s3: Fix bug #9085.
[Samba.git] / source3 / lib / util_unistr.c
blobf53ef94d69eca4b5b8c016b6c9ed75bfbe08457c
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
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.
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.
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/>.
22 #include "includes.h"
24 #ifndef MAXUNI
25 #define MAXUNI 1024
26 #endif
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;
38 /**
39 * Destroy global objects allocated by load_case_tables()
40 **/
41 void gfree_case_tables(void)
43 if ( upcase_table ) {
44 if ( upcase_table_use_unmap )
45 unmap_file(upcase_table, 0x20000);
46 else
47 SAFE_FREE(upcase_table);
50 if ( lowcase_table ) {
51 if ( lowcase_table_use_unmap )
52 unmap_file(lowcase_table, 0x20000);
53 else
54 SAFE_FREE(lowcase_table);
57 if ( valid_table ) {
58 if ( valid_table_use_unmap )
59 unmap_file(valid_table, 0x10000);
60 else
61 SAFE_FREE(valid_table);
63 initialized = false;
66 /**
67 * Load or generate the case handling tables.
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 **/
73 void load_case_tables(void)
75 char *old_locale = NULL, *saved_locale = NULL;
76 int i;
77 TALLOC_CTX *frame = NULL;
79 if (initialized) {
80 return;
82 initialized = true;
84 frame = talloc_stackframe();
86 upcase_table = (smb_ucs2_t *)map_file(data_path("upcase.dat"),
87 0x20000);
88 upcase_table_use_unmap = ( upcase_table != NULL );
90 lowcase_table = (smb_ucs2_t *)map_file(data_path("lowcase.dat"),
91 0x20000);
92 lowcase_table_use_unmap = ( lowcase_table != NULL );
94 #ifdef HAVE_SETLOCALE
95 /* Get the name of the current locale. */
96 old_locale = setlocale(LC_ALL, NULL);
98 if (old_locale) {
99 /* Save it as it is in static storage. */
100 saved_locale = SMB_STRDUP(old_locale);
103 /* We set back the locale to C to get ASCII-compatible toupper/lower functions. */
104 setlocale(LC_ALL, "C");
105 #endif
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;
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);
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;
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);
139 #ifdef HAVE_SETLOCALE
140 /* Restore the old locale. */
141 if (saved_locale) {
142 setlocale (LC_ALL, saved_locale);
143 SAFE_FREE(saved_locale);
145 #endif
146 TALLOC_FREE(frame);
149 static int check_dos_char_slowly(smb_ucs2_t c)
151 char buf[10];
152 smb_ucs2_t c2 = 0;
153 int len1, len2;
155 len1 = convert_string(CH_UTF16LE, CH_DOS, &c, 2, buf, sizeof(buf),False);
156 if (len1 == 0) {
157 return 0;
159 len2 = convert_string(CH_DOS, CH_UTF16LE, buf, len1, &c2, 2,False);
160 if (len2 != 2) {
161 return 0;
163 return (c == c2);
167 * Load the valid character map table from <tt>valid.dat</tt> or
168 * create from the configured codepage.
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.
175 void init_valid_table(void)
177 static int mapped_file;
178 int i;
179 const char *allowed = ".!#$%&'()_-@^`~";
180 uint8 *valid_file;
182 if (mapped_file) {
183 /* Can't unmap files, so stick with what we have */
184 return;
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;
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);
201 /* use free rather than unmap */
202 valid_table_use_unmap = False;
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);
211 lazy_initialize_conv();
213 for (;i<0x10000;i++) {
214 smb_ucs2_t c;
215 SSVAL(&c, 0, i);
216 valid_table[i] = check_dos_char_slowly(c);
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.
225 if null_terminate is True then null terminate the packet (adds 2 bytes)
227 the return value is the length in bytes consumed by the string, including the
228 null termination if applied
229 ********************************************************************/
231 size_t dos_PutUniCode(char *dst,const char *src, size_t len, bool null_terminate)
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);
239 /*******************************************************************
240 Skip past a unicode string, but not more than len. Always move
241 past a terminating zero if found.
242 ********************************************************************/
244 char *skip_unibuf(char *src, size_t len)
246 char *srcend = src + len;
248 while (src < srcend && SVAL(src,0)) {
249 src += 2;
252 if(!SVAL(src,0)) {
253 src += 2;
256 return src;
259 /* Converts a string from internal samba format to unicode
262 int rpcstr_push(void *dest, const char *src, size_t dest_len, int flags)
264 return push_ucs2(NULL, dest, src, dest_len, flags|STR_UNICODE|STR_NOALIGN);
267 /* Converts a string from internal samba format to unicode. Always terminates.
268 * Actually just a wrapper round push_ucs2_talloc().
271 int rpcstr_push_talloc(TALLOC_CTX *ctx, smb_ucs2_t **dest, const char *src)
273 size_t size;
274 if (push_ucs2_talloc(ctx, dest, src, &size))
275 return size;
276 else
277 return -1;
280 /*******************************************************************
281 Convert a wchar to upper case.
282 ********************************************************************/
284 smb_ucs2_t toupper_w(smb_ucs2_t val)
286 return upcase_table[SVAL(&val,0)];
289 /*******************************************************************
290 Convert a wchar to lower case.
291 ********************************************************************/
293 smb_ucs2_t tolower_w( smb_ucs2_t val )
295 return lowcase_table[SVAL(&val,0)];
298 /*******************************************************************
299 Determine if a character is lowercase.
300 ********************************************************************/
302 bool islower_w(smb_ucs2_t c)
304 return upcase_table[SVAL(&c,0)] != c;
307 /*******************************************************************
308 Determine if a character is uppercase.
309 ********************************************************************/
311 bool isupper_w(smb_ucs2_t c)
313 return lowcase_table[SVAL(&c,0)] != c;
316 /*******************************************************************
317 Determine if a character is valid in a 8.3 name.
318 ********************************************************************/
320 bool isvalid83_w(smb_ucs2_t c)
322 return valid_table[SVAL(&c,0)] != 0;
325 /*******************************************************************
326 Count the number of characters in a smb_ucs2_t string.
327 ********************************************************************/
329 size_t strlen_w(const smb_ucs2_t *src)
331 size_t len;
332 smb_ucs2_t c;
334 for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
338 return len;
341 /*******************************************************************
342 Count up to max number of characters in a smb_ucs2_t string.
343 ********************************************************************/
345 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
347 size_t len;
348 smb_ucs2_t c;
350 for(len = 0; (len < max) && *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
354 return len;
357 /*******************************************************************
358 Wide strchr().
359 ********************************************************************/
361 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
363 smb_ucs2_t cp;
364 while (*(COPY_UCS2_CHAR(&cp,s))) {
365 if (c == cp) {
366 return (smb_ucs2_t *)s;
368 s++;
370 if (c == cp) {
371 return (smb_ucs2_t *)s;
374 return NULL;
377 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
379 return strchr_w(s, UCS2_CHAR(c));
382 /*******************************************************************
383 Wide strrchr().
384 ********************************************************************/
386 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
388 smb_ucs2_t cp;
389 const smb_ucs2_t *p = s;
390 int len = strlen_w(s);
392 if (len == 0) {
393 return NULL;
395 p += (len - 1);
396 do {
397 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
398 return (smb_ucs2_t *)p;
400 } while (p-- != s);
401 return NULL;
404 /*******************************************************************
405 Wide version of strrchr that returns after doing strrchr 'n' times.
406 ********************************************************************/
408 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
410 smb_ucs2_t cp;
411 const smb_ucs2_t *p = s;
412 int len = strlen_w(s);
414 if (len == 0 || !n) {
415 return NULL;
417 p += (len - 1);
418 do {
419 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
420 n--;
423 if (!n) {
424 return (smb_ucs2_t *)p;
426 } while (p-- != s);
427 return NULL;
430 /*******************************************************************
431 Wide strstr().
432 ********************************************************************/
434 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
436 smb_ucs2_t *r;
437 size_t inslen;
439 if (!s || !*s || !ins || !*ins) {
440 return NULL;
443 inslen = strlen_w(ins);
444 r = (smb_ucs2_t *)s;
446 while ((r = strchr_w(r, *ins))) {
447 if (strncmp_w(r, ins, inslen) == 0) {
448 return r;
450 r++;
453 return NULL;
456 /*******************************************************************
457 Convert a string to lower case.
458 return True if any char is converted
459 ********************************************************************/
461 bool strlower_w(smb_ucs2_t *s)
463 smb_ucs2_t cp;
464 bool ret = False;
466 while (*(COPY_UCS2_CHAR(&cp,s))) {
467 smb_ucs2_t v = tolower_w(cp);
468 if (v != cp) {
469 COPY_UCS2_CHAR(s,&v);
470 ret = True;
472 s++;
474 return ret;
477 /*******************************************************************
478 Convert a string to upper case.
479 return True if any char is converted
480 ********************************************************************/
482 bool strupper_w(smb_ucs2_t *s)
484 smb_ucs2_t cp;
485 bool ret = False;
486 while (*(COPY_UCS2_CHAR(&cp,s))) {
487 smb_ucs2_t v = toupper_w(cp);
488 if (v != cp) {
489 COPY_UCS2_CHAR(s,&v);
490 ret = True;
492 s++;
494 return ret;
497 /*******************************************************************
498 Convert a string to "normal" form.
499 ********************************************************************/
501 void strnorm_w(smb_ucs2_t *s, int case_default)
503 if (case_default == CASE_UPPER) {
504 strupper_w(s);
505 } else {
506 strlower_w(s);
510 int strcmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
512 smb_ucs2_t cpa, cpb;
514 while ((*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
515 a++;
516 b++;
518 return (*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b)));
519 /* warning: if *a != *b and both are not 0 we return a random
520 greater or lesser than 0 number not realted to which
521 string is longer */
524 int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
526 smb_ucs2_t cpa, cpb;
527 size_t n = 0;
529 while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
530 a++;
531 b++;
532 n++;
534 return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
537 /*******************************************************************
538 Case insensitive string comparison.
539 ********************************************************************/
541 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
543 smb_ucs2_t cpa, cpb;
545 while ((*COPY_UCS2_CHAR(&cpb,b)) && toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb)) {
546 a++;
547 b++;
549 return (tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b))));
552 /*******************************************************************
553 Case insensitive string comparison, length limited.
554 ********************************************************************/
556 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
558 smb_ucs2_t cpa, cpb;
559 size_t n = 0;
561 while ((n < len) && *COPY_UCS2_CHAR(&cpb,b) && (toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb))) {
562 a++;
563 b++;
564 n++;
566 return (len - n)?(tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b)))):0;
569 /*******************************************************************
570 Compare 2 strings.
571 ********************************************************************/
573 bool strequal_w(const smb_ucs2_t *s1, const smb_ucs2_t *s2)
575 if (s1 == s2) {
576 return(True);
578 if (!s1 || !s2) {
579 return(False);
582 return(strcasecmp_w(s1,s2)==0);
585 /*******************************************************************
586 Compare 2 strings up to and including the nth char.
587 ******************************************************************/
589 bool strnequal_w(const smb_ucs2_t *s1,const smb_ucs2_t *s2,size_t n)
591 if (s1 == s2) {
592 return(True);
594 if (!s1 || !s2 || !n) {
595 return(False);
598 return(strncasecmp_w(s1,s2,n)==0);
601 /*******************************************************************
602 Duplicate string.
603 ********************************************************************/
605 smb_ucs2_t *strdup_w(const smb_ucs2_t *src)
607 return strndup_w(src, 0);
610 /* if len == 0 then duplicate the whole string */
612 smb_ucs2_t *strndup_w(const smb_ucs2_t *src, size_t len)
614 smb_ucs2_t *dest;
616 if (!len) {
617 len = strlen_w(src);
619 dest = SMB_MALLOC_ARRAY(smb_ucs2_t, len + 1);
620 if (!dest) {
621 DEBUG(0,("strdup_w: out of memory!\n"));
622 return NULL;
625 memcpy(dest, src, len * sizeof(smb_ucs2_t));
626 dest[len] = 0;
627 return dest;
630 /*******************************************************************
631 Copy a string with max len.
632 ********************************************************************/
634 smb_ucs2_t *strncpy_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
636 smb_ucs2_t cp;
637 size_t len;
639 if (!dest || !src) {
640 return NULL;
643 for (len = 0; (*COPY_UCS2_CHAR(&cp,(src+len))) && (len < max); len++) {
644 cp = *COPY_UCS2_CHAR(dest+len,src+len);
646 cp = 0;
647 for ( /*nothing*/ ; len < max; len++ ) {
648 cp = *COPY_UCS2_CHAR(dest+len,&cp);
651 return dest;
654 /*******************************************************************
655 Append a string of len bytes and add a terminator.
656 ********************************************************************/
658 smb_ucs2_t *strncat_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
660 size_t start;
661 size_t len;
662 smb_ucs2_t z = 0;
664 if (!dest || !src) {
665 return NULL;
668 start = strlen_w(dest);
669 len = strnlen_w(src, max);
671 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
672 z = *COPY_UCS2_CHAR(dest+start+len,&z);
674 return dest;
677 smb_ucs2_t *strcat_w(smb_ucs2_t *dest, const smb_ucs2_t *src)
679 size_t start;
680 size_t len;
681 smb_ucs2_t z = 0;
683 if (!dest || !src) {
684 return NULL;
687 start = strlen_w(dest);
688 len = strlen_w(src);
690 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
691 z = *COPY_UCS2_CHAR(dest+start+len,&z);
693 return dest;
697 /*******************************************************************
698 Replace any occurence of oldc with newc in unicode string.
699 ********************************************************************/
701 void string_replace_w(smb_ucs2_t *s, smb_ucs2_t oldc, smb_ucs2_t newc)
703 smb_ucs2_t cp;
705 for(;*(COPY_UCS2_CHAR(&cp,s));s++) {
706 if(cp==oldc) {
707 COPY_UCS2_CHAR(s,&newc);
712 /*******************************************************************
713 Trim unicode string.
714 ********************************************************************/
716 bool trim_string_w(smb_ucs2_t *s, const smb_ucs2_t *front,
717 const smb_ucs2_t *back)
719 bool ret = False;
720 size_t len, front_len, back_len;
722 if (!s) {
723 return False;
726 len = strlen_w(s);
728 if (front && *front) {
729 front_len = strlen_w(front);
730 while (len && strncmp_w(s, front, front_len) == 0) {
731 memmove(s, (s + front_len), (len - front_len + 1) * sizeof(smb_ucs2_t));
732 len -= front_len;
733 ret = True;
737 if (back && *back) {
738 back_len = strlen_w(back);
739 while (len && strncmp_w((s + (len - back_len)), back, back_len) == 0) {
740 s[len - back_len] = 0;
741 len -= back_len;
742 ret = True;
746 return ret;
750 The *_wa() functions take a combination of 7 bit ascii
751 and wide characters They are used so that you can use string
752 functions combining C string constants with ucs2 strings
754 The char* arguments must NOT be multibyte - to be completely sure
755 of this only pass string constants */
757 int strcmp_wa(const smb_ucs2_t *a, const char *b)
759 smb_ucs2_t cp = 0;
761 while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
762 a++;
763 b++;
765 return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
768 int strncmp_wa(const smb_ucs2_t *a, const char *b, size_t len)
770 smb_ucs2_t cp = 0;
771 size_t n = 0;
773 while ((n < len) && *b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
774 a++;
775 b++;
776 n++;
778 return (len - n)?(*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b)):0;
781 smb_ucs2_t *strpbrk_wa(const smb_ucs2_t *s, const char *p)
783 smb_ucs2_t cp;
785 while (*(COPY_UCS2_CHAR(&cp,s))) {
786 int i;
787 for (i=0; p[i] && cp != UCS2_CHAR(p[i]); i++)
789 if (p[i]) {
790 return (smb_ucs2_t *)s;
792 s++;
794 return NULL;
797 smb_ucs2_t *strstr_wa(const smb_ucs2_t *s, const char *ins)
799 smb_ucs2_t *r;
800 size_t inslen;
802 if (!s || !ins) {
803 return NULL;
806 inslen = strlen(ins);
807 r = (smb_ucs2_t *)s;
809 while ((r = strchr_w(r, UCS2_CHAR(*ins)))) {
810 if (strncmp_wa(r, ins, inslen) == 0)
811 return r;
812 r++;
815 return NULL;
818 /*************************************************************
819 ascii only toupper - saves the need for smbd to be in C locale.
820 *************************************************************/
822 int toupper_ascii(int c)
824 smb_ucs2_t uc = toupper_w(UCS2_CHAR(c));
825 return UCS2_TO_CHAR(uc);
828 /*************************************************************
829 ascii only tolower - saves the need for smbd to be in C locale.
830 *************************************************************/
832 int tolower_ascii(int c)
834 smb_ucs2_t uc = tolower_w(UCS2_CHAR(c));
835 return UCS2_TO_CHAR(uc);
838 /*************************************************************
839 ascii only isupper - saves the need for smbd to be in C locale.
840 *************************************************************/
842 int isupper_ascii(int c)
844 return isupper_w(UCS2_CHAR(c));
847 /*************************************************************
848 ascii only islower - saves the need for smbd to be in C locale.
849 *************************************************************/
851 int islower_ascii(int c)
853 return islower_w(UCS2_CHAR(c));