2 Unix SMB/CIFS implementation.
3 Samba utility functions
5 Copyright (C) Andrew Tridgell 1992-2001
6 Copyright (C) Simo Sorce 2001-2002
7 Copyright (C) Martin Pool 2003
9 This program is free software; you can redistribute it and/or modify
10 it under the terms of the GNU General Public License as published by
11 the Free Software Foundation; either version 2 of the License, or
12 (at your option) any later version.
14 This program is distributed in the hope that it will be useful,
15 but WITHOUT ANY WARRANTY; without even the implied warranty of
16 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 GNU General Public License for more details.
19 You should have received a copy of the GNU General Public License
20 along with this program; if not, write to the Free Software
21 Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
25 #include "system/iconv.h"
28 #include "lib/ldb/include/ldb.h"
32 * @brief String utilities.
36 * Get the next token from a string, return False if none found.
37 * Handles double-quotes.
39 * Based on a routine by GJC@VILLAGE.COM.
40 * Extensively modified by Andrew.Tridgell@anu.edu.au
42 BOOL next_token(const char **ptr,char *buff, const char *sep, size_t bufsize)
53 /* default to simple separators */
57 /* find the first non sep char */
58 while (*s && strchr_m(sep,*s))
65 /* copy over the token */
66 for (quoted = False; len < bufsize && *s && (quoted || !strchr_m(sep,*s)); s++) {
75 *ptr = (*s) ? s+1 : s;
82 Case insensitive string compararison
84 int strcasecmp_m(const char *s1, const char *s2)
86 codepoint_t c1=0, c2=0;
90 c1 = next_codepoint(s1, &size1);
91 c2 = next_codepoint(s2, &size2);
100 if (c1 == INVALID_CODEPOINT ||
101 c2 == INVALID_CODEPOINT) {
102 /* what else can we do?? */
106 if (toupper_w(c1) != toupper_w(c2)) {
117 * @note The comparison is case-insensitive.
119 BOOL strequal(const char *s1, const char *s2)
126 return strcasecmp_m(s1,s2) == 0;
130 Compare 2 strings (case sensitive).
132 BOOL strcsequal(const char *s1,const char *s2)
139 return strcmp(s1,s2) == 0;
144 Do a case-insensitive, whitespace-ignoring string compare.
146 int strwicmp(const char *psz1, const char *psz2)
148 /* if BOTH strings are NULL, return TRUE, if ONE is NULL return */
149 /* appropriate value. */
152 else if (psz1 == NULL)
154 else if (psz2 == NULL)
157 /* sync the strings on first non-whitespace */
159 while (isspace((int)*psz1))
161 while (isspace((int)*psz2))
163 if (toupper((unsigned char)*psz1) != toupper((unsigned char)*psz2)
170 return (*psz1 - *psz2);
175 NOTE: oldc and newc must be 7 bit characters
177 void string_replace(char *s, char oldc, char newc)
181 codepoint_t c = next_codepoint(s, &size);
190 Trim the specified elements off the front and back of a string.
192 BOOL trim_string(char *s,const char *front,const char *back)
199 /* Ignore null or empty strings. */
200 if (!s || (s[0] == '\0'))
203 front_len = front? strlen(front) : 0;
204 back_len = back? strlen(back) : 0;
209 while (len && strncmp(s, front, front_len)==0) {
210 /* Must use memmove here as src & dest can
211 * easily overlap. Found by valgrind. JRA. */
212 memmove(s, s+front_len, (len-front_len)+1);
219 while ((len >= back_len) && strncmp(s+len-back_len,back,back_len)==0) {
220 s[len-back_len]='\0';
229 Find the number of 'c' chars in a string
231 size_t count_chars(const char *s, char c)
237 codepoint_t c2 = next_codepoint(s, &size);
238 if (c2 == c) count++;
246 Safe string copy into a known length string. maxlength does not
247 include the terminating zero.
249 char *safe_strcpy(char *dest,const char *src, size_t maxlength)
254 DEBUG(0,("ERROR: NULL dest in safe_strcpy\n"));
259 /* We intentionally write out at the extremity of the destination
260 * string. If the destination is too short (e.g. pstrcpy into mallocd
261 * or fstring) then this should cause an error under a memory
263 dest[maxlength] = '\0';
264 if (PTR_DIFF(&len, dest) > 0) { /* check if destination is on the stack, ok if so */
265 log_suspicious_usage("safe_strcpy", src);
276 if (len > maxlength) {
277 DEBUG(0,("ERROR: string overflow by %u (%u - %u) in safe_strcpy [%.50s]\n",
278 (uint_t)(len-maxlength), (unsigned)len, (unsigned)maxlength, src));
282 memmove(dest, src, len);
288 Safe string cat into a string. maxlength does not
289 include the terminating zero.
291 char *safe_strcat(char *dest, const char *src, size_t maxlength)
293 size_t src_len, dest_len;
296 DEBUG(0,("ERROR: NULL dest in safe_strcat\n"));
304 if (PTR_DIFF(&src_len, dest) > 0) { /* check if destination is on the stack, ok if so */
305 log_suspicious_usage("safe_strcat", src);
308 src_len = strlen(src);
309 dest_len = strlen(dest);
311 if (src_len + dest_len > maxlength) {
312 DEBUG(0,("ERROR: string overflow by %d in safe_strcat [%.50s]\n",
313 (int)(src_len + dest_len - maxlength), src));
314 if (maxlength > dest_len) {
315 memcpy(&dest[dest_len], src, maxlength - dest_len);
321 memcpy(&dest[dest_len], src, src_len);
322 dest[dest_len + src_len] = 0;
327 Paranoid strcpy into a buffer of given length (includes terminating
328 zero. Strips out all but 'a-Z0-9' and the character in other_safe_chars
329 and replaces with '_'. Deliberately does *NOT* check for multibyte
330 characters. Don't change it !
333 char *alpha_strcpy(char *dest, const char *src, const char *other_safe_chars, size_t maxlength)
337 if (maxlength == 0) {
338 /* can't fit any bytes at all! */
343 DEBUG(0,("ERROR: NULL dest in alpha_strcpy\n"));
353 if (len >= maxlength)
356 if (!other_safe_chars)
357 other_safe_chars = "";
359 for(i = 0; i < len; i++) {
360 int val = (src[i] & 0xff);
361 if (isupper(val) || islower(val) || isdigit(val) || strchr_m(other_safe_chars, val))
373 Like strncpy but always null terminates. Make sure there is room!
374 The variable n should always be one less than the available size.
377 char *StrnCpy(char *dest,const char *src,size_t n)
386 while (n-- && (*d++ = *src++))
394 Routine to get hex characters and turn them into a 16 byte array.
395 the array can be variable length, and any non-hex-numeric
396 characters are skipped. "0xnn" or "0Xnn" is specially catered
399 valid examples: "0A5D15"; "0x15, 0x49, 0xa2"; "59\ta9\te3\n"
403 size_t strhex_to_str(char *p, size_t len, const char *strhex)
406 size_t num_chars = 0;
407 uint8_t lonybble, hinybble;
408 const char *hexchars = "0123456789ABCDEF";
409 char *p1 = NULL, *p2 = NULL;
411 for (i = 0; i < len && strhex[i] != 0; i++) {
412 if (strncasecmp(hexchars, "0x", 2) == 0) {
413 i++; /* skip two chars */
417 if (!(p1 = strchr_m(hexchars, toupper((unsigned char)strhex[i]))))
420 i++; /* next hex digit */
422 if (!(p2 = strchr_m(hexchars, toupper((unsigned char)strhex[i]))))
425 /* get the two nybbles */
426 hinybble = PTR_DIFF(p1, hexchars);
427 lonybble = PTR_DIFF(p2, hexchars);
429 p[num_chars] = (hinybble << 4) | lonybble;
438 DATA_BLOB strhex_to_data_blob(const char *strhex)
440 DATA_BLOB ret_blob = data_blob(NULL, strlen(strhex)/2+1);
442 ret_blob.length = strhex_to_str(ret_blob.data,
451 * Routine to print a buffer as HEX digits, into an allocated string.
453 void hex_encode(const unsigned char *buff_in, size_t len, char **out_hex_buffer)
458 *out_hex_buffer = smb_xmalloc((len*2)+1);
459 hex_buffer = *out_hex_buffer;
461 for (i = 0; i < len; i++)
462 slprintf(&hex_buffer[i*2], 3, "%02X", buff_in[i]);
466 Check if a string is part of a list.
468 BOOL in_list(const char *s, const char *list, BOOL casesensitive)
476 while (next_token(&p,tok,LIST_SEP,sizeof(tok))) {
478 if (strcmp(tok,s) == 0)
481 if (strcasecmp_m(tok,s) == 0)
489 Set a string value, allocing the space for the string
491 static BOOL string_init(char **dest,const char *src)
495 (*dest) = strdup(src);
496 if ((*dest) == NULL) {
497 DEBUG(0,("Out of memory in string_init\n"));
506 void string_free(char **s)
508 if (s) SAFE_FREE(*s);
512 Set a string value, deallocating any existing space, and allocing the space
515 BOOL string_set(char **dest, const char *src)
518 return string_init(dest,src);
522 Substitute a string for a pattern in another string. Make sure there is
525 This routine looks for pattern in s and replaces it with
526 insert. It may do multiple replacements.
528 Any of " ; ' $ or ` in the insert string are replaced with _
529 if len==0 then the string cannot be extended. This is different from the old
530 use of len==0 which was for no length checks to be done.
533 void string_sub(char *s,const char *pattern, const char *insert, size_t len)
538 if (!insert || !pattern || !*pattern || !s)
541 ls = (ssize_t)strlen(s);
542 lp = (ssize_t)strlen(pattern);
543 li = (ssize_t)strlen(insert);
546 len = ls + 1; /* len is number of *bytes* */
548 while (lp <= ls && (p = strstr(s,pattern))) {
549 if (ls + (li-lp) >= len) {
550 DEBUG(0,("ERROR: string overflow by %d in string_sub(%.50s, %d)\n",
551 (int)(ls + (li-lp) - len),
556 memmove(p+li,p+lp,strlen(p+lp)+1);
581 Similar to string_sub() but allows for any character to be substituted.
583 if len==0 then the string cannot be extended. This is different from the old
584 use of len==0 which was for no length checks to be done.
587 void all_string_sub(char *s,const char *pattern,const char *insert, size_t len)
592 if (!insert || !pattern || !s)
595 ls = (ssize_t)strlen(s);
596 lp = (ssize_t)strlen(pattern);
597 li = (ssize_t)strlen(insert);
603 len = ls + 1; /* len is number of *bytes* */
605 while (lp <= ls && (p = strstr(s,pattern))) {
606 if (ls + (li-lp) >= len) {
607 DEBUG(0,("ERROR: string overflow by %d in all_string_sub(%.50s, %d)\n",
608 (int)(ls + (li-lp) - len),
613 memmove(p+li,p+lp,strlen(p+lp)+1);
615 memcpy(p, insert, li);
623 Strchr and strrchr_m are a bit complex on general multi-byte strings.
625 char *strchr_m(const char *s, char c)
627 /* characters below 0x3F are guaranteed to not appear in
628 non-initial position in multi-byte charsets */
629 if ((c & 0xC0) == 0) {
635 codepoint_t c2 = next_codepoint(s, &size);
637 return discard_const(s);
645 char *strrchr_m(const char *s, char c)
649 /* characters below 0x3F are guaranteed to not appear in
650 non-initial position in multi-byte charsets */
651 if ((c & 0xC0) == 0) {
652 return strrchr(s, c);
657 codepoint_t c2 = next_codepoint(s, &size);
659 ret = discard_const(s);
668 return True if any (multi-byte) character is lower case
670 BOOL strhaslower(const char *string)
677 s = next_codepoint(string, &c_size);
683 return True; /* that means it has lower case chars */
691 return True if any (multi-byte) character is upper case
693 BOOL strhasupper(const char *string)
700 s = next_codepoint(string, &c_size);
706 return True; /* that means it has upper case chars */
714 Convert a string to lower case, allocated with talloc
716 char *strlower_talloc(TALLOC_CTX *ctx, const char *src)
721 /* this takes advantage of the fact that upper/lower can't
722 change the length of a character by more than 1 byte */
723 dest = talloc_size(ctx, 2*(strlen(src))+1);
730 codepoint_t c = next_codepoint(src, &c_size);
735 c_size = push_codepoint(dest+size, c);
749 Convert a string to UPPER case, allocated with talloc
751 char *strupper_talloc(TALLOC_CTX *ctx, const char *src)
760 /* this takes advantage of the fact that upper/lower can't
761 change the length of a character by more than 1 byte */
762 dest = talloc_size(ctx, 2*(strlen(src))+1);
769 codepoint_t c = next_codepoint(src, &c_size);
774 c_size = push_codepoint(dest+size, c);
788 Convert a string to lower case.
790 void strlower_m(char *s)
794 /* this is quite a common operation, so we want it to be
795 fast. We optimise for the ascii case, knowing that all our
796 supported multi-byte character sets are ascii-compatible
797 (ie. they match for the first 128 chars) */
798 while (*s && !(((uint8_t)s[0]) & 0x7F)) {
799 *s = tolower((uint8_t)*s);
809 size_t c_size, c_size2;
810 codepoint_t c = next_codepoint(s, &c_size);
811 c_size2 = push_codepoint(d, tolower_w(c));
812 if (c_size2 > c_size) {
813 DEBUG(0,("FATAL: codepoint 0x%x (0x%x) expanded from %d to %d bytes in strlower_m\n",
814 c, tolower_w(c), (int)c_size, (int)c_size2));
815 smb_panic("codepoint expansion in strlower_m\n");
824 Convert a string to UPPER case.
826 void strupper_m(char *s)
830 /* this is quite a common operation, so we want it to be
831 fast. We optimise for the ascii case, knowing that all our
832 supported multi-byte character sets are ascii-compatible
833 (ie. they match for the first 128 chars) */
834 while (*s && !(((uint8_t)s[0]) & 0x7F)) {
835 *s = toupper((uint8_t)*s);
845 size_t c_size, c_size2;
846 codepoint_t c = next_codepoint(s, &c_size);
847 c_size2 = push_codepoint(d, toupper_w(c));
848 if (c_size2 > c_size) {
849 DEBUG(0,("FATAL: codepoint 0x%x (0x%x) expanded from %d to %d bytes in strupper_m\n",
850 c, toupper_w(c), (int)c_size, (int)c_size2));
851 smb_panic("codepoint expansion in strupper_m\n");
860 Count the number of UCS2 characters in a string. Normally this will
861 be the same as the number of bytes in a string for single byte strings,
862 but will be different for multibyte.
864 size_t strlen_m(const char *s)
872 while (*s && !(((uint8_t)s[0]) & 0x7F)) {
883 codepoint_t c = next_codepoint(s, &c_size);
896 Work out the number of multibyte chars in a string, including the NULL
899 size_t strlen_m_term(const char *s)
905 return strlen_m(s) + 1;
909 Unescape a URL encoded string, in place.
912 void rfc1738_unescape(char *buf)
916 while ((p=strchr_m(p,'+')))
921 while (p && *p && (p=strchr_m(p,'%'))) {
925 if (c1 >= '0' && c1 <= '9')
927 else if (c1 >= 'A' && c1 <= 'F')
929 else if (c1 >= 'a' && c1 <= 'f')
931 else {p++; continue;}
933 if (c2 >= '0' && c2 <= '9')
935 else if (c2 >= 'A' && c2 <= 'F')
937 else if (c2 >= 'a' && c2 <= 'f')
939 else {p++; continue;}
943 memmove(p+1, p+3, strlen(p+3)+1);
949 * Decode a base64 string into a DATA_BLOB - simple and slow algorithm
951 DATA_BLOB base64_decode_data_blob(TALLOC_CTX *mem_ctx, const char *s)
953 DATA_BLOB ret = data_blob_talloc(mem_ctx, s, strlen(s)+1);
954 ret.length = ldb_base64_decode(ret.data);
959 * Decode a base64 string in-place - wrapper for the above
961 void base64_decode_inplace(char *s)
963 ldb_base64_decode(s);
967 * Encode a base64 string into a talloc()ed string caller to free.
969 char *base64_encode_data_blob(TALLOC_CTX *mem_ctx, DATA_BLOB data)
971 return ldb_base64_encode(mem_ctx, data.data, data.length);
975 size_t valgrind_strlen(const char *s)
978 for(count = 0; *s++; count++)
986 format a string into length-prefixed dotted domain format, as used in NBT
987 and in some ADS structures
989 const char *str_format_nbt_domain(TALLOC_CTX *mem_ctx, const char *s)
994 return talloc_strdup(mem_ctx, "");
996 ret = talloc_size(mem_ctx, strlen(s)+2);
1001 memcpy(ret+1, s, strlen(s)+1);
1004 for (i=0;ret[i];i++) {
1005 if (ret[i] == '.') {
1006 char *p = strchr(ret+i+1, '.');
1008 ret[i] = p-(ret+i+1);
1010 ret[i] = strlen(ret+i+1);
1018 BOOL add_string_to_array(TALLOC_CTX *mem_ctx,
1019 const char *str, const char ***strings, int *num)
1021 char *dup_str = talloc_strdup(mem_ctx, str);
1023 *strings = talloc_realloc(mem_ctx,
1025 const char *, ((*num)+1));
1027 if ((*strings == NULL) || (dup_str == NULL))
1030 (*strings)[*num] = dup_str;
1039 varient of strcmp() that handles NULL ptrs
1041 int strcmp_safe(const char *s1, const char *s2)
1046 if (s1 == NULL || s2 == NULL) {
1049 return strcmp(s1, s2);
1053 /*******************************************************************
1054 return the number of bytes occupied by a buffer in ASCII format
1055 the result includes the null termination
1056 limited by 'n' bytes
1057 ********************************************************************/
1058 size_t ascii_len_n(const char *src, size_t n)
1062 len = strnlen(src, n);
1071 /*******************************************************************
1072 Return a string representing a CIFS attribute for a file.
1073 ********************************************************************/
1074 char *attrib_string(TALLOC_CTX *mem_ctx, uint32_t attrib)
1081 {'V', FILE_ATTRIBUTE_VOLUME},
1082 {'D', FILE_ATTRIBUTE_DIRECTORY},
1083 {'A', FILE_ATTRIBUTE_ARCHIVE},
1084 {'H', FILE_ATTRIBUTE_HIDDEN},
1085 {'S', FILE_ATTRIBUTE_SYSTEM},
1086 {'N', FILE_ATTRIBUTE_NORMAL},
1087 {'R', FILE_ATTRIBUTE_READONLY},
1088 {'d', FILE_ATTRIBUTE_DEVICE},
1089 {'t', FILE_ATTRIBUTE_TEMPORARY},
1090 {'s', FILE_ATTRIBUTE_SPARSE},
1091 {'r', FILE_ATTRIBUTE_REPARSE_POINT},
1092 {'c', FILE_ATTRIBUTE_COMPRESSED},
1093 {'o', FILE_ATTRIBUTE_OFFLINE},
1094 {'n', FILE_ATTRIBUTE_NONINDEXED},
1095 {'e', FILE_ATTRIBUTE_ENCRYPTED}
1099 ret = talloc_size(mem_ctx, ARRAY_SIZE(attr_strs)+1);
1104 for (len=i=0; i<ARRAY_SIZE(attr_strs); i++) {
1105 if (attrib & attr_strs[i].attr) {
1106 ret[len++] = attr_strs[i].c;