source/lib/util_unistr.c

   1 /*
   2    Unix SMB/CIFS implementation.
   3    Samba utility functions
   4    Copyright (C) Andrew Tridgell 1992-2001
   5    Copyright (C) Simo Sorce 2001
   6    Copyright (C) Jeremy Allison 2005
   7
   8    This program is free software; you can redistribute it and/or modify
   9    it under the terms of the GNU General Public License as published by
  10    the Free Software Foundation; either version 3 of the License, or
  11    (at your option) any later version.
  12
  13    This program is distributed in the hope that it will be useful,
  14    but WITHOUT ANY WARRANTY; without even the implied warranty of
  15    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16    GNU General Public License for more details.
  17
  18    You should have received a copy of the GNU General Public License
  19    along with this program.  If not, see <http://www.gnu.org/licenses/>.
  20 */
  21
  22 #include "includes.h"
  23
  24 #ifndef MAXUNI
  25 #define MAXUNI 1024
  26 #endif
  27
  28 /* these 3 tables define the unicode case handling.  They are loaded
  29    at startup either via mmap() or read() from the lib directory */
  30 static smb_ucs2_t *upcase_table;
  31 static smb_ucs2_t *lowcase_table;
  32 static uint8 *valid_table;
  33 static bool upcase_table_use_unmap;
  34 static bool lowcase_table_use_unmap;
  35 static bool valid_table_use_unmap;
  36 static bool initialized;
  37
  38 /**
  39  * Destroy global objects allocated by load_case_tables()
  40  **/
  41 void gfree_case_tables(void)
  42 {
  43         if ( upcase_table ) {
  44                 if ( upcase_table_use_unmap )
  45                         unmap_file(upcase_table, 0x20000);
  46                 else
  47                         SAFE_FREE(upcase_table);
  48         }
  49
  50         if ( lowcase_table ) {
  51                 if ( lowcase_table_use_unmap )
  52                         unmap_file(lowcase_table, 0x20000);
  53                 else
  54                         SAFE_FREE(lowcase_table);
  55         }
  56
  57         if ( valid_table ) {
  58                 if ( valid_table_use_unmap )
  59                         unmap_file(valid_table, 0x10000);
  60                 else
  61                         SAFE_FREE(valid_table);
  62         }
  63         initialized = false;
  64 }
  65
  66 /**
  67  * Load or generate the case handling tables.
  68  *
  69  * The case tables are defined in UCS2 and don't depend on any
  70  * configured parameters, so they never need to be reloaded.
  71  **/
  72
  73 void load_case_tables(void)
  74 {
  75         char *old_locale = NULL, *saved_locale = NULL;
  76         int i;
  77         TALLOC_CTX *frame = NULL;
  78
  79         if (initialized) {
  80                 return;
  81         }
  82         initialized = true;
  83
  84         frame = talloc_stackframe();
  85
  86         upcase_table = (smb_ucs2_t *)map_file(data_path("upcase.dat"),
  87                                               0x20000);
  88         upcase_table_use_unmap = ( upcase_table != NULL );
  89
  90         lowcase_table = (smb_ucs2_t *)map_file(data_path("lowcase.dat"),
  91                                                0x20000);
  92         lowcase_table_use_unmap = ( lowcase_table != NULL );
  93
  94 #ifdef HAVE_SETLOCALE
  95         /* Get the name of the current locale.  */
  96         old_locale = setlocale(LC_ALL, NULL);
  97
  98         if (old_locale) {
  99                 /* Save it as it is in static storage. */
 100                 saved_locale = SMB_STRDUP(old_locale);
 101         }
 102
 103         /* We set back the locale to C to get ASCII-compatible toupper/lower functions. */
 104         setlocale(LC_ALL, "C");
 105 #endif
 106
 107         /* we would like Samba to limp along even if these tables are
 108            not available */
 109         if (!upcase_table) {
 110                 DEBUG(1,("creating lame upcase table\n"));
 111                 upcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
 112                 for (i=0;i<0x10000;i++) {
 113                         smb_ucs2_t v;
 114                         SSVAL(&v, 0, i);
 115                         upcase_table[v] = i;
 116                 }
 117                 for (i=0;i<256;i++) {
 118                         smb_ucs2_t v;
 119                         SSVAL(&v, 0, UCS2_CHAR(i));
 120                         upcase_table[v] = UCS2_CHAR(islower(i)?toupper(i):i);
 121                 }
 122         }
 123
 124         if (!lowcase_table) {
 125                 DEBUG(1,("creating lame lowcase table\n"));
 126                 lowcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
 127                 for (i=0;i<0x10000;i++) {
 128                         smb_ucs2_t v;
 129                         SSVAL(&v, 0, i);
 130                         lowcase_table[v] = i;
 131                 }
 132                 for (i=0;i<256;i++) {
 133                         smb_ucs2_t v;
 134                         SSVAL(&v, 0, UCS2_CHAR(i));
 135                         lowcase_table[v] = UCS2_CHAR(isupper(i)?tolower(i):i);
 136                 }
 137         }
 138
 139 #ifdef HAVE_SETLOCALE
 140         /* Restore the old locale. */
 141         if (saved_locale) {
 142                 setlocale (LC_ALL, saved_locale);
 143                 SAFE_FREE(saved_locale);
 144         }
 145 #endif
 146         TALLOC_FREE(frame);
 147 }
 148
 149 static int check_dos_char_slowly(smb_ucs2_t c)
 150 {
 151         char buf[10];
 152         smb_ucs2_t c2 = 0;
 153         int len1, len2;
 154
 155         len1 = convert_string(CH_UTF16LE, CH_DOS, &c, 2, buf, sizeof(buf),False);
 156         if (len1 == 0) {
 157                 return 0;
 158         }
 159         len2 = convert_string(CH_DOS, CH_UTF16LE, buf, len1, &c2, 2,False);
 160         if (len2 != 2) {
 161                 return 0;
 162         }
 163         return (c == c2);
 164 }
 165
 166 /**
 167  * Load the valid character map table from <tt>valid.dat</tt> or
 168  * create from the configured codepage.
 169  *
 170  * This function is called whenever the configuration is reloaded.
 171  * However, the valid character table is not changed if it's loaded
 172  * from a file, because we can't unmap files.
 173  **/
 174
 175 void init_valid_table(void)
 176 {
 177         static int mapped_file;
 178         int i;
 179         const char *allowed = ".!#$%&'()_-@^`~";
 180         uint8 *valid_file;
 181
 182         if (mapped_file) {
 183                 /* Can't unmap files, so stick with what we have */
 184                 return;
 185         }
 186
 187         valid_file = (uint8 *)map_file(data_path("valid.dat"), 0x10000);
 188         if (valid_file) {
 189                 valid_table = valid_file;
 190                 mapped_file = 1;
 191                 valid_table_use_unmap = True;
 192                 return;
 193         }
 194
 195         /* Otherwise, we're using a dynamically created valid_table.
 196          * It might need to be regenerated if the code page changed.
 197          * We know that we're not using a mapped file, so we can
 198          * free() the old one. */
 199         SAFE_FREE(valid_table);
 200
 201         /* use free rather than unmap */
 202         valid_table_use_unmap = False;
 203
 204         DEBUG(2,("creating default valid table\n"));
 205         valid_table = (uint8 *)SMB_MALLOC(0x10000);
 206         SMB_ASSERT(valid_table != NULL);
 207         for (i=0;i<128;i++) {
 208                 valid_table[i] = isalnum(i) || strchr(allowed,i);
 209         }
 210
 211         lazy_initialize_conv();
 212
 213         for (;i<0x10000;i++) {
 214                 smb_ucs2_t c;
 215                 SSVAL(&c, 0, i);
 216                 valid_table[i] = check_dos_char_slowly(c);
 217         }
 218 }
 219
 220 /*******************************************************************
 221  Write a string in (little-endian) unicode format. src is in
 222  the current DOS codepage. len is the length in bytes of the
 223  string pointed to by dst.
 224
 225  if null_terminate is True then null terminate the packet (adds 2 bytes)
 226
 227  the return value is the length in bytes consumed by the string, including the
 228  null termination if applied
 229 ********************************************************************/
 230
 231 size_t dos_PutUniCode(char *dst,const char *src, size_t len, bool null_terminate)
 232 {
 233         int flags = null_terminate ? STR_UNICODE|STR_NOALIGN|STR_TERMINATE
 234                                    : STR_UNICODE|STR_NOALIGN;
 235         return push_ucs2(NULL, dst, src, len, flags);
 236 }
 237
 238
 239 /*******************************************************************
 240  Skip past a unicode string, but not more than len. Always move
 241  past a terminating zero if found.
 242 ********************************************************************/
 243
 244 char *skip_unibuf(char *src, size_t len)
 245 {
 246         char *srcend = src + len;
 247
 248         while (src < srcend && SVAL(src,0)) {
 249                 src += 2;
 250         }
 251
 252         if(!SVAL(src,0)) {
 253                 src += 2;
 254         }
 255
 256         return src;
 257 }
 258
 259 /* Copy a string from little-endian or big-endian unicode source (depending
 260  * on flags) to internal samba format destination
 261  */
 262
 263 int rpcstr_pull(char* dest, void *src, int dest_len, int src_len, int flags)
 264 {
 265         if (!src) {
 266                 dest[0] = 0;
 267                 return 0;
 268         }
 269         if(dest_len==-1) {
 270                 dest_len=MAXUNI-3;
 271         }
 272         return pull_ucs2(NULL, dest, src, dest_len, src_len, flags|STR_UNICODE|STR_NOALIGN);
 273 }
 274
 275 /* Copy a string from little-endian or big-endian unicode source (depending
 276  * on flags) to internal samba format destination. Allocates on talloc ctx.
 277  */
 278
 279 int rpcstr_pull_talloc(TALLOC_CTX *ctx,
 280                         char **dest,
 281                         void *src,
 282                         int src_len,
 283                         int flags)
 284 {
 285         return pull_ucs2_base_talloc(ctx,
 286                         NULL,
 287                         dest,
 288                         src,
 289                         src_len,
 290                         flags|STR_UNICODE|STR_NOALIGN);
 291
 292 }
 293
 294 /* Copy a string from a unistr2 source to internal samba format
 295    destination.  Use this instead of direct calls to rpcstr_pull() to avoid
 296    having to determine whether the source string is null terminated. */
 297
 298 int rpcstr_pull_unistr2_fstring(char *dest, UNISTR2 *src)
 299 {
 300         return pull_ucs2(NULL, dest, src->buffer, sizeof(fstring),
 301                          src->uni_str_len * 2, 0);
 302 }
 303
 304 /* Helper function to return a talloc'ed string. I have implemented it with a
 305  * copy because I don't really know how pull_ucs2 and friends calculate the
 306  * target size. If this turns out to be a major bottleneck someone with deeper
 307  * multi-byte knowledge needs to revisit this.
 308  * I just did (JRA :-). No longer uses copy.
 309  * My (VL) use is dsr_getdcname, which returns 6 strings, the alternative would
 310  * have been to manually talloc_strdup them in rpc_client/cli_netlogon.c.
 311  */
 312
 313 char *rpcstr_pull_unistr2_talloc(TALLOC_CTX *ctx, const UNISTR2 *src)
 314 {
 315         char *dest = NULL;
 316         size_t dest_len = convert_string_talloc(ctx,
 317                                 CH_UTF16LE,
 318                                 CH_UNIX,
 319                                 src->buffer,
 320                                 src->uni_str_len * 2,
 321                                 (void *)&dest,
 322                                 true);
 323         if (dest_len == (size_t)-1) {
 324                 return NULL;
 325         }
 326
 327         /* Ensure we're returning a null terminated string. */
 328         if (dest_len) {
 329                 /* Did we already process the terminating zero ? */
 330                 if (dest[dest_len-1] != 0) {
 331                         size_t size = talloc_get_size(dest);
 332                         /* Have we got space to append the '\0' ? */
 333                         if (size <= dest_len) {
 334                                 /* No, realloc. */
 335                                 dest = TALLOC_REALLOC_ARRAY(ctx, dest, char,
 336                                                 dest_len+1);
 337                                 if (!dest) {
 338                                         /* talloc fail. */
 339                                         dest_len = (size_t)-1;
 340                                         return NULL;
 341                                 }
 342                         }
 343                         /* Yay - space ! */
 344                         dest[dest_len] = '\0';
 345                         dest_len++;
 346                 }
 347         } else if (dest) {
 348                 dest[0] = 0;
 349         }
 350
 351         return dest;
 352 }
 353
 354 /* Converts a string from internal samba format to unicode
 355  */
 356
 357 int rpcstr_push(void *dest, const char *src, size_t dest_len, int flags)
 358 {
 359         return push_ucs2(NULL, dest, src, dest_len, flags|STR_UNICODE|STR_NOALIGN);
 360 }
 361
 362 /* Converts a string from internal samba format to unicode. Always terminates.
 363  * Actually just a wrapper round push_ucs2_talloc().
 364  */
 365
 366 int rpcstr_push_talloc(TALLOC_CTX *ctx, smb_ucs2_t **dest, const char *src)
 367 {
 368         return push_ucs2_talloc(ctx, dest, src);
 369 }
 370
 371 /*******************************************************************
 372  Convert a (little-endian) UNISTR2 structure to an ASCII string.
 373 ********************************************************************/
 374
 375 void unistr2_to_ascii(char *dest, const UNISTR2 *str, size_t maxlen)
 376 {
 377         if ((str == NULL) || (str->uni_str_len == 0)) {
 378                 *dest='\0';
 379                 return;
 380         }
 381         pull_ucs2(NULL, dest, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
 382 }
 383
 384 #if 0
 385 /*******************************************************************
 386  Convert a (little-endian) UNISTR3 structure to an ASCII string.
 387 ********************************************************************/
 388
 389 void unistr3_to_ascii(char *dest, const UNISTR3 *str, size_t maxlen)
 390 {
 391         if ((str == NULL) || (str->uni_str_len == 0)) {
 392                 *dest='\0';
 393                 return;
 394         }
 395         pull_ucs2(NULL, dest, str->str.buffer, maxlen, str->uni_str_len*2,
 396                   STR_NOALIGN);
 397 }
 398 #endif
 399
 400 /*******************************************************************
 401  Duplicate a UNISTR2 string into a null terminated char*
 402  using a talloc context.
 403 ********************************************************************/
 404
 405 char *unistr2_to_ascii_talloc(TALLOC_CTX *ctx, const UNISTR2 *str)
 406 {
 407         char *s = NULL;
 408
 409         if (!str || !str->buffer) {
 410                 return NULL;
 411         }
 412         if (pull_ucs2_base_talloc(ctx,
 413                                 NULL,
 414                                 &s,
 415                                 str->buffer,
 416                                 str->uni_str_len*2,
 417                                 STR_NOALIGN) == (size_t)-1) {
 418                 return NULL;
 419         }
 420         return s;
 421 }
 422
 423 /*******************************************************************
 424  Return a string for displaying a UNISTR2. Guarentees to return a
 425  valid string - "" if nothing else.
 426  Changed to use talloc_tos() under the covers.... JRA.
 427 ********************************************************************/
 428
 429 const char *unistr2_static(const UNISTR2 *str)
 430 {
 431         char *dest = NULL;
 432
 433         if ((str == NULL) || (str->uni_str_len == 0)) {
 434                 return "";
 435         }
 436
 437         dest = unistr2_to_ascii_talloc(talloc_tos(), str);
 438         if (!dest) {
 439                 return "";
 440         }
 441
 442         return dest;
 443 }
 444
 445 /*******************************************************************
 446  Convert a wchar to upper case.
 447 ********************************************************************/
 448
 449 smb_ucs2_t toupper_w(smb_ucs2_t val)
 450 {
 451         return upcase_table[SVAL(&val,0)];
 452 }
 453
 454 /*******************************************************************
 455  Convert a wchar to lower case.
 456 ********************************************************************/
 457
 458 smb_ucs2_t tolower_w( smb_ucs2_t val )
 459 {
 460         return lowcase_table[SVAL(&val,0)];
 461 }
 462
 463 /*******************************************************************
 464  Determine if a character is lowercase.
 465 ********************************************************************/
 466
 467 bool islower_w(smb_ucs2_t c)
 468 {
 469         return upcase_table[SVAL(&c,0)] != c;
 470 }
 471
 472 /*******************************************************************
 473  Determine if a character is uppercase.
 474 ********************************************************************/
 475
 476 bool isupper_w(smb_ucs2_t c)
 477 {
 478         return lowcase_table[SVAL(&c,0)] != c;
 479 }
 480
 481 /*******************************************************************
 482  Determine if a character is valid in a 8.3 name.
 483 ********************************************************************/
 484
 485 bool isvalid83_w(smb_ucs2_t c)
 486 {
 487         return valid_table[SVAL(&c,0)] != 0;
 488 }
 489
 490 /*******************************************************************
 491  Count the number of characters in a smb_ucs2_t string.
 492 ********************************************************************/
 493
 494 size_t strlen_w(const smb_ucs2_t *src)
 495 {
 496         size_t len;
 497         smb_ucs2_t c;
 498
 499         for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
 500                 ;
 501         }
 502
 503         return len;
 504 }
 505
 506 /*******************************************************************
 507  Count up to max number of characters in a smb_ucs2_t string.
 508 ********************************************************************/
 509
 510 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
 511 {
 512         size_t len;
 513         smb_ucs2_t c;
 514
 515         for(len = 0; (len < max) && *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
 516                 ;
 517         }
 518
 519         return len;
 520 }
 521
 522 /*******************************************************************
 523  Wide strchr().
 524 ********************************************************************/
 525
 526 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
 527 {
 528         smb_ucs2_t cp;
 529         while (*(COPY_UCS2_CHAR(&cp,s))) {
 530                 if (c == cp) {
 531                         return (smb_ucs2_t *)s;
 532                 }
 533                 s++;
 534         }
 535         if (c == cp) {
 536                 return (smb_ucs2_t *)s;
 537         }
 538
 539         return NULL;
 540 }
 541
 542 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
 543 {
 544         return strchr_w(s, UCS2_CHAR(c));
 545 }
 546
 547 /*******************************************************************
 548  Wide strrchr().
 549 ********************************************************************/
 550
 551 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
 552 {
 553         smb_ucs2_t cp;
 554         const smb_ucs2_t *p = s;
 555         int len = strlen_w(s);
 556
 557         if (len == 0) {
 558                 return NULL;
 559         }
 560         p += (len - 1);
 561         do {
 562                 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
 563                         return (smb_ucs2_t *)p;
 564                 }
 565         } while (p-- != s);
 566         return NULL;
 567 }
 568
 569 /*******************************************************************
 570  Wide version of strrchr that returns after doing strrchr 'n' times.
 571 ********************************************************************/
 572
 573 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
 574 {
 575         smb_ucs2_t cp;
 576         const smb_ucs2_t *p = s;
 577         int len = strlen_w(s);
 578
 579         if (len == 0 || !n) {
 580                 return NULL;
 581         }
 582         p += (len - 1);
 583         do {
 584                 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
 585                         n--;
 586                 }
 587
 588                 if (!n) {
 589                         return (smb_ucs2_t *)p;
 590                 }
 591         } while (p-- != s);
 592         return NULL;
 593 }
 594
 595 /*******************************************************************
 596  Wide strstr().
 597 ********************************************************************/
 598
 599 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
 600 {
 601         smb_ucs2_t *r;
 602         size_t inslen;
 603
 604         if (!s || !*s || !ins || !*ins) {
 605                 return NULL;
 606         }
 607
 608         inslen = strlen_w(ins);
 609         r = (smb_ucs2_t *)s;
 610
 611         while ((r = strchr_w(r, *ins))) {
 612                 if (strncmp_w(r, ins, inslen) == 0) {
 613                         return r;
 614                 }
 615                 r++;
 616         }
 617
 618         return NULL;
 619 }
 620
 621 /*******************************************************************
 622  Convert a string to lower case.
 623  return True if any char is converted
 624 ********************************************************************/
 625
 626 bool strlower_w(smb_ucs2_t *s)
 627 {
 628         smb_ucs2_t cp;
 629         bool ret = False;
 630
 631         while (*(COPY_UCS2_CHAR(&cp,s))) {
 632                 smb_ucs2_t v = tolower_w(cp);
 633                 if (v != cp) {
 634                         COPY_UCS2_CHAR(s,&v);
 635                         ret = True;
 636                 }
 637                 s++;
 638         }
 639         return ret;
 640 }
 641
 642 /*******************************************************************
 643  Convert a string to upper case.
 644  return True if any char is converted
 645 ********************************************************************/
 646
 647 bool strupper_w(smb_ucs2_t *s)
 648 {
 649         smb_ucs2_t cp;
 650         bool ret = False;
 651         while (*(COPY_UCS2_CHAR(&cp,s))) {
 652                 smb_ucs2_t v = toupper_w(cp);
 653                 if (v != cp) {
 654                         COPY_UCS2_CHAR(s,&v);
 655                         ret = True;
 656                 }
 657                 s++;
 658         }
 659         return ret;
 660 }
 661
 662 /*******************************************************************
 663  Convert a string to "normal" form.
 664 ********************************************************************/
 665
 666 void strnorm_w(smb_ucs2_t *s, int case_default)
 667 {
 668         if (case_default == CASE_UPPER) {
 669                 strupper_w(s);
 670         } else {
 671                 strlower_w(s);
 672         }
 673 }
 674
 675 int strcmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
 676 {
 677         smb_ucs2_t cpa, cpb;
 678
 679         while ((*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
 680                 a++;
 681                 b++;
 682         }
 683         return (*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b)));
 684         /* warning: if *a != *b and both are not 0 we return a random
 685                 greater or lesser than 0 number not realted to which
 686                 string is longer */
 687 }
 688
 689 int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
 690 {
 691         smb_ucs2_t cpa, cpb;
 692         size_t n = 0;
 693
 694         while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
 695                 a++;
 696                 b++;
 697                 n++;
 698         }
 699         return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
 700 }
 701
 702 /*******************************************************************
 703  Case insensitive string comparison.
 704 ********************************************************************/
 705
 706 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
 707 {
 708         smb_ucs2_t cpa, cpb;
 709
 710         while ((*COPY_UCS2_CHAR(&cpb,b)) && toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb)) {
 711                 a++;
 712                 b++;
 713         }
 714         return (tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b))));
 715 }
 716
 717 /*******************************************************************
 718  Case insensitive string comparison, length limited.
 719 ********************************************************************/
 720
 721 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
 722 {
 723         smb_ucs2_t cpa, cpb;
 724         size_t n = 0;
 725
 726         while ((n < len) && *COPY_UCS2_CHAR(&cpb,b) && (toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb))) {
 727                 a++;
 728                 b++;
 729                 n++;
 730         }
 731         return (len - n)?(tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b)))):0;
 732 }
 733
 734 /*******************************************************************
 735  Compare 2 strings.
 736 ********************************************************************/
 737
 738 bool strequal_w(const smb_ucs2_t *s1, const smb_ucs2_t *s2)
 739 {
 740         if (s1 == s2) {
 741                 return(True);
 742         }
 743         if (!s1 || !s2) {
 744                 return(False);
 745         }
 746
 747         return(strcasecmp_w(s1,s2)==0);
 748 }
 749
 750 /*******************************************************************
 751  Compare 2 strings up to and including the nth char.
 752 ******************************************************************/
 753
 754 bool strnequal_w(const smb_ucs2_t *s1,const smb_ucs2_t *s2,size_t n)
 755 {
 756         if (s1 == s2) {
 757                 return(True);
 758         }
 759         if (!s1 || !s2 || !n) {
 760                 return(False);
 761         }
 762
 763         return(strncasecmp_w(s1,s2,n)==0);
 764 }
 765
 766 /*******************************************************************
 767  Duplicate string.
 768 ********************************************************************/
 769
 770 smb_ucs2_t *strdup_w(const smb_ucs2_t *src)
 771 {
 772         return strndup_w(src, 0);
 773 }
 774
 775 /* if len == 0 then duplicate the whole string */
 776
 777 smb_ucs2_t *strndup_w(const smb_ucs2_t *src, size_t len)
 778 {
 779         smb_ucs2_t *dest;
 780
 781         if (!len) {
 782                 len = strlen_w(src);
 783         }
 784         dest = SMB_MALLOC_ARRAY(smb_ucs2_t, len + 1);
 785         if (!dest) {
 786                 DEBUG(0,("strdup_w: out of memory!\n"));
 787                 return NULL;
 788         }
 789
 790         memcpy(dest, src, len * sizeof(smb_ucs2_t));
 791         dest[len] = 0;
 792         return dest;
 793 }
 794
 795 /*******************************************************************
 796  Copy a string with max len.
 797 ********************************************************************/
 798
 799 smb_ucs2_t *strncpy_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
 800 {
 801         smb_ucs2_t cp;
 802         size_t len;
 803
 804         if (!dest || !src) {
 805                 return NULL;
 806         }
 807
 808         for (len = 0; (*COPY_UCS2_CHAR(&cp,(src+len))) && (len < max); len++) {
 809                 cp = *COPY_UCS2_CHAR(dest+len,src+len);
 810         }
 811         cp = 0;
 812         for ( /*nothing*/ ; len < max; len++ ) {
 813                 cp = *COPY_UCS2_CHAR(dest+len,&cp);
 814         }
 815
 816         return dest;
 817 }
 818
 819 /*******************************************************************
 820  Append a string of len bytes and add a terminator.
 821 ********************************************************************/
 822
 823 smb_ucs2_t *strncat_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
 824 {
 825         size_t start;
 826         size_t len;
 827         smb_ucs2_t z = 0;
 828
 829         if (!dest || !src) {
 830                 return NULL;
 831         }
 832
 833         start = strlen_w(dest);
 834         len = strnlen_w(src, max);
 835
 836         memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
 837         z = *COPY_UCS2_CHAR(dest+start+len,&z);
 838
 839         return dest;
 840 }
 841
 842 smb_ucs2_t *strcat_w(smb_ucs2_t *dest, const smb_ucs2_t *src)
 843 {
 844         size_t start;
 845         size_t len;
 846         smb_ucs2_t z = 0;
 847
 848         if (!dest || !src) {
 849                 return NULL;
 850         }
 851
 852         start = strlen_w(dest);
 853         len = strlen_w(src);
 854
 855         memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
 856         z = *COPY_UCS2_CHAR(dest+start+len,&z);
 857
 858         return dest;
 859 }
 860
 861
 862 /*******************************************************************
 863  Replace any occurence of oldc with newc in unicode string.
 864 ********************************************************************/
 865
 866 void string_replace_w(smb_ucs2_t *s, smb_ucs2_t oldc, smb_ucs2_t newc)
 867 {
 868         smb_ucs2_t cp;
 869
 870         for(;*(COPY_UCS2_CHAR(&cp,s));s++) {
 871                 if(cp==oldc) {
 872                         COPY_UCS2_CHAR(s,&newc);
 873                 }
 874         }
 875 }
 876
 877 /*******************************************************************
 878  Trim unicode string.
 879 ********************************************************************/
 880
 881 bool trim_string_w(smb_ucs2_t *s, const smb_ucs2_t *front,
 882                                   const smb_ucs2_t *back)
 883 {
 884         bool ret = False;
 885         size_t len, front_len, back_len;
 886
 887         if (!s) {
 888                 return False;
 889         }
 890
 891         len = strlen_w(s);
 892
 893         if (front && *front) {
 894                 front_len = strlen_w(front);
 895                 while (len && strncmp_w(s, front, front_len) == 0) {
 896                         memmove(s, (s + front_len), (len - front_len + 1) * sizeof(smb_ucs2_t));
 897                         len -= front_len;
 898                         ret = True;
 899                 }
 900         }
 901
 902         if (back && *back) {
 903                 back_len = strlen_w(back);
 904                 while (len && strncmp_w((s + (len - back_len)), back, back_len) == 0) {
 905                         s[len - back_len] = 0;
 906                         len -= back_len;
 907                         ret = True;
 908                 }
 909         }
 910
 911         return ret;
 912 }
 913
 914 /*
 915   The *_wa() functions take a combination of 7 bit ascii
 916   and wide characters They are used so that you can use string
 917   functions combining C string constants with ucs2 strings
 918
 919   The char* arguments must NOT be multibyte - to be completely sure
 920   of this only pass string constants */
 921
 922 int strcmp_wa(const smb_ucs2_t *a, const char *b)
 923 {
 924         smb_ucs2_t cp = 0;
 925
 926         while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
 927                 a++;
 928                 b++;
 929         }
 930         return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
 931 }
 932
 933 int strncmp_wa(const smb_ucs2_t *a, const char *b, size_t len)
 934 {
 935         smb_ucs2_t cp = 0;
 936         size_t n = 0;
 937
 938         while ((n < len) && *b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
 939                 a++;
 940                 b++;
 941                 n++;
 942         }
 943         return (len - n)?(*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b)):0;
 944 }
 945
 946 smb_ucs2_t *strpbrk_wa(const smb_ucs2_t *s, const char *p)
 947 {
 948         smb_ucs2_t cp;
 949
 950         while (*(COPY_UCS2_CHAR(&cp,s))) {
 951                 int i;
 952                 for (i=0; p[i] && cp != UCS2_CHAR(p[i]); i++)
 953                         ;
 954                 if (p[i]) {
 955                         return (smb_ucs2_t *)s;
 956                 }
 957                 s++;
 958         }
 959         return NULL;
 960 }
 961
 962 smb_ucs2_t *strstr_wa(const smb_ucs2_t *s, const char *ins)
 963 {
 964         smb_ucs2_t *r;
 965         size_t inslen;
 966
 967         if (!s || !ins) {
 968                 return NULL;
 969         }
 970
 971         inslen = strlen(ins);
 972         r = (smb_ucs2_t *)s;
 973
 974         while ((r = strchr_w(r, UCS2_CHAR(*ins)))) {
 975                 if (strncmp_wa(r, ins, inslen) == 0)
 976                         return r;
 977                 r++;
 978         }
 979
 980         return NULL;
 981 }
 982
 983 /*******************************************************************
 984  Returns the length in number of wide characters.
 985 ******************************************************************/
 986
 987 int unistrlen(uint16 *s)
 988 {
 989         int len;
 990
 991         if (!s) {
 992                 return -1;
 993         }
 994
 995         for (len=0; SVAL(s,0); s++,len++) {
 996                 ;
 997         }
 998
 999         return len;
1000 }
1001
1002 /*******************************************************************
1003  Strcpy for unicode strings. Returns length (in num of wide chars).
1004  Not odd align safe.
1005 ********************************************************************/
1006
1007 int unistrcpy(uint16 *dst, uint16 *src)
1008 {
1009         int num_wchars = 0;
1010
1011         while (SVAL(src,0)) {
1012                 *dst++ = *src++;
1013                 num_wchars++;
1014         }
1015         *dst = 0;
1016
1017         return num_wchars;
1018 }
1019
1020 /**
1021  * Samba ucs2 type to UNISTR2 conversion
1022  *
1023  * @param ctx Talloc context to create the dst strcture (if null) and the
1024  *            contents of the unicode string.
1025  * @param dst UNISTR2 destination. If equals null, then it's allocated.
1026  * @param src smb_ucs2_t source.
1027  * @param max_len maximum number of unicode characters to copy. If equals
1028  *        null, then null-termination of src is taken
1029  *
1030  * @return copied UNISTR2 destination
1031  **/
1032
1033 UNISTR2* ucs2_to_unistr2(TALLOC_CTX *ctx, UNISTR2* dst, smb_ucs2_t* src)
1034 {
1035         size_t len;
1036
1037         if (!src) {
1038                 return NULL;
1039         }
1040
1041         len = strlen_w(src);
1042
1043         /* allocate UNISTR2 destination if not given */
1044         if (!dst) {
1045                 dst = TALLOC_P(ctx, UNISTR2);
1046                 if (!dst)
1047                         return NULL;
1048         }
1049         if (!dst->buffer) {
1050                 dst->buffer = TALLOC_ARRAY(ctx, uint16, len + 1);
1051                 if (!dst->buffer)
1052                         return NULL;
1053         }
1054
1055         /* set UNISTR2 parameters */
1056         dst->uni_max_len = len + 1;
1057         dst->offset = 0;
1058         dst->uni_str_len = len;
1059
1060         /* copy the actual unicode string */
1061         strncpy_w(dst->buffer, src, dst->uni_max_len);
1062
1063         return dst;
1064 }
1065
1066 /*************************************************************
1067  ascii only toupper - saves the need for smbd to be in C locale.
1068 *************************************************************/
1069
1070 int toupper_ascii(int c)
1071 {
1072         smb_ucs2_t uc = toupper_w(UCS2_CHAR(c));
1073         return UCS2_TO_CHAR(uc);
1074 }
1075
1076 /*************************************************************
1077  ascii only tolower - saves the need for smbd to be in C locale.
1078 *************************************************************/
1079
1080 int tolower_ascii(int c)
1081 {
1082         smb_ucs2_t uc = tolower_w(UCS2_CHAR(c));
1083         return UCS2_TO_CHAR(uc);
1084 }
1085
1086 /*************************************************************
1087  ascii only isupper - saves the need for smbd to be in C locale.
1088 *************************************************************/
1089
1090 int isupper_ascii(int c)
1091 {
1092         return isupper_w(UCS2_CHAR(c));
1093 }
1094
1095 /*************************************************************
1096  ascii only islower - saves the need for smbd to be in C locale.
1097 *************************************************************/
1098
1099 int islower_ascii(int c)
1100 {
1101         return islower_w(UCS2_CHAR(c));
1102 }