to-code=NAME          encoding for output\n"));
 358       printf (_("\
 359       --use-first             use first available translation for each\n\
 360                               message, don't merge several translations\n"));
 361       printf (_("\
 362   -e, --no-escape             do not use C escapes in output (default)\n"));
 363       printf (_("\
 364   -E, --escape                use C escapes in output, no extended chars\n"));
 365       printf (_("\
 366       --force-po              write PO file even if empty\n"));
 367       printf (_("\
 368   -i, --indent                write the .po file using indented style\n"));
 369       printf (_("\
 370       --no-location           do not write '#: filename:line' lines\n"));
 371       printf (_("\
 372   -n, --add-location          generate '#: filename:line' lines (default)\n"));
 373       printf (_("\
 374       --strict                write out strict Uniforum conforming .po file\n"));
 375       printf (_("\
 376   -p, --properties-output     write out a Java .properties file\n"));
 377       printf (_("\
 378       --stringtable-output    write out a NeXTstep/GNUstep .strings file\n"));
 379       printf (_("\
 380   -w, --width=NUMBER          set output page width\n"));
 381       printf (_("\
 382       --no-wrap               do not break long message lines, longer than\n\
 383                               the output page width, into several lines\n"));
 384       printf (_("\
 385   -s, --sort-output           generate sorted output\n"));
 386       printf (_("\
 387   -F, --sort-by-file          sort output by file location\n"));
 388       printf ("\n");
 389       printf (_("\
 390 Informative output:\n"));
 391       printf (_("\
 392   -h, --help                  display this help and exit\n"));
 393       printf (_("\
 394   -V, --version               output version information and exit\n"));
 395       printf ("\n");
 396       fputs (_("Report bugs to <bug-gnu-gettext@gnu.org>.\n"),
 397              stdout);
 398     }
 399
 400   exit (status);
 401 }
 402