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"
29 * @brief String utilities.
33 * Get the next token from a string, return False if none found.
34 * Handles double-quotes.
36 * Based on a routine by GJC@VILLAGE.COM.
37 * Extensively modified by Andrew.Tridgell@anu.edu.au
39 BOOL next_token(const char **ptr,char *buff, const char *sep, size_t bufsize)
50 /* default to simple separators */
54 /* find the first non sep char */
55 while (*s && strchr_m(sep,*s))
62 /* copy over the token */
63 for (quoted = False; len < bufsize && *s && (quoted || !strchr_m(sep,*s)); s++) {
72 *ptr = (*s) ? s+1 : s;
79 Case insensitive string compararison
81 int StrCaseCmp(const char *s1, const char *s2)
83 codepoint_t c1=0, c2=0;
87 c1 = next_codepoint(s1, &size1);
88 c2 = next_codepoint(s2, &size2);
97 if (c1 == INVALID_CODEPOINT ||
98 c2 == INVALID_CODEPOINT) {
99 /* what else can we do?? */
103 if (toupper_w(c1) != toupper_w(c2)) {
114 * @note The comparison is case-insensitive.
116 BOOL strequal(const char *s1, const char *s2)
123 return StrCaseCmp(s1,s2) == 0;
127 Compare 2 strings (case sensitive).
129 BOOL strcsequal(const char *s1,const char *s2)
136 return strcmp(s1,s2) == 0;
141 Do a case-insensitive, whitespace-ignoring string compare.
143 int strwicmp(const char *psz1, const char *psz2)
145 /* if BOTH strings are NULL, return TRUE, if ONE is NULL return */
146 /* appropriate value. */
149 else if (psz1 == NULL)
151 else if (psz2 == NULL)
154 /* sync the strings on first non-whitespace */
156 while (isspace((int)*psz1))
158 while (isspace((int)*psz2))
160 if (toupper(*psz1) != toupper(*psz2) || *psz1 == '\0'
166 return (*psz1 - *psz2);
171 NOTE: oldc and newc must be 7 bit characters
173 void string_replace(char *s, char oldc, char newc)
177 codepoint_t c = next_codepoint(s, &size);
186 Trim the specified elements off the front and back of a string.
188 BOOL trim_string(char *s,const char *front,const char *back)
195 /* Ignore null or empty strings. */
196 if (!s || (s[0] == '\0'))
199 front_len = front? strlen(front) : 0;
200 back_len = back? strlen(back) : 0;
205 while (len && strncmp(s, front, front_len)==0) {
206 /* Must use memmove here as src & dest can
207 * easily overlap. Found by valgrind. JRA. */
208 memmove(s, s+front_len, (len-front_len)+1);
215 while ((len >= back_len) && strncmp(s+len-back_len,back,back_len)==0) {
216 s[len-back_len]='\0';
225 Find the number of 'c' chars in a string
227 size_t count_chars(const char *s, char c)
233 codepoint_t c2 = next_codepoint(s, &size);
234 if (c2 == c) count++;
242 Safe string copy into a known length string. maxlength does not
243 include the terminating zero.
245 char *safe_strcpy(char *dest,const char *src, size_t maxlength)
250 DEBUG(0,("ERROR: NULL dest in safe_strcpy\n"));
255 /* We intentionally write out at the extremity of the destination
256 * string. If the destination is too short (e.g. pstrcpy into mallocd
257 * or fstring) then this should cause an error under a memory
259 dest[maxlength] = '\0';
260 if (PTR_DIFF(&len, dest) > 0) { /* check if destination is on the stack, ok if so */
261 log_suspicious_usage("safe_strcpy", src);
272 if (len > maxlength) {
273 DEBUG(0,("ERROR: string overflow by %u (%u - %u) in safe_strcpy [%.50s]\n",
274 (uint_t)(len-maxlength), len, maxlength, src));
278 memmove(dest, src, len);
284 Safe string cat into a string. maxlength does not
285 include the terminating zero.
287 char *safe_strcat(char *dest, const char *src, size_t maxlength)
289 size_t src_len, dest_len;
292 DEBUG(0,("ERROR: NULL dest in safe_strcat\n"));
300 if (PTR_DIFF(&src_len, dest) > 0) { /* check if destination is on the stack, ok if so */
301 log_suspicious_usage("safe_strcat", src);
304 src_len = strlen(src);
305 dest_len = strlen(dest);
307 if (src_len + dest_len > maxlength) {
308 DEBUG(0,("ERROR: string overflow by %d in safe_strcat [%.50s]\n",
309 (int)(src_len + dest_len - maxlength), src));
310 if (maxlength > dest_len) {
311 memcpy(&dest[dest_len], src, maxlength - dest_len);
317 memcpy(&dest[dest_len], src, src_len);
318 dest[dest_len + src_len] = 0;
323 Paranoid strcpy into a buffer of given length (includes terminating
324 zero. Strips out all but 'a-Z0-9' and the character in other_safe_chars
325 and replaces with '_'. Deliberately does *NOT* check for multibyte
326 characters. Don't change it !
329 char *alpha_strcpy(char *dest, const char *src, const char *other_safe_chars, size_t maxlength)
333 if (maxlength == 0) {
334 /* can't fit any bytes at all! */
339 DEBUG(0,("ERROR: NULL dest in alpha_strcpy\n"));
349 if (len >= maxlength)
352 if (!other_safe_chars)
353 other_safe_chars = "";
355 for(i = 0; i < len; i++) {
356 int val = (src[i] & 0xff);
357 if (isupper(val) || islower(val) || isdigit(val) || strchr_m(other_safe_chars, val))
369 Like strncpy but always null terminates. Make sure there is room!
370 The variable n should always be one less than the available size.
373 char *StrnCpy(char *dest,const char *src,size_t n)
382 while (n-- && (*d++ = *src++))
390 Routine to get hex characters and turn them into a 16 byte array.
391 the array can be variable length, and any non-hex-numeric
392 characters are skipped. "0xnn" or "0Xnn" is specially catered
395 valid examples: "0A5D15"; "0x15, 0x49, 0xa2"; "59\ta9\te3\n"
398 size_t strhex_to_str(char *p, size_t len, const char *strhex)
401 size_t num_chars = 0;
402 uint8_t lonybble, hinybble;
403 const char *hexchars = "0123456789ABCDEF";
404 char *p1 = NULL, *p2 = NULL;
406 for (i = 0; i < len && strhex[i] != 0; i++) {
407 if (strncasecmp(hexchars, "0x", 2) == 0) {
408 i++; /* skip two chars */
412 if (!(p1 = strchr_m(hexchars, toupper(strhex[i]))))
415 i++; /* next hex digit */
417 if (!(p2 = strchr_m(hexchars, toupper(strhex[i]))))
420 /* get the two nybbles */
421 hinybble = PTR_DIFF(p1, hexchars);
422 lonybble = PTR_DIFF(p2, hexchars);
424 p[num_chars] = (hinybble << 4) | lonybble;
433 DATA_BLOB strhex_to_data_blob(const char *strhex)
435 DATA_BLOB ret_blob = data_blob(NULL, strlen(strhex)/2+1);
437 ret_blob.length = strhex_to_str(ret_blob.data,
446 * Routine to print a buffer as HEX digits, into an allocated string.
448 void hex_encode(const unsigned char *buff_in, size_t len, char **out_hex_buffer)
453 *out_hex_buffer = smb_xmalloc((len*2)+1);
454 hex_buffer = *out_hex_buffer;
456 for (i = 0; i < len; i++)
457 slprintf(&hex_buffer[i*2], 3, "%02X", buff_in[i]);
461 Check if a string is part of a list.
463 BOOL in_list(const char *s, const char *list, BOOL casesensitive)
471 while (next_token(&p,tok,LIST_SEP,sizeof(tok))) {
473 if (strcmp(tok,s) == 0)
476 if (StrCaseCmp(tok,s) == 0)
484 Set a string value, allocing the space for the string
486 static BOOL string_init(char **dest,const char *src)
490 (*dest) = strdup(src);
491 if ((*dest) == NULL) {
492 DEBUG(0,("Out of memory in string_init\n"));
501 void string_free(char **s)
503 if (s) SAFE_FREE(*s);
507 Set a string value, deallocating any existing space, and allocing the space
510 BOOL string_set(char **dest, const char *src)
513 return string_init(dest,src);
517 Substitute a string for a pattern in another string. Make sure there is
520 This routine looks for pattern in s and replaces it with
521 insert. It may do multiple replacements.
523 Any of " ; ' $ or ` in the insert string are replaced with _
524 if len==0 then the string cannot be extended. This is different from the old
525 use of len==0 which was for no length checks to be done.
528 void string_sub(char *s,const char *pattern, const char *insert, size_t len)
533 if (!insert || !pattern || !*pattern || !s)
536 ls = (ssize_t)strlen(s);
537 lp = (ssize_t)strlen(pattern);
538 li = (ssize_t)strlen(insert);
541 len = ls + 1; /* len is number of *bytes* */
543 while (lp <= ls && (p = strstr(s,pattern))) {
544 if (ls + (li-lp) >= len) {
545 DEBUG(0,("ERROR: string overflow by %d in string_sub(%.50s, %d)\n",
546 (int)(ls + (li-lp) - len),
551 memmove(p+li,p+lp,strlen(p+lp)+1);
576 Similar to string_sub() but allows for any character to be substituted.
578 if len==0 then the string cannot be extended. This is different from the old
579 use of len==0 which was for no length checks to be done.
582 void all_string_sub(char *s,const char *pattern,const char *insert, size_t len)
587 if (!insert || !pattern || !s)
590 ls = (ssize_t)strlen(s);
591 lp = (ssize_t)strlen(pattern);
592 li = (ssize_t)strlen(insert);
598 len = ls + 1; /* len is number of *bytes* */
600 while (lp <= ls && (p = strstr(s,pattern))) {
601 if (ls + (li-lp) >= len) {
602 DEBUG(0,("ERROR: string overflow by %d in all_string_sub(%.50s, %d)\n",
603 (int)(ls + (li-lp) - len),
608 memmove(p+li,p+lp,strlen(p+lp)+1);
610 memcpy(p, insert, li);
618 Strchr and strrchr_m are a bit complex on general multi-byte strings.
620 char *strchr_m(const char *s, char c)
622 /* characters below 0x3F are guaranteed to not appear in
623 non-initial position in multi-byte charsets */
624 if ((c & 0xC0) == 0) {
630 codepoint_t c2 = next_codepoint(s, &size);
632 return discard_const(s);
640 char *strrchr_m(const char *s, char c)
644 /* characters below 0x3F are guaranteed to not appear in
645 non-initial position in multi-byte charsets */
646 if ((c & 0xC0) == 0) {
647 return strrchr(s, c);
652 codepoint_t c2 = next_codepoint(s, &size);
654 ret = discard_const(s);
663 Convert a string to lower case, allocated with talloc
665 char *strlower_talloc(TALLOC_CTX *ctx, const char *src)
670 /* this takes advantage of the fact that upper/lower can't
671 change the length of a character by more than 1 byte */
672 dest = talloc_size(ctx, 2*(strlen(src))+1);
679 codepoint_t c = next_codepoint(src, &c_size);
684 c_size = push_codepoint(dest+size, c);
698 Convert a string to UPPER case, allocated with talloc
700 char *strupper_talloc(TALLOC_CTX *ctx, const char *src)
705 /* this takes advantage of the fact that upper/lower can't
706 change the length of a character by more than 1 byte */
707 dest = talloc_size(ctx, 2*(strlen(src))+1);
714 codepoint_t c = next_codepoint(src, &c_size);
719 c_size = push_codepoint(dest+size, c);
733 Convert a string to lower case.
735 void strlower_m(char *s)
739 /* this is quite a common operation, so we want it to be
740 fast. We optimise for the ascii case, knowing that all our
741 supported multi-byte character sets are ascii-compatible
742 (ie. they match for the first 128 chars) */
743 while (*s && !(((uint8_t)s[0]) & 0x7F)) {
744 *s = tolower((uint8_t)*s);
754 size_t c_size, c_size2;
755 codepoint_t c = next_codepoint(s, &c_size);
756 c_size2 = push_codepoint(d, tolower_w(c));
757 if (c_size2 > c_size) {
758 DEBUG(0,("FATAL: codepoint 0x%x (0x%x) expanded from %d to %d bytes in strlower_m\n",
759 c, tolower_w(c), c_size, c_size2));
760 smb_panic("codepoint expansion in strlower_m\n");
769 Convert a string to UPPER case.
771 void strupper_m(char *s)
775 /* this is quite a common operation, so we want it to be
776 fast. We optimise for the ascii case, knowing that all our
777 supported multi-byte character sets are ascii-compatible
778 (ie. they match for the first 128 chars) */
779 while (*s && !(((uint8_t)s[0]) & 0x7F)) {
780 *s = toupper((uint8_t)*s);
790 size_t c_size, c_size2;
791 codepoint_t c = next_codepoint(s, &c_size);
792 c_size2 = push_codepoint(d, toupper_w(c));
793 if (c_size2 > c_size) {
794 DEBUG(0,("FATAL: codepoint 0x%x (0x%x) expanded from %d to %d bytes in strupper_m\n",
795 c, toupper_w(c), c_size, c_size2));
796 smb_panic("codepoint expansion in strupper_m\n");
805 Count the number of UCS2 characters in a string. Normally this will
806 be the same as the number of bytes in a string for single byte strings,
807 but will be different for multibyte.
809 size_t strlen_m(const char *s)
817 while (*s && !(((uint8_t)s[0]) & 0x7F)) {
828 codepoint_t c = next_codepoint(s, &c_size);
841 Work out the number of multibyte chars in a string, including the NULL
844 size_t strlen_m_term(const char *s)
850 return strlen_m(s) + 1;
854 Return a RFC2254 binary string representation of a buffer.
855 Used in LDAP filters.
858 char *binary_string(char *buf, int len)
862 const char *hex = "0123456789ABCDEF";
863 s = malloc(len * 3 + 1);
866 for (j=i=0;i<len;i++) {
868 s[j+1] = hex[((uint8_t)buf[i]) >> 4];
869 s[j+2] = hex[((uint8_t)buf[i]) & 0xF];
877 Unescape a URL encoded string, in place.
880 void rfc1738_unescape(char *buf)
884 while ((p=strchr_m(p,'+')))
889 while (p && *p && (p=strchr_m(p,'%'))) {
893 if (c1 >= '0' && c1 <= '9')
895 else if (c1 >= 'A' && c1 <= 'F')
897 else if (c1 >= 'a' && c1 <= 'f')
899 else {p++; continue;}
901 if (c2 >= '0' && c2 <= '9')
903 else if (c2 >= 'A' && c2 <= 'F')
905 else if (c2 >= 'a' && c2 <= 'f')
907 else {p++; continue;}
911 memmove(p+1, p+3, strlen(p+3)+1);
917 * Decode a base64 string into a DATA_BLOB - simple and slow algorithm
919 DATA_BLOB base64_decode_data_blob(const char *s)
921 const char *b64 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
923 int bit_offset, byte_offset, idx, i, n;
924 DATA_BLOB decoded = data_blob(s, strlen(s)+1);
925 uint8_t *d = decoded.data;
930 while (*s && (p=strchr_m(b64,*s))) {
931 idx = (int)(p - b64);
932 byte_offset = (i*6)/8;
933 bit_offset = (i*6)%8;
934 d[byte_offset] &= ~((1<<(8-bit_offset))-1);
935 if (bit_offset < 3) {
936 d[byte_offset] |= (idx << (2-bit_offset));
939 d[byte_offset] |= (idx >> (bit_offset-2));
940 d[byte_offset+1] = 0;
941 d[byte_offset+1] |= (idx << (8-(bit_offset-2))) & 0xFF;
953 * Decode a base64 string in-place - wrapper for the above
955 void base64_decode_inplace(char *s)
957 DATA_BLOB decoded = base64_decode_data_blob(s);
958 memcpy(s, decoded.data, decoded.length);
959 data_blob_free(&decoded);
962 s[decoded.length] = '\0';
966 * Encode a base64 string into a malloc()ed string caller to free.
968 *From SQUID: adopted from http://ftp.sunet.se/pub2/gnu/vm/base64-encode.c with adjustments
970 char * base64_encode_data_blob(DATA_BLOB data)
972 const char *b64 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
976 size_t len = data.length;
977 size_t output_len = data.length * 2;
978 char *result = malloc(output_len); /* get us plenty of space */
980 while (len-- && out_cnt < (data.length * 2) - 5) {
981 int c = (uint8_t) *(data.data++);
984 if (char_count == 3) {
985 result[out_cnt++] = b64[bits >> 18];
986 result[out_cnt++] = b64[(bits >> 12) & 0x3f];
987 result[out_cnt++] = b64[(bits >> 6) & 0x3f];
988 result[out_cnt++] = b64[bits & 0x3f];
995 if (char_count != 0) {
996 bits <<= 16 - (8 * char_count);
997 result[out_cnt++] = b64[bits >> 18];
998 result[out_cnt++] = b64[(bits >> 12) & 0x3f];
999 if (char_count == 1) {
1000 result[out_cnt++] = '=';
1001 result[out_cnt++] = '=';
1003 result[out_cnt++] = b64[(bits >> 6) & 0x3f];
1004 result[out_cnt++] = '=';
1007 result[out_cnt] = '\0'; /* terminate */
1012 size_t valgrind_strlen(const char *s)
1015 for(count = 0; *s++; count++)
1023 format a string into length-prefixed dotted domain format, as used in NBT
1024 and in some ADS structures
1026 const char *str_format_nbt_domain(TALLOC_CTX *mem_ctx, const char *s)
1031 return talloc_strdup(mem_ctx, "");
1033 ret = talloc_size(mem_ctx, strlen(s)+2);
1038 memcpy(ret+1, s, strlen(s)+1);
1041 for (i=0;ret[i];i++) {
1042 if (ret[i] == '.') {
1043 char *p = strchr(ret+i+1, '.');
1045 ret[i] = p-(ret+i+1);
1047 ret[i] = strlen(ret+i+1);
1055 BOOL add_string_to_array(TALLOC_CTX *mem_ctx,
1056 const char *str, const char ***strings, int *num)
1058 char *dup_str = talloc_strdup(mem_ctx, str);
1060 *strings = talloc_realloc(mem_ctx,
1062 const char *, ((*num)+1));
1064 if ((*strings == NULL) || (dup_str == NULL))
1067 (*strings)[*num] = dup_str;
1076 varient of strcmp() that handles NULL ptrs
1078 int strcmp_safe(const char *s1, const char *s2)
1083 if (s1 == NULL || s2 == NULL) {
1086 return strcmp(s1, s2);
1090 /*******************************************************************
1091 return the number of bytes occupied by a buffer in ASCII format
1092 the result includes the null termination
1093 limited by 'n' bytes
1094 ********************************************************************/
1095 size_t ascii_len_n(const char *src, size_t n)
1099 len = strnlen(src, n);
1108 /*******************************************************************
1109 Return a string representing a CIFS attribute for a file.
1110 ********************************************************************/
1111 char *attrib_string(TALLOC_CTX *mem_ctx, uint32_t attrib)
1118 {'V', FILE_ATTRIBUTE_VOLUME},
1119 {'D', FILE_ATTRIBUTE_DIRECTORY},
1120 {'A', FILE_ATTRIBUTE_ARCHIVE},
1121 {'H', FILE_ATTRIBUTE_HIDDEN},
1122 {'S', FILE_ATTRIBUTE_SYSTEM},
1123 {'N', FILE_ATTRIBUTE_NORMAL},
1124 {'R', FILE_ATTRIBUTE_READONLY},
1125 {'d', FILE_ATTRIBUTE_DEVICE},
1126 {'t', FILE_ATTRIBUTE_TEMPORARY},
1127 {'s', FILE_ATTRIBUTE_SPARSE},
1128 {'r', FILE_ATTRIBUTE_REPARSE_POINT},
1129 {'c', FILE_ATTRIBUTE_COMPRESSED},
1130 {'o', FILE_ATTRIBUTE_OFFLINE},
1131 {'n', FILE_ATTRIBUTE_NONINDEXED},
1132 {'e', FILE_ATTRIBUTE_ENCRYPTED}
1136 ret = talloc_size(mem_ctx, ARRAY_SIZE(attr_strs)+1);
1141 for (len=i=0; i<ARRAY_SIZE(attr_strs); i++) {
1142 if (attrib & attr_strs[i].attr) {
1143 ret[len++] = attr_strs[i].c;