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 2 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, write to the Free Software
20 Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
29 /* these 3 tables define the unicode case handling. They are loaded
30 at startup either via mmap() or read() from the lib directory */
31 static smb_ucs2_t *upcase_table;
32 static smb_ucs2_t *lowcase_table;
33 static uint8 *valid_table;
36 * This table says which Unicode characters are valid dos
39 * Each value is just a single bit.
41 static uint8 doschar_table[8192]; /* 65536 characters / 8 bits/byte */
45 * Load or generate the case handling tables.
47 * The case tables are defined in UCS2 and don't depend on any
48 * configured parameters, so they never need to be reloaded.
51 void load_case_tables(void)
53 static int initialised;
54 char *old_locale = NULL, *saved_locale = NULL;
62 upcase_table = map_file(lib_path("upcase.dat"), 0x20000);
63 lowcase_table = map_file(lib_path("lowcase.dat"), 0x20000);
66 /* Get the name of the current locale. */
67 old_locale = setlocale(LC_ALL, NULL);
69 /* Save it as it is in static storage. */
70 saved_locale = SMB_STRDUP(old_locale);
72 /* We set back the locale to C to get ASCII-compatible toupper/lower functions. */
73 setlocale(LC_ALL, "C");
76 /* we would like Samba to limp along even if these tables are
79 DEBUG(1,("creating lame upcase table\n"));
80 upcase_table = SMB_MALLOC(0x20000);
81 for (i=0;i<0x10000;i++) {
88 SSVAL(&v, 0, UCS2_CHAR(i));
89 upcase_table[v] = UCS2_CHAR(islower(i)?toupper(i):i);
94 DEBUG(1,("creating lame lowcase table\n"));
95 lowcase_table = SMB_MALLOC(0x20000);
96 for (i=0;i<0x10000;i++) {
101 for (i=0;i<256;i++) {
103 SSVAL(&v, 0, UCS2_CHAR(i));
104 lowcase_table[v] = UCS2_CHAR(isupper(i)?tolower(i):i);
108 #ifdef HAVE_SETLOCALE
109 /* Restore the old locale. */
110 setlocale (LC_ALL, saved_locale);
111 SAFE_FREE(saved_locale);
116 see if a ucs2 character can be mapped correctly to a dos character
117 and mapped back to the same character in ucs2
120 int check_dos_char(smb_ucs2_t c)
122 lazy_initialize_conv();
124 /* Find the right byte, and right bit within the byte; return
126 return (doschar_table[(c & 0xffff) / 8] & (1 << (c & 7))) != 0;
130 static int check_dos_char_slowly(smb_ucs2_t c)
136 len1 = convert_string(CH_UCS2, CH_DOS, &c, 2, buf, sizeof(buf),False);
140 len2 = convert_string(CH_DOS, CH_UCS2, buf, len1, &c2, 2,False);
149 * Fill out doschar table the hard way, by examining each character
152 void init_doschar_table(void)
156 /* For each byte of packed table */
158 for (i = 0; i <= 0xffff; i += 8) {
160 for (j = 0; j <= 7; j++) {
165 if (check_dos_char_slowly(c)) {
169 doschar_table[i/8] = byteval;
175 * Load the valid character map table from <tt>valid.dat</tt> or
176 * create from the configured codepage.
178 * This function is called whenever the configuration is reloaded.
179 * However, the valid character table is not changed if it's loaded
180 * from a file, because we can't unmap files.
183 void init_valid_table(void)
185 static int mapped_file;
187 const char *allowed = ".!#$%&'()_-@^`~";
191 /* Can't unmap files, so stick with what we have */
195 valid_file = map_file(lib_path("valid.dat"), 0x10000);
197 valid_table = valid_file;
202 /* Otherwise, we're using a dynamically created valid_table.
203 * It might need to be regenerated if the code page changed.
204 * We know that we're not using a mapped file, so we can
205 * free() the old one. */
206 if (valid_table) free(valid_table);
208 DEBUG(2,("creating default valid table\n"));
209 valid_table = SMB_MALLOC(0x10000);
210 for (i=0;i<128;i++) {
211 valid_table[i] = isalnum(i) || strchr(allowed,i);
214 for (;i<0x10000;i++) {
217 valid_table[i] = check_dos_char(c);
221 /*******************************************************************
222 Write a string in (little-endian) unicode format. src is in
223 the current DOS codepage. len is the length in bytes of the
224 string pointed to by dst.
226 if null_terminate is True then null terminate the packet (adds 2 bytes)
228 the return value is the length in bytes consumed by the string, including the
229 null termination if applied
230 ********************************************************************/
232 size_t dos_PutUniCode(char *dst,const char *src, ssize_t len, BOOL null_terminate)
234 int flags = null_terminate ? STR_UNICODE|STR_NOALIGN|STR_TERMINATE
235 : STR_UNICODE|STR_NOALIGN;
236 return push_ucs2(NULL, dst, src, len, flags);
240 /*******************************************************************
241 Skip past a unicode string, but not more than len. Always move
242 past a terminating zero if found.
243 ********************************************************************/
245 char *skip_unibuf(char *src, size_t len)
247 char *srcend = src + len;
249 while (src < srcend && SVAL(src,0)) {
260 /* Copy a string from little-endian or big-endian unicode source (depending
261 * on flags) to internal samba format destination
264 int rpcstr_pull(char* dest, void *src, int dest_len, int src_len, int flags)
273 return pull_ucs2(NULL, dest, src, dest_len, src_len, flags|STR_UNICODE|STR_NOALIGN);
276 /* Copy a string from a unistr2 source to internal samba format
277 destination. Use this instead of direct calls to rpcstr_pull() to avoid
278 having to determine whether the source string is null terminated. */
280 int rpcstr_pull_unistr2_fstring(char *dest, UNISTR2 *src)
282 return pull_ucs2(NULL, dest, src->buffer, sizeof(fstring),
283 src->uni_str_len * 2, 0);
286 /* Helper function to return a talloc'ed string. I have implemented it with a
287 * copy because I don't really know how pull_ucs2 and friends calculate the
288 * target size. If this turns out to be a major bottleneck someone with deeper
289 * multi-byte knowledge needs to revisit this.
290 * My (VL) use is dsr_getdcname, which returns 6 strings, the alternative would
291 * have been to manually talloc_strdup them in rpc_client/cli_netlogon.c.
294 char *rpcstr_pull_unistr2_talloc(TALLOC_CTX *mem_ctx, const UNISTR2 *src)
299 result = pull_ucs2(NULL, tmp, src->buffer, sizeof(tmp),
300 src->uni_str_len * 2, 0);
301 if (result == (size_t)-1) {
305 return talloc_strdup(mem_ctx, tmp);
308 /* Converts a string from internal samba format to unicode
311 int rpcstr_push(void* dest, const char *src, int dest_len, int flags)
313 return push_ucs2(NULL, dest, src, dest_len, flags|STR_UNICODE|STR_NOALIGN);
316 /*******************************************************************
317 Convert a (little-endian) UNISTR2 structure to an ASCII string.
318 ********************************************************************/
320 void unistr2_to_ascii(char *dest, const UNISTR2 *str, size_t maxlen)
326 pull_ucs2(NULL, dest, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
329 /*******************************************************************
330 Convert a (little-endian) UNISTR3 structure to an ASCII string.
331 ********************************************************************/
333 void unistr3_to_ascii(char *dest, const UNISTR3 *str, size_t maxlen)
339 pull_ucs2(NULL, dest, str->str.buffer, maxlen, str->uni_str_len*2,
343 /*******************************************************************
344 Give a static string for displaying a UNISTR2.
345 ********************************************************************/
347 const char *unistr2_static(const UNISTR2 *str)
350 unistr2_to_ascii(ret, str, sizeof(ret));
354 /*******************************************************************
355 Duplicate a UNISTR2 string into a null terminated char*
356 using a talloc context.
357 ********************************************************************/
359 char *unistr2_tdup(TALLOC_CTX *ctx, const UNISTR2 *str)
362 int maxlen = (str->uni_str_len+1)*4;
366 s = (char *)TALLOC(ctx, maxlen); /* convervative */
370 pull_ucs2(NULL, s, str->buffer, maxlen, str->uni_str_len*2, STR_NOALIGN);
374 /*******************************************************************
375 Convert a wchar to upper case.
376 ********************************************************************/
378 smb_ucs2_t toupper_w(smb_ucs2_t val)
380 return upcase_table[SVAL(&val,0)];
383 /*******************************************************************
384 Convert a wchar to lower case.
385 ********************************************************************/
387 smb_ucs2_t tolower_w( smb_ucs2_t val )
389 return lowcase_table[SVAL(&val,0)];
392 /*******************************************************************
393 Determine if a character is lowercase.
394 ********************************************************************/
396 BOOL islower_w(smb_ucs2_t c)
398 return upcase_table[SVAL(&c,0)] != c;
401 /*******************************************************************
402 Determine if a character is uppercase.
403 ********************************************************************/
405 BOOL isupper_w(smb_ucs2_t c)
407 return lowcase_table[SVAL(&c,0)] != c;
410 /*******************************************************************
411 Determine if a character is valid in a 8.3 name.
412 ********************************************************************/
414 BOOL isvalid83_w(smb_ucs2_t c)
416 return valid_table[SVAL(&c,0)] != 0;
419 /*******************************************************************
420 Count the number of characters in a smb_ucs2_t string.
421 ********************************************************************/
423 size_t strlen_w(const smb_ucs2_t *src)
428 for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
435 /*******************************************************************
436 Count up to max number of characters in a smb_ucs2_t string.
437 ********************************************************************/
439 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
444 for(len = 0; *(COPY_UCS2_CHAR(&c,src)) && (len < max); src++, len++) {
451 /*******************************************************************
453 ********************************************************************/
455 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
458 while (*(COPY_UCS2_CHAR(&cp,s))) {
460 return (smb_ucs2_t *)s;
465 return (smb_ucs2_t *)s;
471 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
473 return strchr_w(s, UCS2_CHAR(c));
476 /*******************************************************************
478 ********************************************************************/
480 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
483 const smb_ucs2_t *p = s;
484 int len = strlen_w(s);
491 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
492 return (smb_ucs2_t *)p;
498 /*******************************************************************
499 Wide version of strrchr that returns after doing strrchr 'n' times.
500 ********************************************************************/
502 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
505 const smb_ucs2_t *p = s;
506 int len = strlen_w(s);
508 if (len == 0 || !n) {
513 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
518 return (smb_ucs2_t *)p;
524 /*******************************************************************
526 ********************************************************************/
528 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
533 if (!s || !*s || !ins || !*ins) {
537 inslen = strlen_w(ins);
540 while ((r = strchr_w(r, *ins))) {
541 if (strncmp_w(r, ins, inslen) == 0) {
550 /*******************************************************************
551 Convert a string to lower case.
552 return True if any char is converted
553 ********************************************************************/
555 BOOL strlower_w(smb_ucs2_t *s)
560 while (*(COPY_UCS2_CHAR(&cp,s))) {
561 smb_ucs2_t v = tolower_w(cp);
563 COPY_UCS2_CHAR(s,&v);
571 /*******************************************************************
572 Convert a string to upper case.
573 return True if any char is converted
574 ********************************************************************/
576 BOOL strupper_w(smb_ucs2_t *s)
580 while (*(COPY_UCS2_CHAR(&cp,s))) {
581 smb_ucs2_t v = toupper_w(cp);
583 COPY_UCS2_CHAR(s,&v);
591 /*******************************************************************
592 Convert a string to "normal" form.
593 ********************************************************************/
595 void strnorm_w(smb_ucs2_t *s, int case_default)
597 if (case_default == CASE_UPPER) {
604 int strcmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
608 while ((*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
612 return (*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b)));
613 /* warning: if *a != *b and both are not 0 we return a random
614 greater or lesser than 0 number not realted to which
618 int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
623 while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
628 return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
631 /*******************************************************************
632 Case insensitive string comparison.
633 ********************************************************************/
635 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b)
639 while ((*COPY_UCS2_CHAR(&cpb,b)) && toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb)) {
643 return (tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b))));
646 /*******************************************************************
647 Case insensitive string comparison, length limited.
648 ********************************************************************/
650 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
655 while ((n < len) && *COPY_UCS2_CHAR(&cpb,b) && (toupper_w(*(COPY_UCS2_CHAR(&cpa,a))) == toupper_w(cpb))) {
660 return (len - n)?(tolower_w(*(COPY_UCS2_CHAR(&cpa,a))) - tolower_w(*(COPY_UCS2_CHAR(&cpb,b)))):0;
663 /*******************************************************************
665 ********************************************************************/
667 BOOL strequal_w(const smb_ucs2_t *s1, const smb_ucs2_t *s2)
676 return(strcasecmp_w(s1,s2)==0);
679 /*******************************************************************
680 Compare 2 strings up to and including the nth char.
681 ******************************************************************/
683 BOOL strnequal_w(const smb_ucs2_t *s1,const smb_ucs2_t *s2,size_t n)
688 if (!s1 || !s2 || !n) {
692 return(strncasecmp_w(s1,s2,n)==0);
695 /*******************************************************************
697 ********************************************************************/
699 smb_ucs2_t *strdup_w(const smb_ucs2_t *src)
701 return strndup_w(src, 0);
704 /* if len == 0 then duplicate the whole string */
706 smb_ucs2_t *strndup_w(const smb_ucs2_t *src, size_t len)
713 dest = SMB_MALLOC_ARRAY(smb_ucs2_t, len + 1);
715 DEBUG(0,("strdup_w: out of memory!\n"));
719 memcpy(dest, src, len * sizeof(smb_ucs2_t));
724 /*******************************************************************
725 Copy a string with max len.
726 ********************************************************************/
728 smb_ucs2_t *strncpy_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
737 for (len = 0; (*COPY_UCS2_CHAR(&cp,(src+len))) && (len < max); len++) {
738 cp = *COPY_UCS2_CHAR(dest+len,src+len);
741 for ( /*nothing*/ ; len < max; len++ ) {
742 cp = *COPY_UCS2_CHAR(dest+len,&cp);
748 /*******************************************************************
749 Append a string of len bytes and add a terminator.
750 ********************************************************************/
752 smb_ucs2_t *strncat_w(smb_ucs2_t *dest, const smb_ucs2_t *src, const size_t max)
762 start = strlen_w(dest);
763 len = strnlen_w(src, max);
765 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
766 z = *COPY_UCS2_CHAR(dest+start+len,&z);
771 smb_ucs2_t *strcat_w(smb_ucs2_t *dest, const smb_ucs2_t *src)
781 start = strlen_w(dest);
784 memcpy(&dest[start], src, len*sizeof(smb_ucs2_t));
785 z = *COPY_UCS2_CHAR(dest+start+len,&z);
791 /*******************************************************************
792 Replace any occurence of oldc with newc in unicode string.
793 ********************************************************************/
795 void string_replace_w(smb_ucs2_t *s, smb_ucs2_t oldc, smb_ucs2_t newc)
799 for(;*(COPY_UCS2_CHAR(&cp,s));s++) {
801 COPY_UCS2_CHAR(s,&newc);
806 /*******************************************************************
808 ********************************************************************/
810 BOOL trim_string_w(smb_ucs2_t *s, const smb_ucs2_t *front,
811 const smb_ucs2_t *back)
814 size_t len, front_len, back_len;
822 if (front && *front) {
823 front_len = strlen_w(front);
824 while (len && strncmp_w(s, front, front_len) == 0) {
825 memmove(s, (s + front_len), (len - front_len + 1) * sizeof(smb_ucs2_t));
832 back_len = strlen_w(back);
833 while (len && strncmp_w((s + (len - back_len)), back, back_len) == 0) {
834 s[len - back_len] = 0;
844 The *_wa() functions take a combination of 7 bit ascii
845 and wide characters They are used so that you can use string
846 functions combining C string constants with ucs2 strings
848 The char* arguments must NOT be multibyte - to be completely sure
849 of this only pass string constants */
851 int strcmp_wa(const smb_ucs2_t *a, const char *b)
855 while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
859 return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
862 int strncmp_wa(const smb_ucs2_t *a, const char *b, size_t len)
867 while ((n < len) && *b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
872 return (len - n)?(*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b)):0;
875 smb_ucs2_t *strpbrk_wa(const smb_ucs2_t *s, const char *p)
879 while (*(COPY_UCS2_CHAR(&cp,s))) {
881 for (i=0; p[i] && cp != UCS2_CHAR(p[i]); i++)
884 return (smb_ucs2_t *)s;
891 smb_ucs2_t *strstr_wa(const smb_ucs2_t *s, const char *ins)
900 inslen = strlen(ins);
903 while ((r = strchr_w(r, UCS2_CHAR(*ins)))) {
904 if (strncmp_wa(r, ins, inslen) == 0)
912 BOOL trim_string_wa(smb_ucs2_t *s, const char *front,
918 push_ucs2(NULL, f, front, sizeof(wpstring) - 1, STR_TERMINATE);
923 push_ucs2(NULL, b, back, sizeof(wpstring) - 1, STR_TERMINATE);
927 return trim_string_w(s, f, b);
930 /*******************************************************************
931 Returns the length in number of wide characters.
932 ******************************************************************/
934 int unistrlen(uint16 *s)
942 for (len=0; SVAL(s,0); s++,len++) {
949 /*******************************************************************
950 Strcpy for unicode strings. Returns length (in num of wide chars).
952 ********************************************************************/
954 int unistrcpy(uint16 *dst, uint16 *src)
958 while (SVAL(src,0)) {
968 * Samba ucs2 type to UNISTR2 conversion
970 * @param ctx Talloc context to create the dst strcture (if null) and the
971 * contents of the unicode string.
972 * @param dst UNISTR2 destination. If equals null, then it's allocated.
973 * @param src smb_ucs2_t source.
974 * @param max_len maximum number of unicode characters to copy. If equals
975 * null, then null-termination of src is taken
977 * @return copied UNISTR2 destination
980 UNISTR2* ucs2_to_unistr2(TALLOC_CTX *ctx, UNISTR2* dst, smb_ucs2_t* src)
990 /* allocate UNISTR2 destination if not given */
992 dst = TALLOC_P(ctx, UNISTR2);
997 dst->buffer = TALLOC_ARRAY(ctx, uint16, len + 1);
1002 /* set UNISTR2 parameters */
1003 dst->uni_max_len = len + 1;
1005 dst->uni_str_len = len;
1007 /* copy the actual unicode string */
1008 strncpy_w(dst->buffer, src, dst->uni_max_len);
1013 /*************************************************************
1014 ascii only toupper - saves the need for smbd to be in C locale.
1015 *************************************************************/
1017 int toupper_ascii(int c)
1019 smb_ucs2_t uc = toupper_w(UCS2_CHAR(c));
1020 return UCS2_TO_CHAR(uc);
1023 /*************************************************************
1024 ascii only tolower - saves the need for smbd to be in C locale.
1025 *************************************************************/
1027 int tolower_ascii(int c)
1029 smb_ucs2_t uc = tolower_w(UCS2_CHAR(c));
1030 return UCS2_TO_CHAR(uc);
1033 /*************************************************************
1034 ascii only isupper - saves the need for smbd to be in C locale.
1035 *************************************************************/
1037 int isupper_ascii(int c)
1039 return isupper_w(UCS2_CHAR(c));
1042 /*************************************************************
1043 ascii only islower - saves the need for smbd to be in C locale.
1044 *************************************************************/
1046 int islower_ascii(int c)
1048 return islower_w(UCS2_CHAR(c));