src/simple-gettext.c

   1 /* simple-gettext.c  - a simplified version of gettext.
   2  * Copyright (C) 1995, 1996, 1997, 1998, 1999 Free Software Foundation, Inc.
   3  *
   4  * This file is part of GnuPG.
   5  *
   6  * GnuPG is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * GnuPG is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * along with this program; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA
  19  */
  20
  21 /* This is a simplified version of gettext written by Ulrich Drepper.
  22  * It is used for the Win32 version of GnuPG becuase all the overhead
  23  * of gettext is not needed and we have to do some special Win32 stuff.
  24  * I decided that this is far easier than to tweak gettext for the special
  25  * cases (I tried it but it is a lot of code).  wk 15.09.99
  26  */
  27
  28 #include <config.h>
  29 #ifdef USE_SIMPLE_GETTEXT
  30 #ifndef __MINGW32__
  31   #error This file can only be used with MinGW32
  32 #endif
  33
  34 #include <stdio.h>
  35 #include <stdlib.h>
  36 #include <string.h>
  37 #include <ctype.h>
  38 #include <errno.h>
  39 #include <sys/types.h>
  40 #include <sys/stat.h>
  41 #include <windows.h>
  42 #include "w32reg.h"
  43
  44 typedef unsigned int u32; /* That is fine with MingW32 */
  45
  46 typedef unsigned long ulong;
  47
  48 /* The magic number of the GNU message catalog format.  */
  49 #define MAGIC         0x950412de
  50 #define MAGIC_SWAPPED 0xde120495
  51
  52 /* Revision number of the currently used .mo (binary) file format.  */
  53 #define MO_REVISION_NUMBER 0
  54
  55
  56 /* Header for binary .mo file format.  */
  57 struct mo_file_header
  58 {
  59   /* The magic number.  */
  60   u32 magic;
  61   /* The revision number of the file format.  */
  62   u32 revision;
  63   /* The number of strings pairs.  */
  64   u32 nstrings;
  65   /* Offset of table with start offsets of original strings.  */
  66   u32 orig_tab_offset;
  67   /* Offset of table with start offsets of translation strings.  */
  68   u32 trans_tab_offset;
  69   /* Size of hashing table.  */
  70   u32 hash_tab_size;
  71   /* Offset of first hashing entry.  */
  72   u32 hash_tab_offset;
  73 };
  74
  75 struct string_desc
  76 {
  77   /* Length of addressed string.  */
  78   u32 length;
  79   /* Offset of string in file.  */
  80   u32 offset;
  81 };
  82
  83
  84
  85 struct loaded_domain
  86 {
  87   char *data;
  88   int must_swap;
  89   u32 nstrings;
  90   char *mapped;
  91   struct string_desc *orig_tab;
  92   struct string_desc *trans_tab;
  93   u32 hash_size;
  94   u32 *hash_tab;
  95 };
  96
  97
  98 static struct loaded_domain *the_domain;
  99
 100 static __inline__ u32
 101 do_swap_u32( u32 i )
 102 {
 103   return (i << 24) | ((i & 0xff00) << 8) | ((i >> 8) & 0xff00) | (i >> 24);
 104 }
 105
 106 #define SWAPIT(flag, data) ((flag) ? do_swap_u32(data) : (data) )
 107
 108
 109 /* We assume to have `unsigned long int' value with at least 32 bits.  */
 110 #define HASHWORDBITS 32
 111
 112 /* The so called `hashpjw' function by P.J. Weinberger
 113    [see Aho/Sethi/Ullman, COMPILERS: Principles, Techniques and Tools,
 114    1986, 1987 Bell Telephone Laboratories, Inc.]  */
 115
 116 static __inline__ ulong
 117 hash_string( const char *str_param )
 118 {
 119     unsigned long int hval, g;
 120     const char *str = str_param;
 121
 122     hval = 0;
 123     while (*str != '\0')
 124     {
 125         hval <<= 4;
 126         hval += (unsigned long int) *str++;
 127         g = hval & ((unsigned long int) 0xf << (HASHWORDBITS - 4));
 128         if (g != 0)
 129         {
 130           hval ^= g >> (HASHWORDBITS - 8);
 131           hval ^= g;
 132         }
 133     }
 134     return hval;
 135 }
 136
 137
 138 static struct loaded_domain *
 139 load_domain( const char *filename )
 140 {
 141     FILE *fp;
 142     size_t size;
 143     struct stat st;
 144     struct mo_file_header *data = NULL;
 145     struct loaded_domain *domain = NULL;
 146     size_t to_read;
 147     char *read_ptr;
 148
 149     fp = fopen( filename, "rb" );
 150     if( !fp )
 151        return NULL; /* can't open the file */
 152     /* we must know about the size of the file */
 153     if( fstat( fileno(fp ), &st )
 154         || (size = (size_t)st.st_size) != st.st_size
 155         || size < sizeof (struct mo_file_header) ) {
 156         fclose( fp );
 157         return NULL;
 158     }
 159
 160     data = malloc( size );
 161     if( !data ) {
 162         fclose( fp );
 163         return NULL; /* out of memory */
 164     }
 165
 166     to_read = size;
 167     read_ptr = (char *) data;
 168     do {
 169         long int nb = fread( read_ptr, 1, to_read, fp );
 170         if( nb < to_read ) {
 171             fclose (fp);
 172             free(data);
 173             return NULL; /* read error */
 174         }
 175         read_ptr += nb;
 176         to_read -= nb;
 177     } while( to_read > 0 );
 178     fclose (fp);
 179
 180     /* Using the magic number we can test whether it really is a message
 181      * catalog file.  */
 182     if( data->magic != MAGIC && data->magic != MAGIC_SWAPPED ) {
 183         /* The magic number is wrong: not a message catalog file.  */
 184         free( data );
 185         return NULL;
 186     }
 187
 188     domain = calloc( 1, sizeof *domain );
 189     if( !domain )  {
 190         free( data );
 191         return NULL;
 192     }
 193     domain->data = (char *) data;
 194     domain->must_swap = data->magic != MAGIC;
 195
 196     /* Fill in the information about the available tables.  */
 197     switch( SWAPIT(domain->must_swap, data->revision) ) {
 198       case 0:
 199         domain->nstrings = SWAPIT(domain->must_swap, data->nstrings);
 200         domain->orig_tab = (struct string_desc *)
 201           ((char *) data + SWAPIT(domain->must_swap, data->orig_tab_offset));
 202         domain->trans_tab = (struct string_desc *)
 203           ((char *) data + SWAPIT(domain->must_swap, data->trans_tab_offset));
 204         domain->hash_size = SWAPIT(domain->must_swap, data->hash_tab_size);
 205         domain->hash_tab = (u32 *)
 206           ((char *) data + SWAPIT(domain->must_swap, data->hash_tab_offset));
 207       break;
 208
 209       default: /* This is an invalid revision.  */
 210         free( data );
 211         free( domain );
 212         return NULL;
 213     }
 214
 215     /* allocate an array to keep track of code page mappings */
 216     domain->mapped = calloc( 1, domain->nstrings );
 217     if( !domain->mapped ) {
 218         free( data );
 219         free( domain );
 220         return NULL;
 221     }
 222
 223     return domain;
 224 }
 225
 226
 227 /****************
 228  * Set the file used for translations.  Pass a NULL to disable
 229  * translation.  A new filename may be set at anytime.
 230  * WARNING: After changing the filename you shoudl not access any data
 231  *          retrieved by gettext().
 232  */
 233 int
 234 set_gettext_file( const char *filename )
 235 {
 236     struct loaded_domain *domain = NULL;
 237
 238     if( filename && *filename ) {
 239         if( filename[0] == '/'
 240            #ifdef HAVE_DRIVE_LETTERS
 241             || ( isalpha(filename[0])
 242                  && filename[1] == ':'
 243                  && (filename[2] == '/' || filename[2] == '\\') )
 244            #endif
 245            ) {
 246             /* absolute path - use it as is */
 247             domain = load_domain( filename );
 248         }
 249         else { /* relative path - append ".mo" and get dir from the environment */
 250             char *buf = NULL;
 251             char *dir;
 252
 253             dir = read_w32_registry_string( NULL,
 254                                             "Control Panel\\Mingw32\\NLS",
 255                                             "MODir" );
 256             if( dir && (buf=malloc(strlen(dir)+strlen(filename)+1+3+1)) ) {
 257                 strcpy(stpcpy(stpcpy(stpcpy( buf, dir),"\\"), filename),".mo");
 258                 domain = load_domain( buf );
 259                 free(buf);
 260             }
 261             free(dir);
 262         }
 263         if( !domain ) {
 264             return -1; }
 265     }
 266     if( the_domain ) {
 267         free( the_domain->data );
 268         free( the_domain->mapped );
 269         free( the_domain );
 270         the_domain = NULL;
 271     }
 272     the_domain = domain;
 273     return NULL;
 274 }
 275
 276
 277 static const char*
 278 get_string( struct loaded_domain *domain, u32 idx )
 279 {
 280     char *p = domain->data + SWAPIT(domain->must_swap,
 281                                     domain->trans_tab[idx].offset);
 282
 283         /* status of domain->mapped[idx] is ignored.
 284          * not sure about the consequences.
 285          * perhaps mapped can entirely be removed?
 286          */
 287
 288         /* we assume, strings are already correctly
 289          * encoded.
 290          */
 291
 292     return (const char*)p;
 293 }
 294
 295
 296
 297 const char *
 298 gettext( const char *msgid )
 299 {
 300     struct loaded_domain *domain;
 301     size_t act = 0;
 302     size_t top, bottom;
 303
 304     if( !(domain = the_domain) ) {
 305         goto not_found;
 306         }
 307
 308     /* Locate the MSGID and its translation.  */
 309     if( domain->hash_size > 2 && domain->hash_tab ) {
 310         /* Use the hashing table.  */
 311         u32 len = strlen (msgid);
 312         u32 hash_val = hash_string (msgid);
 313         u32 idx = hash_val % domain->hash_size;
 314         u32 incr = 1 + (hash_val % (domain->hash_size - 2));
 315         u32 nstr = SWAPIT (domain->must_swap, domain->hash_tab[idx]);
 316
 317         if ( !nstr ) /* Hash table entry is empty.  */
 318             goto not_found;
 319
 320         if( SWAPIT(domain->must_swap,
 321                     domain->orig_tab[nstr - 1].length) == len
 322             && !strcmp( msgid,
 323                        domain->data + SWAPIT(domain->must_swap,
 324                                     domain->orig_tab[nstr - 1].offset)) )
 325             return get_string( domain, nstr - 1 );
 326
 327         for(;;) {
 328             if (idx >= domain->hash_size - incr)
 329                 idx -= domain->hash_size - incr;
 330             else
 331                 idx += incr;
 332
 333             nstr = SWAPIT(domain->must_swap, domain->hash_tab[idx]);
 334             if( !nstr )
 335                 goto not_found; /* Hash table entry is empty.  */
 336
 337             if ( SWAPIT(domain->must_swap,
 338                                 domain->orig_tab[nstr - 1].length) == len
 339                  && !strcmp (msgid,
 340                          domain->data + SWAPIT(domain->must_swap,
 341                                            domain->orig_tab[nstr - 1].offset)))
 342                 return get_string( domain, nstr-1 );
 343         }
 344         /* NOTREACHED */
 345     }
 346
 347     /* Now we try the default method:  binary search in the sorted
 348        array of messages.  */
 349     bottom = 0;
 350     top = domain->nstrings;
 351     while( bottom < top ) {
 352         int cmp_val;
 353
 354         act = (bottom + top) / 2;
 355         cmp_val = strcmp(msgid, domain->data
 356                                + SWAPIT(domain->must_swap,
 357                                         domain->orig_tab[act].offset));
 358         if (cmp_val < 0)
 359             top = act;
 360         else if (cmp_val > 0)
 361             bottom = act + 1;
 362         else
 363             return get_string( domain, act );
 364     }
 365
 366   not_found:
 367     return msgid;
 368 }
 369
 370 #if 0
 371        unsigned int cp1, cp2;
 372
 373        cp1 = GetConsoleCP();
 374        cp2 = GetConsoleOutputCP();
 375
 376        log_info("InputCP=%u  OutputCP=%u\n", cp1, cp2 );
 377
 378        if( !SetConsoleOutputCP( 1252 ) )
 379            log_info("SetConsoleOutputCP failed: %d\n", (int)GetLastError() );
 380
 381        cp1 = GetConsoleCP();
 382        cp2 = GetConsoleOutputCP();
 383        log_info("InputCP=%u  OutputCP=%u after switch1\n", cp1, cp2 );
 384 #endif
 385
 386 #endif /* USE_SIMPLE_GETTEXT */