/* Unix SMB/CIFS implementation. Samba utility functions Copyright (C) Andrew Tridgell 1992-2001 Copyright (C) Simo Sorce 2001-2002 Copyright (C) Martin Pool 2003 This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #include "includes.h" /** * @file * @brief String utilities. **/ /** * Get the next token from a string, return False if none found. * Handles double-quotes. * * Based on a routine by GJC@VILLAGE.COM. * Extensively modified by Andrew.Tridgell@anu.edu.au **/ BOOL next_token(const char **ptr,char *buff, const char *sep, size_t bufsize) { char *s; char *pbuf; BOOL quoted; size_t len=1; if (!ptr) return(False); s = (char *)*ptr; /* default to simple separators */ if (!sep) sep = " \t\n\r"; /* find the first non sep char */ while (*s && strchr_m(sep,*s)) s++; /* nothing left? */ if (! *s) return(False); /* copy over the token */ pbuf = buff; for (quoted = False; len < bufsize && *s && (quoted || !strchr_m(sep,*s)); s++) { if (*s == '\"' || *s == '\'') { quoted = !quoted; } else { len++; *pbuf++ = *s; } } *ptr = (*s) ? s+1 : s; *pbuf = 0; return(True); } /** This is like next_token but is not re-entrant and "remembers" the first parameter so you can pass NULL. This is useful for user interface code but beware the fact that it is not re-entrant! **/ static const char *last_ptr=NULL; BOOL next_token_nr(const char **ptr,char *buff, const char *sep, size_t bufsize) { BOOL ret; if (!ptr) ptr = &last_ptr; ret = next_token(ptr, buff, sep, bufsize); last_ptr = *ptr; return ret; } static uint16 tmpbuf[sizeof(pstring)]; void set_first_token(char *ptr) { last_ptr = ptr; } /** Convert list of tokens to array; dependent on above routine. Uses last_ptr from above - bit of a hack. **/ char **toktocliplist(int *ctok, const char *sep) { char *s=(char *)last_ptr; int ictok=0; char **ret, **iret; if (!sep) sep = " \t\n\r"; while(*s && strchr_m(sep,*s)) s++; /* nothing left? */ if (!*s) return(NULL); do { ictok++; while(*s && (!strchr_m(sep,*s))) s++; while(*s && strchr_m(sep,*s)) *s++=0; } while(*s); *ctok=ictok; s=(char *)last_ptr; if (!(ret=iret=malloc(ictok*sizeof(char *)))) return NULL; while(ictok--) { *iret++=s; while(*s++) ; while(!*s) s++; } return ret; } /** * Case insensitive string compararison. * * iconv does not directly give us a way to compare strings in * arbitrary unix character sets -- all we can is convert and then * compare. This is expensive. * * As an optimization, we do a first pass that considers only the * prefix of the strings that is entirely 7-bit. Within this, we * check whether they have the same value. * * Hopefully this will often give the answer without needing to copy. * In particular it should speed comparisons to literal ascii strings * or comparisons of strings that are "obviously" different. * * If we find a non-ascii character we fall back to converting via * iconv. * * This should never be slower than convering the whole thing, and * often faster. * * A different optimization would be to compare for bitwise equality * in the binary encoding. (It would be possible thought hairy to do * both simultaneously.) But in that case if they turn out to be * different, we'd need to restart the whole thing. * * Even better is to implement strcasecmp for each encoding and use a * function pointer. **/ int StrCaseCmp(const char *s, const char *t) { const char * ps, * pt; size_t size; smb_ucs2_t *buffer_s, *buffer_t; int ret; for (ps = s, pt = t; ; ps++, pt++) { char us, ut; if (!*ps && !*pt) return 0; /* both ended */ else if (!*ps) return -1; /* s is a prefix */ else if (!*pt) return +1; /* t is a prefix */ else if ((*ps & 0x80) || (*pt & 0x80)) /* not ascii anymore, do it the hard way from here on in */ break; us = toupper(*ps); ut = toupper(*pt); if (us == ut) continue; else if (us < ut) return -1; else if (us > ut) return +1; } size = push_ucs2_allocate(&buffer_s, s); if (size == (size_t)-1) { return strcmp(s, t); /* Not quite the right answer, but finding the right one under this failure case is expensive, and it's pretty close */ } size = push_ucs2_allocate(&buffer_t, t); if (size == (size_t)-1) { SAFE_FREE(buffer_s); return strcmp(s, t); /* Not quite the right answer, but finding the right one under this failure case is expensive, and it's pretty close */ } ret = strcasecmp_w(buffer_s, buffer_t); SAFE_FREE(buffer_s); SAFE_FREE(buffer_t); return ret; } /** Case insensitive string compararison, length limited. **/ int StrnCaseCmp(const char *s, const char *t, size_t n) { pstring buf1, buf2; unix_strupper(s, strlen(s)+1, buf1, sizeof(buf1)); unix_strupper(t, strlen(t)+1, buf2, sizeof(buf2)); return strncmp(buf1,buf2,n); } /** * Compare 2 strings. * * @note The comparison is case-insensitive. **/ BOOL strequal(const char *s1, const char *s2) { if (s1 == s2) return(True); if (!s1 || !s2) return(False); return(StrCaseCmp(s1,s2)==0); } /** * Compare 2 strings up to and including the nth char. * * @note The comparison is case-insensitive. **/ BOOL strnequal(const char *s1,const char *s2,size_t n) { if (s1 == s2) return(True); if (!s1 || !s2 || !n) return(False); return(StrnCaseCmp(s1,s2,n)==0); } /** Compare 2 strings (case sensitive). **/ BOOL strcsequal(const char *s1,const char *s2) { if (s1 == s2) return(True); if (!s1 || !s2) return(False); return(strcmp(s1,s2)==0); } /** Do a case-insensitive, whitespace-ignoring string compare. **/ int strwicmp(const char *psz1, const char *psz2) { /* if BOTH strings are NULL, return TRUE, if ONE is NULL return */ /* appropriate value. */ if (psz1 == psz2) return (0); else if (psz1 == NULL) return (-1); else if (psz2 == NULL) return (1); /* sync the strings on first non-whitespace */ while (1) { while (isspace((int)*psz1)) psz1++; while (isspace((int)*psz2)) psz2++; if (toupper(*psz1) != toupper(*psz2) || *psz1 == '\0' || *psz2 == '\0') break; psz1++; psz2++; } return (*psz1 - *psz2); } /** Convert a string to upper case, but don't modify it. **/ char *strupper_static(const char *s) { static pstring str; pstrcpy(str, s); strupper_m(str); return str; } /** Convert a string to "normal" form. **/ void strnorm(char *s, int case_default) { if (case_default == CASE_UPPER) strupper_m(s); else strlower_m(s); } /** Check if a string is in "normal" case. **/ BOOL strisnormal(const char *s, int case_default) { if (case_default == CASE_UPPER) return(!strhaslower(s)); return(!strhasupper(s)); } /** String replace. NOTE: oldc and newc must be 7 bit characters **/ void string_replace(pstring s,char oldc,char newc) { unsigned char *p; /* 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 (p = (unsigned char *)s; *p; p++) { if (*p & 0x80) /* mb string - slow path. */ break; if (*p == oldc) *p = newc; } if (!*p) return; /* Slow (mb) path. */ #ifdef BROKEN_UNICODE_COMPOSE_CHARACTERS /* With compose characters we must restart from the beginning. JRA. */ p = s; #endif push_ucs2(NULL, tmpbuf, p, sizeof(tmpbuf), STR_TERMINATE); string_replace_w(tmpbuf, UCS2_CHAR(oldc), UCS2_CHAR(newc)); pull_ucs2(NULL, p, tmpbuf, -1, sizeof(tmpbuf), STR_TERMINATE); } /** Skip past some strings in a buffer. **/ char *skip_string(char *buf,size_t n) { while (n--) buf += strlen(buf) + 1; return(buf); } /** Count the number of 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 str_charnum(const char *s) { uint16 tmpbuf2[sizeof(pstring)]; push_ucs2(NULL, tmpbuf2,s, sizeof(tmpbuf2), STR_TERMINATE); return strlen_w(tmpbuf2); } /** Count the number of 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 str_ascii_charnum(const char *s) { pstring tmpbuf2; push_ascii(tmpbuf2, s, sizeof(tmpbuf2), STR_TERMINATE); return strlen(tmpbuf2); } BOOL trim_char(char *s,char cfront,char cback) { BOOL ret = False; char *ep; char *fp = s; /* Ignore null or empty strings. */ if (!s || (s[0] == '\0')) return False; if (cfront) { while (*fp && *fp == cfront) fp++; if (!*fp) { /* We ate the string. */ s[0] = '\0'; return True; } if (fp != s) ret = True; } ep = fp + strlen(fp) - 1; if (cback) { /* Attempt ascii only. Bail for mb strings. */ while ((ep >= fp) && (*ep == cback)) { ret = True; if ((ep > fp) && (((unsigned char)ep[-1]) & 0x80)) { /* Could be mb... bail back to tim_string. */ char fs[2], bs[2]; if (cfront) { fs[0] = cfront; fs[1] = '\0'; } bs[0] = cback; bs[1] = '\0'; return trim_string(s, cfront ? fs : NULL, bs); } else { ep--; } } if (ep < fp) { /* We ate the string. */ s[0] = '\0'; return True; } } ep[1] = '\0'; memmove(s, fp, ep-fp+2); return ret; } /** Trim the specified elements off the front and back of a string. **/ BOOL trim_string(char *s,const char *front,const char *back) { BOOL ret = False; size_t front_len; size_t back_len; size_t len; /* Ignore null or empty strings. */ if (!s || (s[0] == '\0')) return False; front_len = front? strlen(front) : 0; back_len = back? strlen(back) : 0; len = strlen(s); if (front_len) { while (len && strncmp(s, front, front_len)==0) { /* Must use memmove here as src & dest can * easily overlap. Found by valgrind. JRA. */ memmove(s, s+front_len, (len-front_len)+1); len -= front_len; ret=True; } } if (back_len) { while ((len >= back_len) && strncmp(s+len-back_len,back,back_len)==0) { s[len-back_len]='\0'; len -= back_len; ret=True; } } return ret; } /** Does a string have any uppercase chars in it? **/ BOOL strhasupper(const char *s) { smb_ucs2_t *ptr; push_ucs2(NULL, tmpbuf,s, sizeof(tmpbuf), STR_TERMINATE); for(ptr=tmpbuf;*ptr;ptr++) if(isupper_w(*ptr)) return True; return(False); } /** Does a string have any lowercase chars in it? **/ BOOL strhaslower(const char *s) { smb_ucs2_t *ptr; push_ucs2(NULL, tmpbuf,s, sizeof(tmpbuf), STR_TERMINATE); for(ptr=tmpbuf;*ptr;ptr++) if(islower_w(*ptr)) return True; return(False); } /** Find the number of 'c' chars in a string **/ size_t count_chars(const char *s,char c) { smb_ucs2_t *ptr; int count; smb_ucs2_t *alloc_tmpbuf = NULL; if (push_ucs2_allocate(&alloc_tmpbuf, s) == (size_t)-1) { return 0; } for(count=0,ptr=alloc_tmpbuf;*ptr;ptr++) if(*ptr==UCS2_CHAR(c)) count++; SAFE_FREE(alloc_tmpbuf); return(count); } /** Safe string copy into a known length string. maxlength does not include the terminating zero. **/ char *safe_strcpy_fn(const char *fn, int line, char *dest,const char *src, size_t maxlength) { size_t len; if (!dest) { DEBUG(0,("ERROR: NULL dest in safe_strcpy, called from [%s][%d]\n", fn, line)); return NULL; } #ifdef DEVELOPER clobber_region(fn,line,dest, maxlength+1); #endif if (!src) { *dest = 0; return dest; } len = strnlen(src, maxlength+1); if (len > maxlength) { DEBUG(0,("ERROR: string overflow by %lu (%lu - %lu) in safe_strcpy [%.50s]\n", (unsigned long)(len-maxlength), (unsigned long)len, (unsigned long)maxlength, src)); len = maxlength; } memmove(dest, src, len); dest[len] = 0; return dest; } /** Safe string cat into a string. maxlength does not include the terminating zero. **/ char *safe_strcat_fn(const char *fn, int line, char *dest, const char *src, size_t maxlength) { size_t src_len, dest_len; if (!dest) { DEBUG(0,("ERROR: NULL dest in safe_strcat, called from [%s][%d]\n", fn, line)); return NULL; } if (!src) return dest; src_len = strnlen(src, maxlength + 1); dest_len = strnlen(dest, maxlength + 1); #ifdef DEVELOPER clobber_region(fn, line, dest + dest_len, maxlength + 1 - dest_len); #endif if (src_len + dest_len > maxlength) { DEBUG(0,("ERROR: string overflow by %d in safe_strcat [%.50s]\n", (int)(src_len + dest_len - maxlength), src)); if (maxlength > dest_len) { memcpy(&dest[dest_len], src, maxlength - dest_len); } dest[maxlength] = 0; return NULL; } memcpy(&dest[dest_len], src, src_len); dest[dest_len + src_len] = 0; return dest; } /** Paranoid strcpy into a buffer of given length (includes terminating zero. Strips out all but 'a-Z0-9' and the character in other_safe_chars and replaces with '_'. Deliberately does *NOT* check for multibyte characters. Don't change it ! **/ char *alpha_strcpy_fn(const char *fn, int line, char *dest, const char *src, const char *other_safe_chars, size_t maxlength) { size_t len, i; #ifdef DEVELOPER clobber_region(fn, line, dest, maxlength); #endif if (!dest) { DEBUG(0,("ERROR: NULL dest in alpha_strcpy, called from [%s][%d]\n", fn, line)); return NULL; } if (!src) { *dest = 0; return dest; } len = strlen(src); if (len >= maxlength) len = maxlength - 1; if (!other_safe_chars) other_safe_chars = ""; for(i = 0; i < len; i++) { int val = (src[i] & 0xff); if (isupper(val) || islower(val) || isdigit(val) || strchr_m(other_safe_chars, val)) dest[i] = src[i]; else dest[i] = '_'; } dest[i] = '\0'; return dest; } /** Like strncpy but always null terminates. Make sure there is room! The variable n should always be one less than the available size. **/ char *StrnCpy_fn(const char *fn, int line,char *dest,const char *src,size_t n) { char *d = dest; #ifdef DEVELOPER clobber_region(fn, line, dest, n+1); #endif if (!dest) { DEBUG(0,("ERROR: NULL dest in StrnCpy, called from [%s][%d]\n", fn, line)); return(NULL); } if (!src) { *dest = 0; return(dest); } while (n-- && (*d = *src)) { d++; src++; } *d = 0; return(dest); } #if 0 /** Like strncpy but copies up to the character marker. always null terminates. returns a pointer to the character marker in the source string (src). **/ static char *strncpyn(char *dest, const char *src, size_t n, char c) { char *p; size_t str_len; #ifdef DEVELOPER clobber_region(dest, n+1); #endif p = strchr_m(src, c); if (p == NULL) { DEBUG(5, ("strncpyn: separator character (%c) not found\n", c)); return NULL; } str_len = PTR_DIFF(p, src); strncpy(dest, src, MIN(n, str_len)); dest[str_len] = '\0'; return p; } #endif /** Routine to get hex characters and turn them into a 16 byte array. the array can be variable length, and any non-hex-numeric characters are skipped. "0xnn" or "0Xnn" is specially catered for. valid examples: "0A5D15"; "0x15, 0x49, 0xa2"; "59\ta9\te3\n" **/ size_t strhex_to_str(char *p, size_t len, const char *strhex) { size_t i; size_t num_chars = 0; unsigned char lonybble, hinybble; const char *hexchars = "0123456789ABCDEF"; char *p1 = NULL, *p2 = NULL; for (i = 0; i < len && strhex[i] != 0; i++) { if (strnequal(hexchars, "0x", 2)) { i++; /* skip two chars */ continue; } if (!(p1 = strchr_m(hexchars, toupper(strhex[i])))) break; i++; /* next hex digit */ if (!(p2 = strchr_m(hexchars, toupper(strhex[i])))) break; /* get the two nybbles */ hinybble = PTR_DIFF(p1, hexchars); lonybble = PTR_DIFF(p2, hexchars); p[num_chars] = (hinybble << 4) | lonybble; num_chars++; p1 = NULL; p2 = NULL; } return num_chars; } DATA_BLOB strhex_to_data_blob(const char *strhex) { DATA_BLOB ret_blob = data_blob(NULL, strlen(strhex)/2+1); ret_blob.length = strhex_to_str(ret_blob.data, strlen(strhex), strhex); return ret_blob; } /** * Routine to print a buffer as HEX digits, into an allocated string. */ void hex_encode(const unsigned char *buff_in, size_t len, char **out_hex_buffer) { int i; char *hex_buffer; *out_hex_buffer = smb_xmalloc((len*2)+1); hex_buffer = *out_hex_buffer; for (i = 0; i < len; i++) slprintf(&hex_buffer[i*2], 3, "%02X", buff_in[i]); } /** Check if a string is part of a list. **/ BOOL in_list(char *s,char *list,BOOL casesensitive) { pstring tok; const char *p=list; if (!list) return(False); while (next_token(&p,tok,LIST_SEP,sizeof(tok))) { if (casesensitive) { if (strcmp(tok,s) == 0) return(True); } else { if (StrCaseCmp(tok,s) == 0) return(True); } } return(False); } /* this is used to prevent lots of mallocs of size 1 */ static char *null_string = NULL; /** Set a string value, allocing the space for the string **/ static BOOL string_init(char **dest,const char *src) { size_t l; if (!src) src = ""; l = strlen(src); if (l == 0) { if (!null_string) { if((null_string = (char *)malloc(1)) == NULL) { DEBUG(0,("string_init: malloc fail for null_string.\n")); return False; } *null_string = 0; } *dest = null_string; } else { (*dest) = strdup(src); if ((*dest) == NULL) { DEBUG(0,("Out of memory in string_init\n")); return False; } } return(True); } /** Free a string value. **/ void string_free(char **s) { if (!s || !(*s)) return; if (*s == null_string) *s = NULL; SAFE_FREE(*s); } /** Set a string value, deallocating any existing space, and allocing the space for the string **/ BOOL string_set(char **dest,const char *src) { string_free(dest); return(string_init(dest,src)); } /** Substitute a string for a pattern in another string. Make sure there is enough room! This routine looks for pattern in s and replaces it with insert. It may do multiple replacements. Any of " ; ' $ or ` in the insert string are replaced with _ if len==0 then the string cannot be extended. This is different from the old use of len==0 which was for no length checks to be done. **/ void string_sub(char *s,const char *pattern, const char *insert, size_t len) { char *p; ssize_t ls,lp,li, i; if (!insert || !pattern || !*pattern || !s) return; ls = (ssize_t)strlen(s); lp = (ssize_t)strlen(pattern); li = (ssize_t)strlen(insert); if (len == 0) len = ls + 1; /* len is number of *bytes* */ while (lp <= ls && (p = strstr_m(s,pattern))) { if (ls + (li-lp) >= len) { DEBUG(0,("ERROR: string overflow by %d in string_sub(%.50s, %d)\n", (int)(ls + (li-lp) - len), pattern, (int)len)); break; } if (li != lp) { memmove(p+li,p+lp,strlen(p+lp)+1); } for (i=0;i