source3/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
  37 /**
  38  * Destroy global objects allocated by load_case_tables()
  39  **/
  40 void gfree_case_tables(void)
  41 {
  42         if ( upcase_table ) {
  43                 if ( upcase_table_use_unmap )
  44                         unmap_file(upcase_table, 0x20000);
  45                 else
  46                         SAFE_FREE(upcase_table);
  47         }
  48
  49         if ( lowcase_table ) {
  50                 if ( lowcase_table_use_unmap )
  51                         unmap_file(lowcase_table, 0x20000);
  52                 else
  53                         SAFE_FREE(lowcase_table);
  54         }
  55
  56         if ( valid_table ) {
  57                 if ( valid_table_use_unmap )
  58                         unmap_file(valid_table, 0x10000);
  59                 else
  60                         SAFE_FREE(valid_table);
  61         }
  62 }
  63
  64 /**
  65  * Load or generate the case handling tables.
  66  *
  67  * The case tables are defined in UCS2 and don't depend on any
  68  * configured parameters, so they never need to be reloaded.
  69  **/
  70
  71 void load_case_tables(void)
  72 {
  73         static int initialised;
  74         char *old_locale = NULL, *saved_locale = NULL;
  75         int i;
  76         TALLOC_CTX *frame = NULL;
  77
  78         if (initialised) {
  79                 return;
  80         }
  81         initialised = 1;
  82
  83         frame = talloc_stackframe();
  84
  85         upcase_table = (smb_ucs2_t *)map_file(data_path("upcase.dat"),
  86                                               0x20000);
  87         upcase_table_use_unmap = ( upcase_table != NULL );
  88
  89         lowcase_table = (smb_ucs2_t *)map_file(data_path("lowcase.dat"),
  90                                                0x20000);
  91         lowcase_table_use_unmap = ( lowcase_table != NULL );
  92
  93 #ifdef HAVE_SETLOCALE
  94         /* Get the name of the current locale.  */
  95         old_locale = setlocale(LC_ALL, NULL);
  96
  97         if (old_locale) {
  98                 /* Save it as it is in static storage. */
  99                 saved_locale = SMB_STRDUP(old_locale);
 100         }
 101
 102         /* We set back the locale to C to get ASCII-compatible toupper/lower functions. */
 103         setlocale(LC_ALL, "C");
 104 #endif
 105
 106         /* we would like Samba to limp along even if these tables are
 107            not available */
 108         if (!upcase_table) {
 109                 DEBUG(1,("creating lame upcase table\n"));
 110                 upcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
 111                 for (i=0;i<0x10000;i++) {
 112                         smb_ucs2_t v;
 113                         SSVAL(&v, 0, i);
 114                         upcase_table[v] = i;
 115                 }
 116                 for (i=0;i<256;i++) {
 117                         smb_ucs2_t v;
 118                         SSVAL(&v, 0, UCS2_CHAR(i));
 119                         upcase_table[v] = UCS2_CHAR(islower(i)?toupper(i):i);
 120                 }
 121         }
 122
 123         if (!lowcase_table) {
 124                 DEBUG(1,("creating lame lowcase table\n"));
 125                 lowcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
 126                 for (i=0;i<0x10000;i++) {
 127                         smb_ucs2_t v;
 128                         SSVAL(&v, 0, i);
 129                         lowcase_table[v] = i;
 130                 }
 131                 for (i=0;i<256;i++) {
 132                         smb_ucs2_t v;
 133                         SSVAL(&v, 0, UCS2_CHAR(i));
 134                         lowcase_table[v] = UCS2_CHAR(isupper(i)?tolower(i):i);
 135                 }
 136         }
 137
 138 #ifdef HAVE_SETLOCALE
 139         /* Restore the old locale. */
 140         if (saved_locale) {
 141                 setlocale (LC_ALL, saved_locale);
 142                 SAFE_FREE(saved_locale);
 143         }
 144 #endif
 145         TALLOC_FREE(frame);
 146 }
 147
 148 static int check_dos_char_slowly(smb_ucs2_t c)
 149 {
 150         char buf[10];
 151         smb_ucs2_t c2 = 0;
 152         int len1, len2;
 153
 154         len1 = convert_string(CH_UTF16LE, CH_DOS, &c, 2, buf, sizeof(buf),False);
 155         if (len1 == 0) {
 156                 return 0;
 157         }
 158         len2 = convert_string(CH_DOS, CH_UTF16LE, buf, len1, &c2, 2,False);
 159         if (len2 != 2) {
 160                 return 0;
 161         }
 162         return (c == c2);
 163 }
 164
 165 /**
 166  * Load the valid character map table from <tt>valid.dat</tt> or
 167  * create from the configured codepage.
 168  *
 169  * This function is called whenever the configuration is reloaded.
 170  * However, the valid character table is not changed if it's loaded
 171  * from a file, because we can't unmap files.
 172  **/
 173
 174 void init_valid_table(void)
 175 {
 176         static int mapped_file;
 177         int i;
 178         const char *allowed = ".!#$%&'()_-@^`~";
 179         uint8 *valid_file;
 180
 181         if (mapped_file) {
 182                 /* Can't unmap files, so stick with what we have */
 183                 return;
 184         }
 185
 186         valid_file = (uint8 *)map_file(data_path("valid.dat"), 0x10000);
 187         if (valid_file) {
 188                 valid_table = valid_file;
 189                 mapped_file = 1;
 190                 valid_table_use_unmap = True;
 191                 return;
 192         }
 193
 194         /* Otherwise, we're using a dynamically created valid_table.
 195          * It might need to be regenerated if the code page changed.
 196          * We know that we're not using a mapped file, so we can
 197          * free() the old one. */
 198         SAFE_FREE(valid_table);
 199
 200         /* use free rather than unmap */
 201         valid_table_use_unmap = False;
 202
 203         DEBUG(2,("creating default valid table\n"));
 204         valid_table = (uint8 *)SMB_MALLOC(0x10000);
 205         SMB_ASSERT(valid_table != NULL);
 206         for (i=0;i<128;i++) {
 207                 valid_table[i] = isalnum(i) || strchr(allowed,i);
 208         }
 209
 210         lazy_initialize_conv();
 211
 212         for (;i<0x10000;i++) {
 213                 smb_ucs2_t c;
 214                 SSVAL(&c, 0, i);
 215                 valid_table[i] = check_dos_char_slowly(c);
 216         }
 217 }
 218
 219 /*******************************************************************
 220  Write a string in (little-endian) unicode format. src is in
 221  the current DOS codepage. len is the length in bytes of the
 222  string pointed to by dst.
 223
 224  if null_terminate is True then null terminate the packet (adds 2 bytes)
 225
 226  the return value is the length in bytes consumed by the string, including the
 227  null termination if applied
 228 ********************************************************************/
 229
 230 size_t dos_PutUniCode(char *dst,const char *src, size_t len, bool null_terminate)
 231 {
 232         int flags = null_terminate ? STR_UNICODE|STR_NOALIGN|STR_TERMINATE
 233                                    : STR_UNICODE|STR_NOALIGN;
 234         return push_ucs2(NULL, dst, src, len, flags);
 235 }
 236
 237
 238 /*******************************************************************
 239  Skip past a unicode string, but not more than len. Always move
 240  past a terminating zero if found.
 241 ********************************************************************/
 242
 243 char *skip_unibuf(char *src, size_t len)
 244 {
 245         char *srcend = src + len;
 246
 247         while (src < srcend && SVAL(src,0)) {
 248                 src += 2;
 249         }
 250
 251         if(!SVAL(src,0)) {
 252                 src += 2;
 253         }
 254
 255         return src;
 256 }
 257
 258 /* Copy a string from little-endian or big-endian unicode source (depending
 259  * on flags) to internal samba format destination
 260  */
 261
 262 int rpcstr_pull(char* dest, void *src, int dest_len, int src_len, int flags)
 263 {
 264         if (!src) {
 265                 dest[0] = 0;
 266                 return 0;
 267         }
 268         if(dest_len==-1) {
 269                 dest_len=MAXUNI-3;
 270         }
 271         return pull_ucs2(NULL, dest, src, dest_len, src_len, flags|STR_UNICODE|STR_NOALIGN);
 272 }
 273
 274 /* Copy a string from little-endian or big-endian unicode source (depending
 275  * on flags) to internal samba format destination. Allocates on talloc ctx.
 276  */
 277
 278 int rpcstr_pull_talloc(TALLOC_CTX *ctx,
 279                         char **dest,
 280                         void *src,
 281                         int src_len,
 282                         int flags)
 283 {
 284         return pull_ucs2_base_talloc(ctx,
 285                         NULL,
 286                         dest,
 287                         src,
 288                         src_len,
 289                         flags|STR_UNICODE|STR_NOALIGN);
 290
 291 }
 292
 293 /* Copy a string from a unistr2 source to internal samba format
 294    destination.  Use this instead of direct calls to rpcstr_pull() to avoid
 295    having to determine whether the source string is null terminated. */
 296
 297 int rpcstr_pull_unistr2_fstring(char *dest, UNISTR2 *src)
 298 {
 299         return pull_ucs2(NULL, dest, src->buffer, sizeof(fstring),
 300                          src->uni_str_len * 2, 0);
 301 }
 302
 303 /* Helper function to return a talloc'ed string. I have implemented it with a
 304  * copy because I don't really know how pull_ucs2 and friends calculate the
 305  * target size. If this turns out to be a major bottleneck someone with deeper
 306  * multi-byte knowledge needs to revisit this.
 307  * I just did (JRA :-). No longer uses copy.
 308  * My (VL) use is dsr_getdcname, which returns 6 strings, the alternative would
 309  * have been to manually talloc_strdup them in rpc_client/cli_netlogon.c.
 310  */
 311
 312 char *rpcstr_pull_unistr2_talloc(TALLOC_CTX *ctx, const UNISTR2 *src)
 313 {
 314         char *dest = NULL;
 315         size_t dest_len;
 316
 317         if (!convert_string_talloc(ctx, CH_UTF16LE, CH_UNIX, src->buffer,
 318                                    src->uni_str_len * 2, (void *)&dest,
 319                                    &dest_len, true))
 320         {
 321                 return NULL;
 322         }
 323
 324         /* Ensure we're returning a null terminated string. */
 325         if (dest_len) {
 326                 /* Did we already process the terminating zero ? */
 327                 if (dest[dest_len-1] != 0) {
 328                         size_t size = talloc_get_size(dest);
 329                         /* Have we got space to append the '\0' ? */
 330                         if (size <= dest_len) {
 331                                 /* No, realloc. */
 332                                 dest = TALLOC_REALLOC_ARRAY(ctx, dest, char,
 333                                                 dest_len+1);
 334                                 if (!dest) {
 335                                         /* talloc fail. */
 336                                         dest_len = (size_t)-1;
 337                                         return NULL;
 338                                 }
 339                         }
 340                         /* Yay - space ! */
 341                         dest[dest_len] = '\0';
 342                         dest_len++;
 343                 }
 344         } else if (dest) {
 345                 dest[0] = 0;
 346         }
 347
 348         return dest;
 349 }
 350
 351 /* Converts a string from internal samba format to unicode
 352  */
 353
 354 int rpcstr_push(void *dest, const char *src, size_t dest_len, int flags)
 355 {
 356         return push_ucs2(NULL, dest, src, dest_len, flags|STR_UNICODE|STR_NOALIGN);
 357 }
 358
 359 /* Converts a string from internal samba format to unicode. Always terminates.
 360  * Actually just a wrapper round push_ucs2_talloc().
 361  */
 362
 363 int rpcstr_push_talloc(TALLOC_CTX *ctx, smb_ucs2_t **dest, const char *src)
 364 {
 365         size_t size;
 366         if (push_ucs2_talloc(ctx, dest, src, &size))
 367                 return size;
 368         else
 369                 return -1;
 370 }
 371
 372 /*******************************************************************
 373  Convert a (little-endian) UNISTR2 structure to an ASCII string.
 374 ********************************************************************/
 375
 376 void unistr2_to_ascii(char *dest, const UNISTR2 *str, size_t maxlen)
 377 {
 378         if ((str == NULL) || (str->uni_str_len == 0)) {
 379                 *dest='\0';
 380                 return;
 381         }
 382         pull_ucs2(NULL, dest, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
 383 }
 384
 385 #if 0
 386 /*******************************************************************
 387  Convert a (little-endian) UNISTR3 structure to an ASCII string.
 388 ********************************************************************/
 389
 390 void unistr3_to_ascii(char *dest, const UNISTR3 *str, size_t maxlen)
 391 {
 392         if ((str == NULL) || (str->uni_str_len == 0)) {
 393                 *dest='\0';
 394                 return;
 395         }
 396         pull_ucs2(NULL, dest, str->str.buffer, maxlen, str->uni_str_len*2,
 397                   STR_NOALIGN);
 398 }
 399 #endif
 400
 401 /*******************************************************************
 402  Duplicate a UNISTR2 string into a null terminated char*
 403  using a talloc context.
 404 ********************************************************************/
 405
 406 char *unistr2_to_ascii_talloc(TALLOC_CTX *ctx, const UNISTR2 *str)
 407 {
 408         char *s = NULL;
 409
 410         if (!str || !str->buffer) {
 411                 return NULL;
 412         }
 413         if (pull_ucs2_base_talloc(ctx,
 414                                 NULL,
 415                                 &s,
 416                                 str->buffer,
 417                                 str->uni_str_len*2,
 418                                 STR_NOALIGN) == (size_t)-1) {
 419                 return NULL;
 420         }
 421         return s;
 422 }
 423
 424 /*******************************************************************
 425  Return a string for displaying a UNISTR2. Guarentees to return a
 426  valid string - "" if nothing else.
 427  Changed to use talloc_tos() under the covers.... JRA.
 428 ********************************************************************/
 429
 430 const char *unistr2_static(const UNISTR2 *str)
 431 {
 432         char *dest = NULL;
 433
 434         if ((str == NULL) || (str->uni_str_len == 0)) {
 435                 return "";
 436         }
 437
 438         dest = unistr2_to_ascii_talloc(talloc_tos(), str);
 439         if (!dest) {
 440                 return "";
 441         }
 442
 443         return dest;
 444 }
 445
 446 /*******************************************************************
 447  Convert a wchar to upper case.
 448 ********************************************************************/
 449
 450 smb_ucs2_t toupper_w(smb_ucs2_t val)
 451 {
 452         return upcase_table[SVAL(&val,0)];
 453 }
 454
 455 /*******************************************************************
 456  Convert a wchar to lower case.
 457 ********************************************************************/
 458
 459 smb_ucs2_t tolower_w( smb_ucs2_t val )
 460 {
 461         return lowcase_table[SVAL(&val,0)];
 462 }
 463
 464 /*******************************************************************
 465  Determine if a character is lowercase.
 466 ********************************************************************/
 467
 468 bool islower_w(smb_ucs2_t c)
 469 {
 470         return upcase_table[SVAL(&c,0)] != c;
 471 }
 472
 473 /*******************************************************************
 474  Determine if a character is uppercase.
 475 ********************************************************************/
 476
 477 bool isupper_w(smb_ucs2_t c)
 478 {
 479         return lowcase_table[SVAL(&c,0)] != c;
 480 }
 481
 482 /*******************************************************************
 483  Determine if a character is valid in a 8.3 name.
 484 ********************************************************************/
 485
 486 bool isvalid83_w(smb_ucs2_t c)
 487 {
 488         return valid_table[SVAL(&c,0)] != 0;
 489 }
 490
 491 /*******************************************************************
 492  Count the number of characters in a smb_ucs2_t string.
 493 ********************************************************************/
 494
 495 size_t strlen_w(const smb_ucs2_t *src)
 496 {
 497         size_t len;
 498         smb_ucs2_t c;
 499
 500         for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
 501                 ;
 502         }
 503
 504         return len;
 505 }
 506
 507 /*******************************************************************
 508  Count up to max number of characters in a smb_ucs2_t string.
 509 ********************************************************************/
 510
 511 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
 512 {
 513         size_t len;
 514         smb_ucs2_t c;
 515
 516         for(len = 0; (len < max) && *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
 517                 ;
 518         }
 519
 520         return len;
 521 }
 522
 523 /*******************************************************************
 524  Wide strchr().
 525 ********************************************************************/
 526
 527 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
 528 {
 529         smb_ucs2_t cp;
 530         while (*(COPY_UCS2_CHAR(&cp,s))) {
 531                 if (c == cp) {
 532                         return (smb_ucs2_t *)s;
 533                 }
 534                 s++;
 535         }
 536         if (c == cp) {
 537                 return (smb_ucs2_t *)s;
 538         }
 539
 540         return NULL;
 541 }
 542
 543 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
 544 {
 545         return strchr_w(s, UCS2_CHAR(c));
 546 }
 547
 548 /*******************************************************************
 549  Wide strrchr().
 550 ********************************************************************/
 551
 552 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
 553 {
 554         smb_ucs2_t cp;
 555         const smb_ucs2_t *p = s;
 556         int len = strlen_w(s);
 557
 558         if (len == 0) {
 559                 return NULL;
 560         }
 561         p += (len - 1);
 562         do {
 563                 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
 564                         return (smb_ucs2_t *)p;
 565                 }
 566         } while (p-- != s);
 567         return NULL;
 568 }
 569
 570 /*******************************************************************
 571  Wide version of strrchr that returns after doing strrchr 'n' times.
 572 ********************************************************************/
 573
 574 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
 575 {
 576         smb_ucs2_t cp;
 577         const smb_ucs2_t *p = s;
 578         int len = strlen_w(s);
 579
 580         if (len == 0 || !n) {
 581                 return NULL;
 582         }
 583         p += (len - 1);
 584         do {
 585                 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
 586                         n--;
 587                 }
 588
 589                 if (!n) {
 590                         return (smb_ucs2_t *)p;
 591                 }
 592         } while (p-- != s);
 593         return NULL;
 594 }
 595
 596 /*******************************************************************
 597  Wide strstr().
 598 ********************************************************************/
 599
 600 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
 601 {
 602         smb_ucs2_t *r;
 603         size_t inslen;
 604
 605         if (!s || !*s || !ins || !*ins) {
 606                 return NULL;
 607         }
 608
 609         inslen = strlen_w(ins);
 610         r = (smb_ucs2_t *)s;
 611
 612         while ((r = strchr_w(r, *ins))) {
 613                 if (strncmp_w(r, ins, inslen) == 0) {
 614                         return r;
 615                 }
 616                 r++;
 617         }
 618
 619         return NULL;
 620 }
 621
 622 /*******************************************************************
 623  Convert a string to lower case.
 624  return True if any char is converted
 625 ********************************************************************/
 626
 627 bool strlower_w(smb_ucs2_t *s)
 628 {
 629         smb_ucs2_t cp;
 630         bool ret = False;
 631
 632         while (*(COPY_UCS2_CHAR(&cp,s))) {
 633                 smb_ucs2_t v = tolower_w(cp);
 634                 if (v != cp) {
 635                         COPY_UCS2_CHAR(s,&v);
 636                         ret = True;
 637                 }
 638                 s++;
 639         }
 640         return ret;
 641 }
 642
 643 /*******************************************************************
 644  Convert a string to upper case.
 645  return True if any char is converted
 646 ********************************************************************/
 647
 648 bool strupper_w(smb_ucs2_t *s)
 649 {
 650         smb_ucs2_t cp;
 651         bool ret = False;
 652         while (*(COPY_UCS2_CHAR(&cp,s))) {
 653                 smb_ucs2_t v = toupper_w(cp);
 654                 if (v != cp) {
 655                         COPY_UCS2_CHAR(s,&v);
 656                         ret = True;
 657                 }
 658                 s++;
 659         }
 660         return ret;
 661 }
 662
 663 /*******************************************************************
 664  Convert a string to "normal" form.
 665 ********************************************************************/
 666
 667 void strnorm_w(smb_ucs2_t *s, int case_default)
 668 {
 669         if (case_default == CASE_UPPER) {
 670                 strupper_w(s);
 671         } else {
 672                 strlower_w(s);
 673         }
 674 }
 675
 676 int strcmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
 677 {
 678         smb_ucs2_t cpa, cpb;
 679
 680         while ((*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
 681                 a++;
 682                 b++;
 683         }
 684         return (*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b)));
 685         /* warning: if *a != *b and both are not 0 we return a random
 686                 greater or lesser than 0 number not realted to which
 687                 string is longer */
 688 }
 689
 690 int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
 691 {
 692         smb_ucs2_t cpa, cpb;
 693         size_t n = 0;
 694
 695         while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
 696                 a++;
 697                 b++;
 698                 n++;
 699         }
 700         return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
 701 }
 702
 703 /*******************************************************************
 704  Case insensitive string comparison.
 705 ********************************************************************/
 706
 707 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
 708 {
 709         smb_ucs2_t cpa, cpb;
 710
 711         while ((*COPY_UCS2_CHAR(&cpb,b)) && toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb)) {
 712                 a++;
 713                 b++;
 714         }
 715         return (tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b))));
 716 }
 717
 718 /*******************************************************************
 719  Case insensitive string comparison, length limited.
 720 ********************************************************************/
 721
 722 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
 723 {
 724         smb_ucs2_t cpa, cpb;
 725         size_t n = 0;
 726
 727         while ((n < len) && *COPY_UCS2_CHAR(&cpb,b) && (toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb))) {
 728                 a++;
 729                 b++;
 730                 n++;
 731         }
 732         return (len - n)?(tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b)))):0;
 733 }
 734
 735 /*******************************************************************
 736  Compare 2 strings.
 737 ********************************************************************/
 738
 739 bool strequal_w(const smb_ucs2_t *s1, const smb_ucs2_t *s2)
 740 {
 741         if (s1 == s2) {
 742                 return(True);
 743         }
 744         if (!s1 || !s2) {
 745                 return(False);
 746         }
 747
 748         return(strcasecmp_w(s1,s2)==0);
 749 }
 750
 751 /*******************************************************************
 752  Compare 2 strings up to and including the nth char.
 753 ******************************************************************/
 754
 755 bool strnequal_w(const smb_ucs2_t *s1,const smb_ucs2_t *s2,size_t n)
 756 {
 757         if (s1 == s2) {
 758                 return(True);
 759         }
 760         if (!s1 || !s2 || !n) {
 761                 return(False);
 762         }
 763
 764         return(strncasecmp_w(s1,s2,n)==0);
 765 }
 766
 767 /*******************************************************************
 768  Duplicate string.
 769 ********************************************************************/
 770
 771 smb_ucs2_t *strdup_w(const smb_ucs2_t *src)
 772 {
 773         return strndup_w(src, 0);
 774 }
 775
 776 /* if len == 0 then duplicate the whole string */
 777
 778 smb_ucs2_t *strndup_w(const smb_ucs2_t *src, size_t len)
 779 {
 780         smb_ucs2_t *dest;
 781
 782         if (!len) {
 783                 len = strlen_w(src);
 784         }
 785         dest = SMB_MALLOC_ARRAY(smb_ucs2_t, len + 1);
 786         if (!dest) {
 787                 DEBUG(0,("strdup_w: out of memory!\n"));
 788                 return NULL;
 789         }
 790
 791         memcpy(dest, src, len * sizeof(smb_ucs2_t));
 792         dest[len] = 0;
 793         return dest;
 794 }
 795
 796 /*******************************************************************
 797  Copy a string with max len.
 798 ********************************************************************/
 799
 800 smb_ucs2_t *strncpy_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
 801 {
 802         smb_ucs2_t cp;
 803         size_t len;
 804
 805         if (!dest || !src) {
 806                 return NULL;
 807         }
 808
 809         for (len = 0; (*COPY_UCS2_CHAR(&cp,(src+len))) && (len < max); len++) {
 810                 cp = *COPY_UCS2_CHAR(dest+len,src+len);
 811         }
 812         cp = 0;
 813         for ( /*nothing*/ ; len < max; len++ ) {
 814                 cp = *COPY_UCS2_CHAR(dest+len,&cp);
 815         }
 816
 817         return dest;
 818 }
 819
 820 /*******************************************************************
 821  Append a string of len bytes and add a terminator.
 822 ********************************************************************/
 823
 824 smb_ucs2_t *strncat_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
 825 {
 826         size_t start;
 827         size_t len;
 828         smb_ucs2_t z = 0;
 829
 830         if (!dest || !src) {
 831                 return NULL;
 832         }
 833
 834         start = strlen_w(dest);
 835         len = strnlen_w(src, max);
 836
 837         memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
 838         z = *COPY_UCS2_CHAR(dest+start+len,&z);
 839
 840         return dest;
 841 }
 842
 843 smb_ucs2_t *strcat_w(smb_ucs2_t *dest, const smb_ucs2_t *src)
 844 {
 845         size_t start;
 846         size_t len;
 847         smb_ucs2_t z = 0;
 848
 849         if (!dest || !src) {
 850                 return NULL;
 851         }
 852
 853         start = strlen_w(dest);
 854         len = strlen_w(src);
 855
 856         memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
 857         z = *COPY_UCS2_CHAR(dest+start+len,&z);
 858
 859         return dest;
 860 }
 861
 862
 863 /*******************************************************************
 864  Replace any occurence of oldc with newc in unicode string.
 865 ********************************************************************/
 866
 867 void string_replace_w(smb_ucs2_t *s, smb_ucs2_t oldc, smb_ucs2_t newc)
 868 {
 869         smb_ucs2_t cp;
 870
 871         for(;*(COPY_UCS2_CHAR(&cp,s));s++) {
 872                 if(cp==oldc) {
 873                         COPY_UCS2_CHAR(s,&newc);
 874                 }
 875         }
 876 }
 877
 878 /*******************************************************************
 879  Trim unicode string.
 880 ********************************************************************/
 881
 882 bool trim_string_w(smb_ucs2_t *s, const smb_ucs2_t *front,
 883                                   const smb_ucs2_t *back)
 884 {
 885         bool ret = False;
 886         size_t len, front_len, back_len;
 887
 888         if (!s) {
 889                 return False;
 890         }
 891
 892         len = strlen_w(s);
 893
 894         if (front && *front) {
 895                 front_len = strlen_w(front);
 896                 while (len && strncmp_w(s, front, front_len) == 0) {
 897                         memmove(s, (s + front_len), (len - front_len + 1) * sizeof(smb_ucs2_t));
 898                         len -= front_len;
 899                         ret = True;
 900                 }
 901         }
 902
 903         if (back && *back) {
 904                 back_len = strlen_w(back);
 905                 while (len && strncmp_w((s + (len - back_len)), back, back_len) == 0) {
 906                         s[len - back_len] = 0;
 907                         len -= back_len;
 908                         ret = True;
 909                 }
 910         }
 911
 912         return ret;
 913 }
 914
 915 /*
 916   The *_wa() functions take a combination of 7 bit ascii
 917   and wide characters They are used so that you can use string
 918   functions combining C string constants with ucs2 strings
 919
 920   The char* arguments must NOT be multibyte - to be completely sure
 921   of this only pass string constants */
 922
 923 int strcmp_wa(const smb_ucs2_t *a, const char *b)
 924 {
 925         smb_ucs2_t cp = 0;
 926
 927         while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
 928                 a++;
 929                 b++;
 930         }
 931         return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
 932 }
 933
 934 int strncmp_wa(const smb_ucs2_t *a, const char *b, size_t len)
 935 {
 936         smb_ucs2_t cp = 0;
 937         size_t n = 0;
 938
 939         while ((n < len) && *b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
 940                 a++;
 941                 b++;
 942                 n++;
 943         }
 944         return (len - n)?(*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b)):0;
 945 }
 946
 947 smb_ucs2_t *strpbrk_wa(const smb_ucs2_t *s, const char *p)
 948 {
 949         smb_ucs2_t cp;
 950
 951         while (*(COPY_UCS2_CHAR(&cp,s))) {
 952                 int i;
 953                 for (i=0; p[i] && cp != UCS2_CHAR(p[i]); i++)
 954                         ;
 955                 if (p[i]) {
 956                         return (smb_ucs2_t *)s;
 957                 }
 958                 s++;
 959         }
 960         return NULL;
 961 }
 962
 963 smb_ucs2_t *strstr_wa(const smb_ucs2_t *s, const char *ins)
 964 {
 965         smb_ucs2_t *r;
 966         size_t inslen;
 967
 968         if (!s || !ins) {
 969                 return NULL;
 970         }
 971
 972         inslen = strlen(ins);
 973         r = (smb_ucs2_t *)s;
 974
 975         while ((r = strchr_w(r, UCS2_CHAR(*ins)))) {
 976                 if (strncmp_wa(r, ins, inslen) == 0)
 977                         return r;
 978                 r++;
 979         }
 980
 981         return NULL;
 982 }
 983
 984 /*******************************************************************
 985  Returns the length in number of wide characters.
 986 ******************************************************************/
 987
 988 int unistrlen(uint16 *s)
 989 {
 990         int len;
 991
 992         if (!s) {
 993                 return -1;
 994         }
 995
 996         for (len=0; SVAL(s,0); s++,len++) {
 997                 ;
 998         }
 999
1000         return len;
1001 }
1002
1003 /*******************************************************************
1004  Strcpy for unicode strings. Returns length (in num of wide chars).
1005  Not odd align safe.
1006 ********************************************************************/
1007
1008 int unistrcpy(uint16 *dst, uint16 *src)
1009 {
1010         int num_wchars = 0;
1011
1012         while (SVAL(src,0)) {
1013                 *dst++ = *src++;
1014                 num_wchars++;
1015         }
1016         *dst = 0;
1017
1018         return num_wchars;
1019 }
1020
1021 /**
1022  * Samba ucs2 type to UNISTR2 conversion
1023  *
1024  * @param ctx Talloc context to create the dst strcture (if null) and the
1025  *            contents of the unicode string.
1026  * @param dst UNISTR2 destination. If equals null, then it's allocated.
1027  * @param src smb_ucs2_t source.
1028  * @param max_len maximum number of unicode characters to copy. If equals
1029  *        null, then null-termination of src is taken
1030  *
1031  * @return copied UNISTR2 destination
1032  **/
1033
1034 UNISTR2* ucs2_to_unistr2(TALLOC_CTX *ctx, UNISTR2* dst, smb_ucs2_t* src)
1035 {
1036         size_t len;
1037
1038         if (!src) {
1039                 return NULL;
1040         }
1041
1042         len = strlen_w(src);
1043
1044         /* allocate UNISTR2 destination if not given */
1045         if (!dst) {
1046                 dst = TALLOC_P(ctx, UNISTR2);
1047                 if (!dst)
1048                         return NULL;
1049         }
1050         if (!dst->buffer) {
1051                 dst->buffer = TALLOC_ARRAY(ctx, uint16, len + 1);
1052                 if (!dst->buffer)
1053                         return NULL;
1054         }
1055
1056         /* set UNISTR2 parameters */
1057         dst->uni_max_len = len + 1;
1058         dst->offset = 0;
1059         dst->uni_str_len = len;
1060
1061         /* copy the actual unicode string */
1062         strncpy_w(dst->buffer, src, dst->uni_max_len);
1063
1064         return dst;
1065 }
1066
1067 /*************************************************************
1068  ascii only toupper - saves the need for smbd to be in C locale.
1069 *************************************************************/
1070
1071 int toupper_ascii(int c)
1072 {
1073         smb_ucs2_t uc = toupper_w(UCS2_CHAR(c));
1074         return UCS2_TO_CHAR(uc);
1075 }
1076
1077 /*************************************************************
1078  ascii only tolower - saves the need for smbd to be in C locale.
1079 *************************************************************/
1080
1081 int tolower_ascii(int c)
1082 {
1083         smb_ucs2_t uc = tolower_w(UCS2_CHAR(c));
1084         return UCS2_TO_CHAR(uc);
1085 }
1086
1087 /*************************************************************
1088  ascii only isupper - saves the need for smbd to be in C locale.
1089 *************************************************************/
1090
1091 int isupper_ascii(int c)
1092 {
1093         return isupper_w(UCS2_CHAR(c));
1094 }
1095
1096 /*************************************************************
1097  ascii only islower - saves the need for smbd to be in C locale.
1098 *************************************************************/
1099
1100 int islower_ascii(int c)
1101 {
1102         return islower_w(UCS2_CHAR(c));
1103 }