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

   1 /* GNU gettext - internationalization aids
   2    Copyright (C) 1995-1998, 2000-2005 Free Software Foundation, Inc.
   3    This file was written by Peter Miller <millerp@canb.auug.org.au>
   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 #ifdef HAVE_CONFIG_H
  20 # include <config.h>
  21 #endif
  22
  23 #include <getopt.h>
  24 #include <limits.h>
  25 #include <stdbool.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 "error.h"
  33 #include "error-progname.h"
  34 #include "progname.h"
  35 #include "relocatable.h"
  36 #include "basename.h"
  37 #include "message.h"
  38 #include "exit.h"
  39 #include "read-po.h"
  40 #include "msgl-iconv.h"
  41 #include "strstr.h"
  42 #include "c-strcase.h"
  43 #include "gettext.h"
  44
  45 #define _(str) gettext (str)
  46
  47
  48 /* Apply the .pot file to each of the domains in the PO file.  */
  49 static bool multi_domain_mode = false;
  50
  51 /* Long options.  */
  52 static const struct option long_options[] =
  53 {
  54   { "directory", required_argument, NULL, 'D' },
  55   { "help", no_argument, NULL, 'h' },
  56   { "multi-domain", no_argument, NULL, 'm' },
  57   { "properties-input", no_argument, NULL, 'P' },
  58   { "stringtable-input", no_argument, NULL, CHAR_MAX + 1 },
  59   { "version", no_argument, NULL, 'V' },
  60   { NULL, 0, NULL, 0 }
  61 };
  62
  63
  64 /* Forward declaration of local functions.  */
  65 static void usage (int status)
  66 #if defined __GNUC__ && ((__GNUC__ == 2 && __GNUC_MINOR__ >= 5) || __GNUC__ > 2)
  67         __attribute__ ((noreturn))
  68 #endif
  69 ;
  70 static void compare (const char *fn1, const char *fn2);
  71
  72
  73 int
  74 main (int argc, char *argv[])
  75 {
  76   int optchar;
  77   bool do_help;
  78   bool do_version;
  79
  80   /* Set program name for messages.  */
  81   set_program_name (argv[0]);
  82   error_print_progname = maybe_print_progname;
  83   gram_max_allowed_errors = UINT_MAX;
  84
  85 #ifdef HAVE_SETLOCALE
  86   /* Set locale via LC_ALL.  */
  87   setlocale (LC_ALL, "");
  88 #endif
  89
  90   /* Set the text message domain.  */
  91   bindtextdomain (PACKAGE, relocate (LOCALEDIR));
  92   textdomain (PACKAGE);
  93
  94   /* Ensure that write errors on stdout are detected.  */
  95   atexit (close_stdout);
  96
  97   do_help = false;
  98   do_version = false;
  99   while ((optchar = getopt_long (argc, argv, "D:hmPV", long_options, NULL))
 100          != EOF)
 101     switch (optchar)
 102       {
 103       case '\0':                /* long option */
 104         break;
 105
 106       case 'D':
 107         dir_list_append (optarg);
 108         break;
 109
 110       case 'h':
 111         do_help = true;
 112         break;
 113
 114       case 'm':
 115         multi_domain_mode = true;
 116         break;
 117
 118       case 'P':
 119         input_syntax = syntax_properties;
 120         break;
 121
 122       case 'V':
 123         do_version = true;
 124         break;
 125
 126       case CHAR_MAX + 1: /* --stringtable-input */
 127         input_syntax = syntax_stringtable;
 128         break;
 129
 130       default:
 131         usage (EXIT_FAILURE);
 132         break;
 133       }
 134
 135   /* Version information is requested.  */
 136   if (do_version)
 137     {
 138       printf ("%s (GNU %s) %s\n", basename (program_name), PACKAGE, VERSION);
 139       /* xgettext: no-wrap */
 140       printf (_("Copyright (C) %s Free Software Foundation, Inc.\n\
 141 This is free software; see the source for copying conditions.  There is NO\n\
 142 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\
 143 "),
 144               "1995-1998, 2000-2005");
 145       printf (_("Written by %s.\n"), "Peter Miller");
 146       exit (EXIT_SUCCESS);
 147     }
 148
 149   /* Help is requested.  */
 150   if (do_help)
 151     usage (EXIT_SUCCESS);
 152
 153   /* Test whether we have an .po file name as argument.  */
 154   if (optind >= argc)
 155     {
 156       error (EXIT_SUCCESS, 0, _("no input files given"));
 157       usage (EXIT_FAILURE);
 158     }
 159   if (optind + 2 != argc)
 160     {
 161       error (EXIT_SUCCESS, 0, _("exactly 2 input files required"));
 162       usage (EXIT_FAILURE);
 163     }
 164
 165   /* compare the two files */
 166   compare (argv[optind], argv[optind + 1]);
 167   exit (EXIT_SUCCESS);
 168 }
 169
 170
 171 /* Display usage information and exit.  */
 172 static void
 173 usage (int status)
 174 {
 175   if (status != EXIT_SUCCESS)
 176     fprintf (stderr, _("Try `%s --help' for more information.\n"),
 177              program_name);
 178   else
 179     {
 180       printf (_("\
 181 Usage: %s [OPTION] def.po ref.pot\n\
 182 "), program_name);
 183       printf ("\n");
 184       /* xgettext: no-wrap */
 185       printf (_("\
 186 Compare two Uniforum style .po files to check that both contain the same\n\
 187 set of msgid strings.  The def.po file is an existing PO file with the\n\
 188 translations.  The ref.pot file is the last created PO file, or a PO Template\n\
 189 file (generally created by xgettext).  This is useful for checking that\n\
 190 you have translated each and every message in your program.  Where an exact\n\
 191 match cannot be found, fuzzy matching is used to produce better diagnostics.\n\
 192 "));
 193       printf ("\n");
 194       printf (_("\
 195 Mandatory arguments to long options are mandatory for short options too.\n"));
 196       printf ("\n");
 197       printf (_("\
 198 Input file location:\n"));
 199       printf (_("\
 200   def.po                      translations\n"));
 201       printf (_("\
 202   ref.pot                     references to the sources\n"));
 203       printf (_("\
 204   -D, --directory=DIRECTORY   add DIRECTORY to list for input files search\n"));
 205       printf ("\n");
 206       printf (_("\
 207 Operation modifiers:\n"));
 208       printf (_("\
 209   -m, --multi-domain          apply ref.pot to each of the domains in def.po\n"));
 210       printf ("\n");
 211       printf (_("\
 212 Input file syntax:\n"));
 213       printf (_("\
 214   -P, --properties-input      input files are in Java .properties syntax\n"));
 215       printf (_("\
 216       --stringtable-input     input files are in NeXTstep/GNUstep .strings\n\
 217                               syntax\n"));
 218       printf ("\n");
 219       printf (_("\
 220 Informative output:\n"));
 221       printf (_("\
 222   -h, --help                  display this help and exit\n"));
 223       printf (_("\
 224   -V, --version               output version information and exit\n"));
 225       printf ("\n");
 226       fputs (_("Report bugs to <bug-gnu-gettext@gnu.org>.\n"), stdout);
 227     }
 228
 229   exit (status);
 230 }
 231
 232
 233 /* Return true if a message should be kept.  */
 234 static bool
 235 is_message_selected (const message_ty *mp)
 236 {
 237   /* Always keep the header entry.  */
 238   if (mp->msgid[0] == '\0')
 239     return true;
 240
 241   return !mp->obsolete;
 242 }
 243
 244
 245 /* Remove obsolete messages from a message list.  Return the modified list.  */
 246 static msgdomain_list_ty *
 247 remove_obsoletes (msgdomain_list_ty *mdlp)
 248 {
 249   size_t k;
 250
 251   for (k = 0; k < mdlp->nitems; k++)
 252     message_list_remove_if_not (mdlp->item[k]->messages, is_message_selected);
 253
 254   return mdlp;
 255 }
 256
 257
 258 static void
 259 match_domain (const char *fn1, const char *fn2,
 260               message_list_ty *defmlp, message_list_ty *refmlp,
 261               int *nerrors)
 262 {
 263   size_t j;
 264
 265   for (j = 0; j < refmlp->nitems; j++)
 266     {
 267       message_ty *refmsg;
 268       message_ty *defmsg;
 269
 270       refmsg = refmlp->item[j];
 271
 272       /* See if it is in the other file.  */
 273       defmsg = message_list_search (defmlp, refmsg->msgid);
 274       if (defmsg)
 275         defmsg->used = 1;
 276       else
 277         {
 278           /* If the message was not defined at all, try to find a very
 279              similar message, it could be a typo, or the suggestion may
 280              help.  */
 281           (*nerrors)++;
 282           defmsg = message_list_search_fuzzy (defmlp, refmsg->msgid);
 283           if (defmsg)
 284             {
 285               po_gram_error_at_line (&refmsg->pos, _("\
 286 this message is used but not defined..."));
 287               po_gram_error_at_line (&defmsg->pos, _("\
 288 ...but this definition is similar"));
 289               defmsg->used = 1;
 290             }
 291           else
 292             po_gram_error_at_line (&refmsg->pos, _("\
 293 this message is used but not defined in %s"), fn1);
 294         }
 295     }
 296 }
 297
 298
 299 static void
 300 compare (const char *fn1, const char *fn2)
 301 {
 302   msgdomain_list_ty *def;
 303   msgdomain_list_ty *ref;
 304   int nerrors;
 305   size_t j, k;
 306   message_list_ty *empty_list;
 307
 308   /* This is the master file, created by a human.  */
 309   def = remove_obsoletes (read_po_file (fn1));
 310
 311   /* This is the generated file, created by groping the sources with
 312      the xgettext program.  */
 313   ref = remove_obsoletes (read_po_file (fn2));
 314
 315   /* The references file can be either in ASCII or in UTF-8.  If it is
 316      in UTF-8, we have to convert the definitions to UTF-8 as well.  */
 317   {
 318     bool was_utf8 = false;
 319     for (k = 0; k < ref->nitems; k++)
 320       {
 321         message_list_ty *mlp = ref->item[k]->messages;
 322
 323         for (j = 0; j < mlp->nitems; j++)
 324           if (mlp->item[j]->msgid[0] == '\0' /* && !mlp->item[j]->obsolete */)
 325             {
 326               const char *header = mlp->item[j]->msgstr;
 327
 328               if (header != NULL)
 329                 {
 330                   const char *charsetstr = strstr (header, "charset=");
 331
 332                   if (charsetstr != NULL)
 333                     {
 334                       size_t len;
 335
 336                       charsetstr += strlen ("charset=");
 337                       len = strcspn (charsetstr, " \t\n");
 338                       if (len == strlen ("UTF-8")
 339                           && c_strncasecmp (charsetstr, "UTF-8", len) == 0)
 340                         was_utf8 = true;
 341                     }
 342                 }
 343             }
 344         }
 345     if (was_utf8)
 346       def = iconv_msgdomain_list (def, "UTF-8", fn1);
 347   }
 348
 349   empty_list = message_list_alloc (false);
 350
 351   /* Every entry in the xgettext generated file must be matched by a
 352      (single) entry in the human created file.  */
 353   nerrors = 0;
 354   if (!multi_domain_mode)
 355     for (k = 0; k < ref->nitems; k++)
 356       {
 357         const char *domain = ref->item[k]->domain;
 358         message_list_ty *refmlp = ref->item[k]->messages;
 359         message_list_ty *defmlp;
 360
 361         defmlp = msgdomain_list_sublist (def, domain, false);
 362         if (defmlp == NULL)
 363           defmlp = empty_list;
 364
 365         match_domain (fn1, fn2, defmlp, refmlp, &nerrors);
 366       }
 367   else
 368     {
 369       /* Apply the references messages in the default domain to each of
 370          the definition domains.  */
 371       message_list_ty *refmlp = ref->item[0]->messages;
 372
 373       for (k = 0; k < def->nitems; k++)
 374         {
 375           message_list_ty *defmlp = def->item[k]->messages;
 376
 377           /* Ignore the default message domain if it has no messages.  */
 378           if (k > 0 || defmlp->nitems > 0)
 379             match_domain (fn1, fn2, defmlp, refmlp, &nerrors);
 380         }
 381     }
 382
 383   /* Look for messages in the definition file, which are not present
 384      in the reference file, indicating messages which defined but not
 385      used in the program.  */
 386   for (k = 0; k < def->nitems; ++k)
 387     {
 388       message_list_ty *defmlp = def->item[k]->messages;
 389
 390       for (j = 0; j < defmlp->nitems; j++)
 391         {
 392           message_ty *defmsg = defmlp->item[j];
 393
 394           if (!defmsg->used)
 395             po_gram_error_at_line (&defmsg->pos,
 396                                    _("warning: this message is not used"));
 397         }
 398     }
 399
 400   /* Exit with status 1 on any error.  */
 401   if (nerrors > 0)
 402     error (EXIT_FAILURE, 0,
 403            ngettext ("found %d fatal error", "found %d fatal errors", nerrors),
 404            nerrors);
 405 }