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/>.
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;
39 * Destroy global objects allocated by load_case_tables()
41 void gfree_case_tables(void)
44 if ( upcase_table_use_unmap )
45 unmap_file(upcase_table, 0x20000);
47 SAFE_FREE(upcase_table);
50 if ( lowcase_table ) {
51 if ( lowcase_table_use_unmap )
52 unmap_file(lowcase_table, 0x20000);
54 SAFE_FREE(lowcase_table);
58 if ( valid_table_use_unmap )
59 unmap_file(valid_table, 0x10000);
61 SAFE_FREE(valid_table);
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.
73 void load_case_tables(void)
75 char *old_locale = NULL, *saved_locale = NULL;
77 TALLOC_CTX *frame = NULL;
84 frame = talloc_stackframe();
86 upcase_table = (smb_ucs2_t *)map_file(data_path("upcase.dat"),
88 upcase_table_use_unmap = ( upcase_table != NULL );
90 lowcase_table = (smb_ucs2_t *)map_file(data_path("lowcase.dat"),
92 lowcase_table_use_unmap = ( lowcase_table != NULL );
95 /* Get the name of the current locale. */
96 old_locale = setlocale(LC_ALL, NULL);
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");
107 /* we would like Samba to limp along even if these tables are
110 DEBUG(1,("creating lame upcase table\n"));
111 upcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
112 for (i=0;i<0x10000;i++) {
117 for (i=0;i<256;i++) {
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++) {
130 lowcase_table[v] = i;
132 for (i=0;i<256;i++) {
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. */
142 setlocale (LC_ALL, saved_locale);
143 SAFE_FREE(saved_locale);
149 static int check_dos_char_slowly(smb_ucs2_t c)
155 len1 = convert_string(CH_UTF16LE, CH_DOS, &c, 2, buf, sizeof(buf),False);
159 len2 = convert_string(CH_DOS, CH_UTF16LE, buf, len1, &c2, 2,False);
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;
179 const char *allowed = ".!#$%&'()_-@^`~";
183 /* Can't unmap files, so stick with what we have */
187 valid_file = (uint8 *)map_file(data_path("valid.dat"), 0x10000);
189 valid_table = valid_file;
191 valid_table_use_unmap = True;
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++) {
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)) {
259 /* Copy a string from little-endian or big-endian unicode source (depending
260 * on flags) to internal samba format destination
263 int rpcstr_pull(char* dest, void *src, int dest_len, int src_len, int flags)
272 return pull_ucs2(NULL, dest, src, dest_len, src_len, flags|STR_UNICODE|STR_NOALIGN);
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.
279 int rpcstr_pull_talloc(TALLOC_CTX *ctx,
285 return pull_ucs2_base_talloc(ctx,
290 flags|STR_UNICODE|STR_NOALIGN);
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. */
298 int rpcstr_pull_unistr2_fstring(char *dest, UNISTR2 *src)
300 return pull_ucs2(NULL, dest, src->buffer, sizeof(fstring),
301 src->uni_str_len * 2, 0);
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.
313 char *rpcstr_pull_unistr2_talloc(TALLOC_CTX *ctx, const UNISTR2 *src)
318 if (!convert_string_talloc(ctx, CH_UTF16LE, CH_UNIX, src->buffer,
319 src->uni_str_len * 2, (void *)&dest,
325 /* Ensure we're returning a null terminated string. */
327 /* Did we already process the terminating zero ? */
328 if (dest[dest_len-1] != 0) {
329 size_t size = talloc_get_size(dest);
330 /* Have we got space to append the '\0' ? */
331 if (size <= dest_len) {
333 dest = TALLOC_REALLOC_ARRAY(ctx, dest, char,
337 dest_len = (size_t)-1;
342 dest[dest_len] = '\0';
352 /* Converts a string from internal samba format to unicode
355 int rpcstr_push(void *dest, const char *src, size_t dest_len, int flags)
357 return push_ucs2(NULL, dest, src, dest_len, flags|STR_UNICODE|STR_NOALIGN);
360 /* Converts a string from internal samba format to unicode. Always terminates.
361 * Actually just a wrapper round push_ucs2_talloc().
364 int rpcstr_push_talloc(TALLOC_CTX *ctx, smb_ucs2_t **dest, const char *src)
367 if (push_ucs2_talloc(ctx, dest, src, &size))
373 /*******************************************************************
374 Convert a (little-endian) UNISTR2 structure to an ASCII string.
375 ********************************************************************/
377 void unistr2_to_ascii(char *dest, const UNISTR2 *str, size_t maxlen)
379 if ((str == NULL) || (str->uni_str_len == 0)) {
383 pull_ucs2(NULL, dest, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
386 /*******************************************************************
387 Duplicate a UNISTR2 string into a null terminated char*
388 using a talloc context.
389 ********************************************************************/
391 char *unistr2_to_ascii_talloc(TALLOC_CTX *ctx, const UNISTR2 *str)
395 if (!str || !str->buffer) {
398 if (pull_ucs2_base_talloc(ctx,
403 STR_NOALIGN) == (size_t)-1) {
409 /*******************************************************************
410 Return a string for displaying a UNISTR2. Guarentees to return a
411 valid string - "" if nothing else.
412 Changed to use talloc_tos() under the covers.... JRA.
413 ********************************************************************/
415 const char *unistr2_static(const UNISTR2 *str)
419 if ((str == NULL) || (str->uni_str_len == 0)) {
423 dest = unistr2_to_ascii_talloc(talloc_tos(), str);
431 /*******************************************************************
432 Convert a wchar to upper case.
433 ********************************************************************/
435 smb_ucs2_t toupper_w(smb_ucs2_t val)
437 return upcase_table[SVAL(&val,0)];
440 /*******************************************************************
441 Convert a wchar to lower case.
442 ********************************************************************/
444 smb_ucs2_t tolower_w( smb_ucs2_t val )
446 return lowcase_table[SVAL(&val,0)];
449 /*******************************************************************
450 Determine if a character is lowercase.
451 ********************************************************************/
453 bool islower_w(smb_ucs2_t c)
455 return upcase_table[SVAL(&c,0)] != c;
458 /*******************************************************************
459 Determine if a character is uppercase.
460 ********************************************************************/
462 bool isupper_w(smb_ucs2_t c)
464 return lowcase_table[SVAL(&c,0)] != c;
467 /*******************************************************************
468 Determine if a character is valid in a 8.3 name.
469 ********************************************************************/
471 bool isvalid83_w(smb_ucs2_t c)
473 return valid_table[SVAL(&c,0)] != 0;
476 /*******************************************************************
477 Count the number of characters in a smb_ucs2_t string.
478 ********************************************************************/
480 size_t strlen_w(const smb_ucs2_t *src)
485 for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
492 /*******************************************************************
493 Count up to max number of characters in a smb_ucs2_t string.
494 ********************************************************************/
496 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
501 for(len = 0; (len < max) && *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
508 /*******************************************************************
510 ********************************************************************/
512 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
515 while (*(COPY_UCS2_CHAR(&cp,s))) {
517 return (smb_ucs2_t *)s;
522 return (smb_ucs2_t *)s;
528 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
530 return strchr_w(s, UCS2_CHAR(c));
533 /*******************************************************************
535 ********************************************************************/
537 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
540 const smb_ucs2_t *p = s;
541 int len = strlen_w(s);
548 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
549 return (smb_ucs2_t *)p;
555 /*******************************************************************
556 Wide version of strrchr that returns after doing strrchr 'n' times.
557 ********************************************************************/
559 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
562 const smb_ucs2_t *p = s;
563 int len = strlen_w(s);
565 if (len == 0 || !n) {
570 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
575 return (smb_ucs2_t *)p;
581 /*******************************************************************
583 ********************************************************************/
585 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
590 if (!s || !*s || !ins || !*ins) {
594 inslen = strlen_w(ins);
597 while ((r = strchr_w(r, *ins))) {
598 if (strncmp_w(r, ins, inslen) == 0) {
607 /*******************************************************************
608 Convert a string to lower case.
609 return True if any char is converted
610 ********************************************************************/
612 bool strlower_w(smb_ucs2_t *s)
617 while (*(COPY_UCS2_CHAR(&cp,s))) {
618 smb_ucs2_t v = tolower_w(cp);
620 COPY_UCS2_CHAR(s,&v);
628 /*******************************************************************
629 Convert a string to upper case.
630 return True if any char is converted
631 ********************************************************************/
633 bool strupper_w(smb_ucs2_t *s)
637 while (*(COPY_UCS2_CHAR(&cp,s))) {
638 smb_ucs2_t v = toupper_w(cp);
640 COPY_UCS2_CHAR(s,&v);
648 /*******************************************************************
649 Convert a string to "normal" form.
650 ********************************************************************/
652 void strnorm_w(smb_ucs2_t *s, int case_default)
654 if (case_default == CASE_UPPER) {
661 int strcmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
665 while ((*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
669 return (*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b)));
670 /* warning: if *a != *b and both are not 0 we return a random
671 greater or lesser than 0 number not realted to which
675 int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
680 while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
685 return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
688 /*******************************************************************
689 Case insensitive string comparison.
690 ********************************************************************/
692 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
696 while ((*COPY_UCS2_CHAR(&cpb,b)) && toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb)) {
700 return (tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b))));
703 /*******************************************************************
704 Case insensitive string comparison, length limited.
705 ********************************************************************/
707 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
712 while ((n < len) && *COPY_UCS2_CHAR(&cpb,b) && (toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb))) {
717 return (len - n)?(tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b)))):0;
720 /*******************************************************************
722 ********************************************************************/
724 bool strequal_w(const smb_ucs2_t *s1, const smb_ucs2_t *s2)
733 return(strcasecmp_w(s1,s2)==0);
736 /*******************************************************************
737 Compare 2 strings up to and including the nth char.
738 ******************************************************************/
740 bool strnequal_w(const smb_ucs2_t *s1,const smb_ucs2_t *s2,size_t n)
745 if (!s1 || !s2 || !n) {
749 return(strncasecmp_w(s1,s2,n)==0);
752 /*******************************************************************
754 ********************************************************************/
756 smb_ucs2_t *strdup_w(const smb_ucs2_t *src)
758 return strndup_w(src, 0);
761 /* if len == 0 then duplicate the whole string */
763 smb_ucs2_t *strndup_w(const smb_ucs2_t *src, size_t len)
770 dest = SMB_MALLOC_ARRAY(smb_ucs2_t, len + 1);
772 DEBUG(0,("strdup_w: out of memory!\n"));
776 memcpy(dest, src, len * sizeof(smb_ucs2_t));
781 /*******************************************************************
782 Copy a string with max len.
783 ********************************************************************/
785 smb_ucs2_t *strncpy_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
794 for (len = 0; (*COPY_UCS2_CHAR(&cp,(src+len))) && (len < max); len++) {
795 cp = *COPY_UCS2_CHAR(dest+len,src+len);
798 for ( /*nothing*/ ; len < max; len++ ) {
799 cp = *COPY_UCS2_CHAR(dest+len,&cp);
805 /*******************************************************************
806 Append a string of len bytes and add a terminator.
807 ********************************************************************/
809 smb_ucs2_t *strncat_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
819 start = strlen_w(dest);
820 len = strnlen_w(src, max);
822 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
823 z = *COPY_UCS2_CHAR(dest+start+len,&z);
828 smb_ucs2_t *strcat_w(smb_ucs2_t *dest, const smb_ucs2_t *src)
838 start = strlen_w(dest);
841 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
842 z = *COPY_UCS2_CHAR(dest+start+len,&z);
848 /*******************************************************************
849 Replace any occurence of oldc with newc in unicode string.
850 ********************************************************************/
852 void string_replace_w(smb_ucs2_t *s, smb_ucs2_t oldc, smb_ucs2_t newc)
856 for(;*(COPY_UCS2_CHAR(&cp,s));s++) {
858 COPY_UCS2_CHAR(s,&newc);
863 /*******************************************************************
865 ********************************************************************/
867 bool trim_string_w(smb_ucs2_t *s, const smb_ucs2_t *front,
868 const smb_ucs2_t *back)
871 size_t len, front_len, back_len;
879 if (front && *front) {
880 front_len = strlen_w(front);
881 while (len && strncmp_w(s, front, front_len) == 0) {
882 memmove(s, (s + front_len), (len - front_len + 1) * sizeof(smb_ucs2_t));
889 back_len = strlen_w(back);
890 while (len && strncmp_w((s + (len - back_len)), back, back_len) == 0) {
891 s[len - back_len] = 0;
901 The *_wa() functions take a combination of 7 bit ascii
902 and wide characters They are used so that you can use string
903 functions combining C string constants with ucs2 strings
905 The char* arguments must NOT be multibyte - to be completely sure
906 of this only pass string constants */
908 int strcmp_wa(const smb_ucs2_t *a, const char *b)
912 while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
916 return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
919 int strncmp_wa(const smb_ucs2_t *a, const char *b, size_t len)
924 while ((n < len) && *b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
929 return (len - n)?(*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b)):0;
932 smb_ucs2_t *strpbrk_wa(const smb_ucs2_t *s, const char *p)
936 while (*(COPY_UCS2_CHAR(&cp,s))) {
938 for (i=0; p[i] && cp != UCS2_CHAR(p[i]); i++)
941 return (smb_ucs2_t *)s;
948 smb_ucs2_t *strstr_wa(const smb_ucs2_t *s, const char *ins)
957 inslen = strlen(ins);
960 while ((r = strchr_w(r, UCS2_CHAR(*ins)))) {
961 if (strncmp_wa(r, ins, inslen) == 0)
969 /*******************************************************************
970 Returns the length in number of wide characters.
971 ******************************************************************/
973 int unistrlen(uint16 *s)
981 for (len=0; SVAL(s,0); s++,len++) {
988 /*******************************************************************
989 Strcpy for unicode strings. Returns length (in num of wide chars).
991 ********************************************************************/
993 int unistrcpy(uint16 *dst, uint16 *src)
997 while (SVAL(src,0)) {
1006 /*************************************************************
1007 ascii only toupper - saves the need for smbd to be in C locale.
1008 *************************************************************/
1010 int toupper_ascii(int c)
1012 smb_ucs2_t uc = toupper_w(UCS2_CHAR(c));
1013 return UCS2_TO_CHAR(uc);
1016 /*************************************************************
1017 ascii only tolower - saves the need for smbd to be in C locale.
1018 *************************************************************/
1020 int tolower_ascii(int c)
1022 smb_ucs2_t uc = tolower_w(UCS2_CHAR(c));
1023 return UCS2_TO_CHAR(uc);
1026 /*************************************************************
1027 ascii only isupper - saves the need for smbd to be in C locale.
1028 *************************************************************/
1030 int isupper_ascii(int c)
1032 return isupper_w(UCS2_CHAR(c));
1035 /*************************************************************
1036 ascii only islower - saves the need for smbd to be in C locale.
1037 *************************************************************/
1039 int islower_ascii(int c)
1041 return islower_w(UCS2_CHAR(c));