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;
38 * This table says which Unicode characters are valid dos
41 * Each value is just a single bit.
43 static uint8 doschar_table[8192]; /* 65536 characters / 8 bits/byte */
46 * Destroy global objects allocated by load_case_tables()
48 void gfree_case_tables(void)
51 if ( upcase_table_use_unmap )
52 unmap_file(upcase_table, 0x20000);
54 SAFE_FREE(upcase_table);
57 if ( lowcase_table ) {
58 if ( lowcase_table_use_unmap )
59 unmap_file(lowcase_table, 0x20000);
61 SAFE_FREE(lowcase_table);
65 if ( valid_table_use_unmap )
66 unmap_file(valid_table, 0x10000);
68 SAFE_FREE(valid_table);
73 * Load or generate the case handling tables.
75 * The case tables are defined in UCS2 and don't depend on any
76 * configured parameters, so they never need to be reloaded.
79 void load_case_tables(void)
81 static int initialised;
82 char *old_locale = NULL, *saved_locale = NULL;
90 upcase_table = (smb_ucs2_t *)map_file(lib_path("upcase.dat"),
92 upcase_table_use_unmap = ( upcase_table != NULL );
94 lowcase_table = (smb_ucs2_t *)map_file(lib_path("lowcase.dat"),
96 lowcase_table_use_unmap = ( lowcase_table != NULL );
99 /* Get the name of the current locale. */
100 old_locale = setlocale(LC_ALL, NULL);
103 /* Save it as it is in static storage. */
104 saved_locale = SMB_STRDUP(old_locale);
107 /* We set back the locale to C to get ASCII-compatible toupper/lower functions. */
108 setlocale(LC_ALL, "C");
111 /* we would like Samba to limp along even if these tables are
114 DEBUG(1,("creating lame upcase table\n"));
115 upcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
116 for (i=0;i<0x10000;i++) {
121 for (i=0;i<256;i++) {
123 SSVAL(&v, 0, UCS2_CHAR(i));
124 upcase_table[v] = UCS2_CHAR(islower(i)?toupper(i):i);
128 if (!lowcase_table) {
129 DEBUG(1,("creating lame lowcase table\n"));
130 lowcase_table = (smb_ucs2_t *)SMB_MALLOC(0x20000);
131 for (i=0;i<0x10000;i++) {
134 lowcase_table[v] = i;
136 for (i=0;i<256;i++) {
138 SSVAL(&v, 0, UCS2_CHAR(i));
139 lowcase_table[v] = UCS2_CHAR(isupper(i)?tolower(i):i);
143 #ifdef HAVE_SETLOCALE
144 /* Restore the old locale. */
146 setlocale (LC_ALL, saved_locale);
147 SAFE_FREE(saved_locale);
153 see if a ucs2 character can be mapped correctly to a dos character
154 and mapped back to the same character in ucs2
157 int check_dos_char(smb_ucs2_t c)
159 lazy_initialize_conv();
161 /* Find the right byte, and right bit within the byte; return
163 return (doschar_table[(c & 0xffff) / 8] & (1 << (c & 7))) != 0;
167 static int check_dos_char_slowly(smb_ucs2_t c)
173 len1 = convert_string(CH_UTF16LE, CH_DOS, &c, 2, buf, sizeof(buf),False);
177 len2 = convert_string(CH_DOS, CH_UTF16LE, buf, len1, &c2, 2,False);
186 * Fill out doschar table the hard way, by examining each character
189 void init_doschar_table(void)
193 /* For each byte of packed table */
195 for (i = 0; i <= 0xffff; i += 8) {
197 for (j = 0; j <= 7; j++) {
202 if (check_dos_char_slowly(c)) {
206 doschar_table[i/8] = byteval;
212 * Load the valid character map table from <tt>valid.dat</tt> or
213 * create from the configured codepage.
215 * This function is called whenever the configuration is reloaded.
216 * However, the valid character table is not changed if it's loaded
217 * from a file, because we can't unmap files.
220 void init_valid_table(void)
222 static int mapped_file;
224 const char *allowed = ".!#$%&'()_-@^`~";
228 /* Can't unmap files, so stick with what we have */
232 valid_file = (uint8 *)map_file(lib_path("valid.dat"), 0x10000);
234 valid_table = valid_file;
236 valid_table_use_unmap = True;
240 /* Otherwise, we're using a dynamically created valid_table.
241 * It might need to be regenerated if the code page changed.
242 * We know that we're not using a mapped file, so we can
243 * free() the old one. */
245 SAFE_FREE(valid_table);
247 /* use free rather than unmap */
248 valid_table_use_unmap = False;
250 DEBUG(2,("creating default valid table\n"));
251 valid_table = (uint8 *)SMB_MALLOC(0x10000);
252 for (i=0;i<128;i++) {
253 valid_table[i] = isalnum(i) || strchr(allowed,i);
256 for (;i<0x10000;i++) {
259 valid_table[i] = check_dos_char(c);
263 /*******************************************************************
264 Write a string in (little-endian) unicode format. src is in
265 the current DOS codepage. len is the length in bytes of the
266 string pointed to by dst.
268 if null_terminate is True then null terminate the packet (adds 2 bytes)
270 the return value is the length in bytes consumed by the string, including the
271 null termination if applied
272 ********************************************************************/
274 size_t dos_PutUniCode(char *dst,const char *src, size_t len, bool null_terminate)
276 int flags = null_terminate ? STR_UNICODE|STR_NOALIGN|STR_TERMINATE
277 : STR_UNICODE|STR_NOALIGN;
278 return push_ucs2(NULL, dst, src, len, flags);
282 /*******************************************************************
283 Skip past a unicode string, but not more than len. Always move
284 past a terminating zero if found.
285 ********************************************************************/
287 char *skip_unibuf(char *src, size_t len)
289 char *srcend = src + len;
291 while (src < srcend && SVAL(src,0)) {
302 /* Copy a string from little-endian or big-endian unicode source (depending
303 * on flags) to internal samba format destination
306 int rpcstr_pull(char* dest, void *src, int dest_len, int src_len, int flags)
315 return pull_ucs2(NULL, dest, src, dest_len, src_len, flags|STR_UNICODE|STR_NOALIGN);
318 /* Copy a string from a unistr2 source to internal samba format
319 destination. Use this instead of direct calls to rpcstr_pull() to avoid
320 having to determine whether the source string is null terminated. */
322 int rpcstr_pull_unistr2_fstring(char *dest, UNISTR2 *src)
324 return pull_ucs2(NULL, dest, src->buffer, sizeof(fstring),
325 src->uni_str_len * 2, 0);
328 /* Helper function to return a talloc'ed string. I have implemented it with a
329 * copy because I don't really know how pull_ucs2 and friends calculate the
330 * target size. If this turns out to be a major bottleneck someone with deeper
331 * multi-byte knowledge needs to revisit this.
332 * My (VL) use is dsr_getdcname, which returns 6 strings, the alternative would
333 * have been to manually talloc_strdup them in rpc_client/cli_netlogon.c.
336 char *rpcstr_pull_unistr2_talloc(TALLOC_CTX *mem_ctx, const UNISTR2 *src)
341 result = pull_ucs2(NULL, tmp, src->buffer, sizeof(tmp),
342 src->uni_str_len * 2, 0);
343 if (result == (size_t)-1) {
347 return talloc_strdup(mem_ctx, tmp);
350 /* Converts a string from internal samba format to unicode
353 int rpcstr_push(void* dest, const char *src, size_t dest_len, int flags)
355 return push_ucs2(NULL, dest, src, dest_len, flags|STR_UNICODE|STR_NOALIGN);
358 /*******************************************************************
359 Convert a (little-endian) UNISTR2 structure to an ASCII string.
360 ********************************************************************/
362 void unistr2_to_ascii(char *dest, const UNISTR2 *str, size_t maxlen)
364 if ((str == NULL) || (str->uni_str_len == 0)) {
368 pull_ucs2(NULL, dest, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
371 /*******************************************************************
372 Convert a (little-endian) UNISTR3 structure to an ASCII string.
373 ********************************************************************/
375 void unistr3_to_ascii(char *dest, const UNISTR3 *str, size_t maxlen)
377 if ((str == NULL) || (str->uni_str_len == 0)) {
381 pull_ucs2(NULL, dest, str->str.buffer, maxlen, str->uni_str_len*2,
385 /*******************************************************************
386 Give a static string for displaying a UNISTR2.
387 ********************************************************************/
389 const char *unistr2_static(const UNISTR2 *str)
392 unistr2_to_ascii(ret, str, sizeof(ret));
396 /*******************************************************************
397 Duplicate a UNISTR2 string into a null terminated char*
398 using a talloc context.
399 ********************************************************************/
401 char *unistr2_tdup(TALLOC_CTX *ctx, const UNISTR2 *str)
404 int maxlen = (str->uni_str_len+1)*4;
408 s = (char *)TALLOC(ctx, maxlen); /* convervative */
412 pull_ucs2(NULL, s, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
416 /*******************************************************************
417 Convert a wchar to upper case.
418 ********************************************************************/
420 smb_ucs2_t toupper_w(smb_ucs2_t val)
422 return upcase_table[SVAL(&val,0)];
425 /*******************************************************************
426 Convert a wchar to lower case.
427 ********************************************************************/
429 smb_ucs2_t tolower_w( smb_ucs2_t val )
431 return lowcase_table[SVAL(&val,0)];
434 /*******************************************************************
435 Determine if a character is lowercase.
436 ********************************************************************/
438 bool islower_w(smb_ucs2_t c)
440 return upcase_table[SVAL(&c,0)] != c;
443 /*******************************************************************
444 Determine if a character is uppercase.
445 ********************************************************************/
447 bool isupper_w(smb_ucs2_t c)
449 return lowcase_table[SVAL(&c,0)] != c;
452 /*******************************************************************
453 Determine if a character is valid in a 8.3 name.
454 ********************************************************************/
456 bool isvalid83_w(smb_ucs2_t c)
458 return valid_table[SVAL(&c,0)] != 0;
461 /*******************************************************************
462 Count the number of characters in a smb_ucs2_t string.
463 ********************************************************************/
465 size_t strlen_w(const smb_ucs2_t *src)
470 for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
477 /*******************************************************************
478 Count up to max number of characters in a smb_ucs2_t string.
479 ********************************************************************/
481 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
486 for(len = 0; (len < max) && *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
493 /*******************************************************************
495 ********************************************************************/
497 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
500 while (*(COPY_UCS2_CHAR(&cp,s))) {
502 return (smb_ucs2_t *)s;
507 return (smb_ucs2_t *)s;
513 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
515 return strchr_w(s, UCS2_CHAR(c));
518 /*******************************************************************
520 ********************************************************************/
522 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
525 const smb_ucs2_t *p = s;
526 int len = strlen_w(s);
533 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
534 return (smb_ucs2_t *)p;
540 /*******************************************************************
541 Wide version of strrchr that returns after doing strrchr 'n' times.
542 ********************************************************************/
544 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
547 const smb_ucs2_t *p = s;
548 int len = strlen_w(s);
550 if (len == 0 || !n) {
555 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
560 return (smb_ucs2_t *)p;
566 /*******************************************************************
568 ********************************************************************/
570 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
575 if (!s || !*s || !ins || !*ins) {
579 inslen = strlen_w(ins);
582 while ((r = strchr_w(r, *ins))) {
583 if (strncmp_w(r, ins, inslen) == 0) {
592 /*******************************************************************
593 Convert a string to lower case.
594 return True if any char is converted
595 ********************************************************************/
597 bool strlower_w(smb_ucs2_t *s)
602 while (*(COPY_UCS2_CHAR(&cp,s))) {
603 smb_ucs2_t v = tolower_w(cp);
605 COPY_UCS2_CHAR(s,&v);
613 /*******************************************************************
614 Convert a string to upper case.
615 return True if any char is converted
616 ********************************************************************/
618 bool strupper_w(smb_ucs2_t *s)
622 while (*(COPY_UCS2_CHAR(&cp,s))) {
623 smb_ucs2_t v = toupper_w(cp);
625 COPY_UCS2_CHAR(s,&v);
633 /*******************************************************************
634 Convert a string to "normal" form.
635 ********************************************************************/
637 void strnorm_w(smb_ucs2_t *s, int case_default)
639 if (case_default == CASE_UPPER) {
646 int strcmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
650 while ((*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
654 return (*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b)));
655 /* warning: if *a != *b and both are not 0 we return a random
656 greater or lesser than 0 number not realted to which
660 int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
665 while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
670 return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
673 /*******************************************************************
674 Case insensitive string comparison.
675 ********************************************************************/
677 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
681 while ((*COPY_UCS2_CHAR(&cpb,b)) && toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb)) {
685 return (tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b))));
688 /*******************************************************************
689 Case insensitive string comparison, length limited.
690 ********************************************************************/
692 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
697 while ((n < len) && *COPY_UCS2_CHAR(&cpb,b) && (toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb))) {
702 return (len - n)?(tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b)))):0;
705 /*******************************************************************
707 ********************************************************************/
709 bool strequal_w(const smb_ucs2_t *s1, const smb_ucs2_t *s2)
718 return(strcasecmp_w(s1,s2)==0);
721 /*******************************************************************
722 Compare 2 strings up to and including the nth char.
723 ******************************************************************/
725 bool strnequal_w(const smb_ucs2_t *s1,const smb_ucs2_t *s2,size_t n)
730 if (!s1 || !s2 || !n) {
734 return(strncasecmp_w(s1,s2,n)==0);
737 /*******************************************************************
739 ********************************************************************/
741 smb_ucs2_t *strdup_w(const smb_ucs2_t *src)
743 return strndup_w(src, 0);
746 /* if len == 0 then duplicate the whole string */
748 smb_ucs2_t *strndup_w(const smb_ucs2_t *src, size_t len)
755 dest = SMB_MALLOC_ARRAY(smb_ucs2_t, len + 1);
757 DEBUG(0,("strdup_w: out of memory!\n"));
761 memcpy(dest, src, len * sizeof(smb_ucs2_t));
766 /*******************************************************************
767 Copy a string with max len.
768 ********************************************************************/
770 smb_ucs2_t *strncpy_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
779 for (len = 0; (*COPY_UCS2_CHAR(&cp,(src+len))) && (len < max); len++) {
780 cp = *COPY_UCS2_CHAR(dest+len,src+len);
783 for ( /*nothing*/ ; len < max; len++ ) {
784 cp = *COPY_UCS2_CHAR(dest+len,&cp);
790 /*******************************************************************
791 Append a string of len bytes and add a terminator.
792 ********************************************************************/
794 smb_ucs2_t *strncat_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
804 start = strlen_w(dest);
805 len = strnlen_w(src, max);
807 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
808 z = *COPY_UCS2_CHAR(dest+start+len,&z);
813 smb_ucs2_t *strcat_w(smb_ucs2_t *dest, const smb_ucs2_t *src)
823 start = strlen_w(dest);
826 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
827 z = *COPY_UCS2_CHAR(dest+start+len,&z);
833 /*******************************************************************
834 Replace any occurence of oldc with newc in unicode string.
835 ********************************************************************/
837 void string_replace_w(smb_ucs2_t *s, smb_ucs2_t oldc, smb_ucs2_t newc)
841 for(;*(COPY_UCS2_CHAR(&cp,s));s++) {
843 COPY_UCS2_CHAR(s,&newc);
848 /*******************************************************************
850 ********************************************************************/
852 bool trim_string_w(smb_ucs2_t *s, const smb_ucs2_t *front,
853 const smb_ucs2_t *back)
856 size_t len, front_len, back_len;
864 if (front && *front) {
865 front_len = strlen_w(front);
866 while (len && strncmp_w(s, front, front_len) == 0) {
867 memmove(s, (s + front_len), (len - front_len + 1) * sizeof(smb_ucs2_t));
874 back_len = strlen_w(back);
875 while (len && strncmp_w((s + (len - back_len)), back, back_len) == 0) {
876 s[len - back_len] = 0;
886 The *_wa() functions take a combination of 7 bit ascii
887 and wide characters They are used so that you can use string
888 functions combining C string constants with ucs2 strings
890 The char* arguments must NOT be multibyte - to be completely sure
891 of this only pass string constants */
893 int strcmp_wa(const smb_ucs2_t *a, const char *b)
897 while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
901 return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
904 int strncmp_wa(const smb_ucs2_t *a, const char *b, size_t len)
909 while ((n < len) && *b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
914 return (len - n)?(*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b)):0;
917 smb_ucs2_t *strpbrk_wa(const smb_ucs2_t *s, const char *p)
921 while (*(COPY_UCS2_CHAR(&cp,s))) {
923 for (i=0; p[i] && cp != UCS2_CHAR(p[i]); i++)
926 return (smb_ucs2_t *)s;
933 smb_ucs2_t *strstr_wa(const smb_ucs2_t *s, const char *ins)
942 inslen = strlen(ins);
945 while ((r = strchr_w(r, UCS2_CHAR(*ins)))) {
946 if (strncmp_wa(r, ins, inslen) == 0)
954 bool trim_string_wa(smb_ucs2_t *s, const char *front,
960 push_ucs2(NULL, f, front, sizeof(wpstring) - 1, STR_TERMINATE);
965 push_ucs2(NULL, b, back, sizeof(wpstring) - 1, STR_TERMINATE);
969 return trim_string_w(s, f, b);
972 /*******************************************************************
973 Returns the length in number of wide characters.
974 ******************************************************************/
976 int unistrlen(uint16 *s)
984 for (len=0; SVAL(s,0); s++,len++) {
991 /*******************************************************************
992 Strcpy for unicode strings. Returns length (in num of wide chars).
994 ********************************************************************/
996 int unistrcpy(uint16 *dst, uint16 *src)
1000 while (SVAL(src,0)) {
1010 * Samba ucs2 type to UNISTR2 conversion
1012 * @param ctx Talloc context to create the dst strcture (if null) and the
1013 * contents of the unicode string.
1014 * @param dst UNISTR2 destination. If equals null, then it's allocated.
1015 * @param src smb_ucs2_t source.
1016 * @param max_len maximum number of unicode characters to copy. If equals
1017 * null, then null-termination of src is taken
1019 * @return copied UNISTR2 destination
1022 UNISTR2* ucs2_to_unistr2(TALLOC_CTX *ctx, UNISTR2* dst, smb_ucs2_t* src)
1030 len = strlen_w(src);
1032 /* allocate UNISTR2 destination if not given */
1034 dst = TALLOC_P(ctx, UNISTR2);
1039 dst->buffer = TALLOC_ARRAY(ctx, uint16, len + 1);
1044 /* set UNISTR2 parameters */
1045 dst->uni_max_len = len + 1;
1047 dst->uni_str_len = len;
1049 /* copy the actual unicode string */
1050 strncpy_w(dst->buffer, src, dst->uni_max_len);
1055 /*************************************************************
1056 ascii only toupper - saves the need for smbd to be in C locale.
1057 *************************************************************/
1059 int toupper_ascii(int c)
1061 smb_ucs2_t uc = toupper_w(UCS2_CHAR(c));
1062 return UCS2_TO_CHAR(uc);
1065 /*************************************************************
1066 ascii only tolower - saves the need for smbd to be in C locale.
1067 *************************************************************/
1069 int tolower_ascii(int c)
1071 smb_ucs2_t uc = tolower_w(UCS2_CHAR(c));
1072 return UCS2_TO_CHAR(uc);
1075 /*************************************************************
1076 ascii only isupper - saves the need for smbd to be in C locale.
1077 *************************************************************/
1079 int isupper_ascii(int c)
1081 return isupper_w(UCS2_CHAR(c));
1084 /*************************************************************
1085 ascii only islower - saves the need for smbd to be in C locale.
1086 *************************************************************/
1088 int islower_ascii(int c)
1090 return islower_w(UCS2_CHAR(c));