locale/programs/localedef.c

   1 /* Copyright (C) 1995-1999, 2000, 2001, 2002 Free Software Foundation, Inc.
   2    This file is part of the GNU C Library.
   3    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1995.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Lesser General Public
   7    License as published by the Free Software Foundation; either
   8    version 2.1 of the License, or (at your option) any later version.
   9
  10    The GNU C Library 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 GNU
  13    Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public
  16    License along with the GNU C Library; if not, write to the Free
  17    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
  18    02111-1307 USA.  */
  19
  20 #ifdef HAVE_CONFIG_H
  21 # include <config.h>
  22 #endif
  23
  24 #include <argp.h>
  25 #include <errno.h>
  26 #include <fcntl.h>
  27 #include <libintl.h>
  28 #include <locale.h>
  29 #include <mcheck.h>
  30 #include <stdio.h>
  31 #include <stdlib.h>
  32 #include <string.h>
  33 #include <unistd.h>
  34 #include <sys/mman.h>
  35 #include <sys/stat.h>
  36
  37 #include "error.h"
  38 #include "charmap.h"
  39 #include "locfile.h"
  40
  41 /* Undefine the following line in the production version.  */
  42 /* #define NDEBUG 1 */
  43 #include <assert.h>
  44
  45
  46 /* List of copied locales.  */
  47 struct copy_def_list_t *copy_list;
  48
  49 /* If this is defined be POSIX conform.  */
  50 int posix_conformance;
  51
  52 /* If not zero give a lot more messages.  */
  53 int verbose;
  54
  55 /* If not zero suppress warnings and information messages.  */
  56 int be_quiet;
  57
  58 /* If not zero, produce old-style hash table instead of 3-level access tables.  */
  59 int oldstyle_tables;
  60
  61 /* If not zero force output even if warning were issued.  */
  62 static int force_output;
  63
  64 /* Prefix for output files.  */
  65 const char *output_prefix;
  66
  67 /* Name of the character map file.  */
  68 static const char *charmap_file;
  69
  70 /* Name of the locale definition file.  */
  71 static const char *input_file;
  72
  73 /* Name of the repertoire map file.  */
  74 const char *repertoire_global;
  75
  76 /* List of all locales.  */
  77 static struct localedef_t *locales;
  78
  79
  80 /* Name and version of program.  */
  81 static void print_version (FILE *stream, struct argp_state *state);
  82 void (*argp_program_version_hook) (FILE *, struct argp_state *) = print_version;
  83
  84 #define OPT_POSIX 1
  85 #define OPT_QUIET 2
  86 #define OPT_OLDSTYLE 3
  87 #define OPT_PREFIX 4
  88
  89 /* Definitions of arguments for argp functions.  */
  90 static const struct argp_option options[] =
  91 {
  92   { NULL, 0, NULL, 0, N_("Input Files:") },
  93   { "charmap", 'f', "FILE", 0,
  94     N_("Symbolic character names defined in FILE") },
  95   { "inputfile", 'i', "FILE", 0, N_("Source definitions are found in FILE") },
  96   { "repertoire-map", 'u', "FILE", 0,
  97     N_("FILE contains mapping from symbolic names to UCS4 values") },
  98
  99   { NULL, 0, NULL, 0, N_("Output control:") },
 100   { "force", 'c', NULL, 0,
 101     N_("Create output even if warning messages were issued") },
 102   { "old-style", OPT_OLDSTYLE, NULL, 0, N_("Create old-style tables") },
 103   { "prefix", OPT_PREFIX, "PATH", 0, N_("Optional output file prefix") },
 104   { "posix", OPT_POSIX, NULL, 0, N_("Be strictly POSIX conform") },
 105   { "quiet", OPT_QUIET, NULL, 0,
 106     N_("Suppress warnings and information messages") },
 107   { "verbose", 'v', NULL, 0, N_("Print more messages") },
 108   { NULL, 0, NULL, 0, NULL }
 109 };
 110
 111 /* Short description of program.  */
 112 static const char doc[] = N_("Compile locale specification");
 113
 114 /* Strings for arguments in help texts.  */
 115 static const char args_doc[] = N_("NAME");
 116
 117 /* Prototype for option handler.  */
 118 static error_t parse_opt (int key, char *arg, struct argp_state *state);
 119
 120 /* Function to print some extra text in the help message.  */
 121 static char *more_help (int key, const char *text, void *input);
 122
 123 /* Data structure to communicate with argp functions.  */
 124 static struct argp argp =
 125 {
 126   options, parse_opt, args_doc, doc, NULL, more_help
 127 };
 128
 129
 130 /* Prototypes for global functions.  */
 131 extern void *xmalloc (size_t __n);
 132
 133 /* Prototypes for local functions.  */
 134 static void error_print (void);
 135 static const char *construct_output_path (char *path);
 136 static const char *normalize_codeset (const char *codeset, size_t name_len);
 137
 138
 139 int
 140 main (int argc, char *argv[])
 141 {
 142   const char *output_path;
 143   int cannot_write_why;
 144   struct charmap_t *charmap;
 145   struct localedef_t global;
 146   int remaining;
 147
 148   /* Set initial values for global variables.  */
 149   copy_list = NULL;
 150   posix_conformance = getenv ("POSIXLY_CORRECT") != NULL;
 151   error_print_progname = error_print;
 152
 153   /* Set locale.  Do not set LC_ALL because the other categories must
 154      not be affected (according to POSIX.2).  */
 155   setlocale (LC_MESSAGES, "");
 156   setlocale (LC_CTYPE, "");
 157
 158   /* Initialize the message catalog.  */
 159   textdomain (_libc_intl_domainname);
 160
 161   /* Parse and process arguments.  */
 162   argp_err_exit_status = 4;
 163   argp_parse (&argp, argc, argv, 0, &remaining, NULL);
 164
 165   /* POSIX.2 requires to be verbose about missing characters in the
 166      character map.  */
 167   verbose |= posix_conformance;
 168
 169   if (argc - remaining != 1)
 170     {
 171       /* We need exactly one non-option parameter.  */
 172       argp_help (&argp, stdout, ARGP_HELP_SEE | ARGP_HELP_EXIT_ERR,
 173                  program_invocation_short_name);
 174       exit (4);
 175     }
 176
 177   /* The parameter describes the output path of the constructed files.
 178      If the described files cannot be written return a NULL pointer.  */
 179   output_path  = construct_output_path (argv[remaining]);
 180   if (output_path == NULL)
 181     error (4, errno, _("cannot create directory for output files"));
 182   cannot_write_why = errno;
 183
 184   /* Now that the parameters are processed we have to reset the local
 185      ctype locale.  (P1003.2 4.35.5.2)  */
 186   setlocale (LC_CTYPE, "POSIX");
 187
 188   /* Look whether the system really allows locale definitions.  POSIX
 189      defines error code 3 for this situation so I think it must be
 190      a fatal error (see P1003.2 4.35.8).  */
 191   if (sysconf (_SC_2_LOCALEDEF) < 0)
 192     error (3, 0, _("FATAL: system does not define `_POSIX2_LOCALEDEF'"));
 193
 194   /* Process charmap file.  */
 195   charmap = charmap_read (charmap_file, verbose, be_quiet, 1);
 196
 197   /* Add the first entry in the locale list.  */
 198   memset (&global, '\0', sizeof (struct localedef_t));
 199   global.name = input_file ?: "/dev/stdin";
 200   global.needed = ALL_LOCALES;
 201   locales = &global;
 202
 203   /* Now read the locale file.  */
 204   if (locfile_read (&global, charmap) != 0)
 205     error (4, errno, _("cannot open locale definition file `%s'"), input_file);
 206
 207   /* Perhaps we saw some `copy' instructions.  */
 208   while (1)
 209     {
 210       struct localedef_t *runp = locales;
 211
 212       while (runp != NULL && (runp->needed & runp->avail) == runp->needed)
 213         runp = runp->next;
 214
 215       if (runp == NULL)
 216         /* Everything read.  */
 217         break;
 218
 219       if (locfile_read (runp, charmap) != 0)
 220         error (4, errno, _("cannot open locale definition file `%s'"),
 221                runp->name);
 222     }
 223
 224   /* Check the categories we processed in source form.  */
 225   check_all_categories (locales, charmap);
 226
 227   /* We are now able to write the data files.  If warning were given we
 228      do it only if it is explicitly requested (--force).  */
 229   if (error_message_count == 0 || force_output != 0)
 230     {
 231       if (cannot_write_why != 0)
 232         error (4, cannot_write_why, _("cannot write output files to `%s'"),
 233                output_path);
 234       else
 235         write_all_categories (locales, charmap, output_path);
 236     }
 237   else
 238     error (4, 0, _("no output file produced because warning were issued"));
 239
 240   /* This exit status is prescribed by POSIX.2 4.35.7.  */
 241   exit (error_message_count != 0);
 242 }
 243
 244
 245 /* Handle program arguments.  */
 246 static error_t
 247 parse_opt (int key, char *arg, struct argp_state *state)
 248 {
 249   switch (key)
 250     {
 251     case OPT_QUIET:
 252       be_quiet = 1;
 253       break;
 254     case OPT_POSIX:
 255       posix_conformance = 1;
 256       break;
 257     case OPT_OLDSTYLE:
 258       oldstyle_tables = 1;
 259       break;
 260     case OPT_PREFIX:
 261       output_prefix = arg;
 262       break;
 263     case 'c':
 264       force_output = 1;
 265       break;
 266     case 'f':
 267       charmap_file = arg;
 268       break;
 269     case 'i':
 270       input_file = arg;
 271       break;
 272     case 'u':
 273       repertoire_global = arg;
 274       break;
 275     case 'v':
 276       verbose = 1;
 277       break;
 278     default:
 279       return ARGP_ERR_UNKNOWN;
 280     }
 281   return 0;
 282 }
 283
 284
 285 static char *
 286 more_help (int key, const char *text, void *input)
 287 {
 288   char *cp;
 289
 290   switch (key)
 291     {
 292     case ARGP_KEY_HELP_EXTRA:
 293       /* We print some extra information.  */
 294       asprintf (&cp, gettext ("\
 295 System's directory for character maps : %s\n\
 296                        repertoire maps: %s\n\
 297                        locale path    : %s\n\
 298 %s"),
 299                 CHARMAP_PATH, REPERTOIREMAP_PATH, LOCALE_PATH, gettext ("\
 300 Report bugs using the `glibcbug' script to <bugs@gnu.org>.\n"));
 301       return cp;
 302     default:
 303       break;
 304     }
 305   return (char *) text;
 306 }
 307
 308 /* Print the version information.  */
 309 static void
 310 print_version (FILE *stream, struct argp_state *state)
 311 {
 312   fprintf (stream, "localedef (GNU %s) %s\n", PACKAGE, VERSION);
 313   fprintf (stream, gettext ("\
 314 Copyright (C) %s Free Software Foundation, Inc.\n\
 315 This is free software; see the source for copying conditions.  There is NO\n\
 316 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\
 317 "), "2002");
 318   fprintf (stream, gettext ("Written by %s.\n"), "Ulrich Drepper");
 319 }
 320
 321
 322 /* The address of this function will be assigned to the hook in the error
 323    functions.  */
 324 static void
 325 error_print (void)
 326 {
 327 }
 328
 329
 330 /* The parameter to localedef describes the output path.  If it does
 331    contain a '/' character it is a relative path.  Otherwise it names the
 332    locale this definition is for.  */
 333 static const char *
 334 construct_output_path (char *path)
 335 {
 336   const char *normal = NULL;
 337   char *result;
 338   char *endp;
 339
 340   if (strchr (path, '/') == NULL)
 341     {
 342       /* This is a system path.  First examine whether the locale name
 343          contains a reference to the codeset.  This should be
 344          normalized.  */
 345       char *startp;
 346       size_t n;
 347
 348       startp = path;
 349       /* We must be prepared for finding a CEN name or a location of
 350          the introducing `.' where it is not possible anymore.  */
 351       while (*startp != '\0' && *startp != '@' && *startp != '.'
 352              && *startp != '+' && *startp != ',')
 353         ++startp;
 354       if (*startp == '.')
 355         {
 356           /* We found a codeset specification.  Now find the end.  */
 357           endp = ++startp;
 358           while (*endp != '\0' && *endp != '@')
 359             ++endp;
 360
 361           if (endp > startp)
 362             normal = normalize_codeset (startp, endp - startp);
 363         }
 364       else
 365         /* This is to keep gcc quiet.  */
 366         endp = NULL;
 367
 368       /* We put an additional '\0' at the end of the string because at
 369          the end of the function we need another byte for the trailing
 370          '/'.  */
 371       if (normal == NULL)
 372         n = asprintf (&result, "%s%s/%s%c",
 373                       output_prefix ?: "", LOCALEDIR, path, '\0');
 374       else
 375         n = asprintf (&result, "%s%s/%.*s%s%s%c",
 376                       output_prefix ?: "", LOCALEDIR,
 377                       (int) (startp - path), path, normal, endp, '\0');
 378
 379       if (n < 0)
 380         return NULL;
 381
 382       endp = result + n - 1;
 383     }
 384   else
 385     {
 386       /* This is a user path.  Please note the additional byte in the
 387          memory allocation.  */
 388       size_t len = strlen (path) + 1;
 389       result = xmalloc (len + 1);
 390       endp = mempcpy (result, path, len) - 1;
 391     }
 392
 393   errno = 0;
 394
 395   if (euidaccess (result, W_OK) == -1)
 396     /* Perhaps the directory does not exist now.  Try to create it.  */
 397     if (errno == ENOENT)
 398       {
 399         errno = 0;
 400         if (mkdir (result, 0777) < 0)
 401           return NULL;
 402       }
 403
 404   *endp++ = '/';
 405   *endp = '\0';
 406
 407   return result;
 408 }
 409
 410
 411 /* Normalize codeset name.  There is no standard for the codeset
 412    names.  Normalization allows the user to use any of the common
 413    names.  */
 414 static const char *
 415 normalize_codeset (codeset, name_len)
 416      const char *codeset;
 417      size_t name_len;
 418 {
 419   int len = 0;
 420   int only_digit = 1;
 421   char *retval;
 422   char *wp;
 423   size_t cnt;
 424
 425   for (cnt = 0; cnt < name_len; ++cnt)
 426     if (isalnum (codeset[cnt]))
 427       {
 428         ++len;
 429
 430         if (isalpha (codeset[cnt]))
 431           only_digit = 0;
 432       }
 433
 434   retval = (char *) malloc ((only_digit ? 3 : 0) + len + 1);
 435
 436   if (retval != NULL)
 437     {
 438       if (only_digit)
 439         wp = stpcpy (retval, "iso");
 440       else
 441         wp = retval;
 442
 443       for (cnt = 0; cnt < name_len; ++cnt)
 444         if (isalpha (codeset[cnt]))
 445           *wp++ = tolower (codeset[cnt]);
 446         else if (isdigit (codeset[cnt]))
 447           *wp++ = codeset[cnt];
 448
 449       *wp = '\0';
 450     }
 451
 452   return (const char *) retval;
 453 }
 454
 455
 456 struct localedef_t *
 457 add_to_readlist (int locale, const char *name, const char *repertoire_name,
 458                  int generate, struct localedef_t *copy_locale)
 459 {
 460   struct localedef_t *runp = locales;
 461
 462   while (runp != NULL && strcmp (name, runp->name) != 0)
 463     runp = runp->next;
 464
 465   if (runp == NULL)
 466     {
 467       /* Add a new entry at the end.  */
 468       struct localedef_t *newp;
 469
 470       assert (generate == 1);
 471
 472       newp = xcalloc (1, sizeof (struct localedef_t));
 473       newp->name = name;
 474       newp->repertoire_name = repertoire_name;
 475
 476       if (locales == NULL)
 477         runp = locales = newp;
 478       else
 479         {
 480           runp = locales;
 481           while (runp->next != NULL)
 482             runp = runp->next;
 483           runp = runp->next = newp;
 484         }
 485     }
 486
 487   if (generate && (runp->needed & (1 << locale)) != 0)
 488     error (5, 0, _("circular dependencies between locale definitions"));
 489
 490   if (copy_locale != NULL)
 491     {
 492       if (runp->categories[locale].generic != NULL)
 493         error (5, 0, _("cannot add already read locale `%s' a second time"),
 494                name);
 495       else
 496         runp->categories[locale].generic =
 497           copy_locale->categories[locale].generic;
 498     }
 499
 500   runp->needed |= 1 << locale;
 501
 502   return runp;
 503 }
 504
 505
 506 struct localedef_t *
 507 find_locale (int locale, const char *name, const char *repertoire_name,
 508              const struct charmap_t *charmap)
 509 {
 510   struct localedef_t *result;
 511
 512   /* Find the locale, but do not generate it since this would be a bug.  */
 513   result = add_to_readlist (locale, name, repertoire_name, 0, NULL);
 514
 515   assert (result != NULL);
 516
 517   if ((result->avail & (1 << locale)) == 0
 518       && locfile_read (result, charmap) != 0)
 519     error (4, errno, _("cannot open locale definition file `%s'"),
 520            result->name);
 521
 522   return result;
 523 }
 524
 525
 526 struct localedef_t *
 527 load_locale (int locale, const char *name, const char *repertoire_name,
 528              const struct charmap_t *charmap, struct localedef_t *copy_locale)
 529 {
 530   struct localedef_t *result;
 531
 532   /* Generate the locale if it does not exist.  */
 533   result = add_to_readlist (locale, name, repertoire_name, 1, copy_locale);
 534
 535   assert (result != NULL);
 536
 537   if ((result->avail & (1 << locale)) == 0
 538       && locfile_read (result, charmap) != 0)
 539     error (4, errno, _("cannot open locale definition file `%s'"),
 540            result->name);
 541
 542   return result;
 543 }
 544
 545 static void
 546 turn_on_mcheck (void)
 547 {
 548   /* Enable `malloc' debugging.  */
 549   mcheck (NULL);
 550   /* Use the following line for a more thorough but much slower testing.  */
 551   /* mcheck_pedantic (NULL); */
 552 }
 553
 554 void (*__malloc_initialize_hook) (void) = turn_on_mcheck;