{
size_t ret, converted_size;
char *tmpbuf2 = NULL;
- if (!push_ascii_allocate(&tmpbuf2, s, &converted_size)) {
+ if (!push_ascii_talloc(NULL, &tmpbuf2, s, &converted_size)) {
return 0;
}
ret = strlen(tmpbuf2);
- SAFE_FREE(tmpbuf2);
+ TALLOC_FREE(tmpbuf2);
return ret;
}
return s;
}
-/**
- Strchr and strrchr_m are very hard to do on general multi-byte strings.
- We convert via ucs2 for now.
-**/
-
-char *strchr_m(const char *src, char c)
-{
- smb_ucs2_t *ws = NULL;
- char *s2 = NULL;
- smb_ucs2_t *p;
- const char *s;
- char *ret;
- size_t converted_size;
-
- /* characters below 0x3F are guaranteed to not appear in
- non-initial position in multi-byte charsets */
- if ((c & 0xC0) == 0) {
- return strchr(src, c);
- }
-
- /* this is quite a common operation, so we want it to be
- fast. We optimise for the ascii case, knowing that all our
- supported multi-byte character sets are ascii-compatible
- (ie. they match for the first 128 chars) */
-
- for (s = src; *s && !(((unsigned char)s[0]) & 0x80); s++) {
- if (*s == c)
- return (char *)s;
- }
-
- if (!*s)
- return NULL;
-
-#ifdef BROKEN_UNICODE_COMPOSE_CHARACTERS
- /* With compose characters we must restart from the beginning. JRA. */
- s = src;
-#endif
-
- if (!push_ucs2_allocate(&ws, s, &converted_size)) {
- /* Wrong answer, but what can we do... */
- return strchr(src, c);
- }
- p = strchr_w(ws, UCS2_CHAR(c));
- if (!p) {
- SAFE_FREE(ws);
- return NULL;
- }
- *p = 0;
- if (!pull_ucs2_allocate(&s2, ws, &converted_size)) {
- SAFE_FREE(ws);
- /* Wrong answer, but what can we do... */
- return strchr(src, c);
- }
- ret = (char *)(s+strlen(s2));
- SAFE_FREE(ws);
- SAFE_FREE(s2);
- return ret;
-}
-
-char *strrchr_m(const char *s, char c)
-{
- /* characters below 0x3F are guaranteed to not appear in
- non-initial position in multi-byte charsets */
- if ((c & 0xC0) == 0) {
- return strrchr(s, c);
- }
-
- /* this is quite a common operation, so we want it to be
- fast. We optimise for the ascii case, knowing that all our
- supported multi-byte character sets are ascii-compatible
- (ie. they match for the first 128 chars). Also, in Samba
- we only search for ascii characters in 'c' and that
- in all mb character sets with a compound character
- containing c, if 'c' is not a match at position
- p, then p[-1] > 0x7f. JRA. */
-
- {
- size_t len = strlen(s);
- const char *cp = s;
- bool got_mb = false;
-
- if (len == 0)
- return NULL;
- cp += (len - 1);
- do {
- if (c == *cp) {
- /* Could be a match. Part of a multibyte ? */
- if ((cp > s) &&
- (((unsigned char)cp[-1]) & 0x80)) {
- /* Yep - go slow :-( */
- got_mb = true;
- break;
- }
- /* No - we have a match ! */
- return (char *)cp;
- }
- } while (cp-- != s);
- if (!got_mb)
- return NULL;
- }
-
- /* String contained a non-ascii char. Slow path. */
- {
- smb_ucs2_t *ws = NULL;
- char *s2 = NULL;
- smb_ucs2_t *p;
- char *ret;
- size_t converted_size;
-
- if (!push_ucs2_allocate(&ws, s, &converted_size)) {
- /* Wrong answer, but what can we do. */
- return strrchr(s, c);
- }
- p = strrchr_w(ws, UCS2_CHAR(c));
- if (!p) {
- SAFE_FREE(ws);
- return NULL;
- }
- *p = 0;
- if (!pull_ucs2_allocate(&s2, ws, &converted_size)) {
- SAFE_FREE(ws);
- /* Wrong answer, but what can we do. */
- return strrchr(s, c);
- }
- ret = (char *)(s+strlen(s2));
- SAFE_FREE(ws);
- SAFE_FREE(s2);
- return ret;
- }
-}
-
/***********************************************************************
Return the equivalent of doing strrchr 'n' times - always going
backwards.
return retp;
}
-/**
- Convert a string to lower case.
-**/
-
-void strlower_m(char *s)
-{
- size_t len;
- int errno_save;
-
- /* this is quite a common operation, so we want it to be
- fast. We optimise for the ascii case, knowing that all our
- supported multi-byte character sets are ascii-compatible
- (ie. they match for the first 128 chars) */
-
- while (*s && !(((unsigned char)s[0]) & 0x80)) {
- *s = tolower_ascii((unsigned char)*s);
- s++;
- }
-
- if (!*s)
- return;
-
- /* I assume that lowercased string takes the same number of bytes
- * as source string even in UTF-8 encoding. (VIV) */
- len = strlen(s) + 1;
- errno_save = errno;
- errno = 0;
- unix_strlower(s,len,s,len);
- /* Catch mb conversion errors that may not terminate. */
- if (errno)
- s[len-1] = '\0';
- errno = errno_save;
-}
-
-/**
- Convert a string to upper case.
-**/
-
-void strupper_m(char *s)
-{
- size_t len;
- int errno_save;
-
- /* this is quite a common operation, so we want it to be
- fast. We optimise for the ascii case, knowing that all our
- supported multi-byte character sets are ascii-compatible
- (ie. they match for the first 128 chars) */
-
- while (*s && !(((unsigned char)s[0]) & 0x80)) {
- *s = toupper_ascii_fast((unsigned char)*s);
- s++;
- }
-
- if (!*s)
- return;
-
- /* I assume that lowercased string takes the same number of bytes
- * as source string even in multibyte encoding. (VIV) */
- len = strlen(s) + 1;
- errno_save = errno;
- errno = 0;
- unix_strupper(s,len,s,len);
- /* Catch mb conversion errors that may not terminate. */
- if (errno)
- s[len-1] = '\0';
- errno = errno_save;
-}
-
-/**
- Count the number of UCS2 characters in a string. Normally this will
- be the same as the number of bytes in a string for single byte strings,
- but will be different for multibyte.
-**/
-
-size_t strlen_m(const char *s)
-{
- size_t count = 0;
-
- if (!s) {
- return 0;
- }
-
- while (*s && !(((uint8_t)*s) & 0x80)) {
- s++;
- count++;
- }
-
- if (!*s) {
- return count;
- }
-
- while (*s) {
- size_t c_size;
- codepoint_t c = next_codepoint(s, &c_size);
- if (c < 0x10000) {
- /* Unicode char fits into 16 bits. */
- count += 1;
- } else {
- /* Double-width unicode char - 32 bits. */
- count += 2;
- }
- s += c_size;
- }
-
- return count;
-}
-
-/**
- Count the number of UCS2 characters in a string including the null
- terminator.
-**/
-
-size_t strlen_m_term(const char *s)
-{
- if (!s) {
- return 0;
- }
- return strlen_m(s) + 1;
-}
-
-/*
- * Weird helper routine for the winreg pipe: If nothing is around, return 0,
- * if a string is there, include the terminator.
- */
-
-size_t strlen_m_term_null(const char *s)
-{
- size_t len;
- if (!s) {
- return 0;
- }
- len = strlen_m(s);
- if (len == 0) {
- return 0;
- }
-
- return len+1;
-}
/**
Return a RFC2254 binary string representation of a buffer.
Used in LDAP filters.
git.samba.org / abartlet / samba.git / .git / blobdiff