lib/util/charset/util_unistr_w.c

   1 /*
   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
   7
   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 3 of the License, or
  11    (at your option) any later version.
  12
  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.
  17
  18    You should have received a copy of the GNU General Public License
  19    along with this program.  If not, see <http://www.gnu.org/licenses/>.
  20 */
  21
  22 #include "includes.h"
  23
  24 /* Copy into a smb_ucs2_t from a possibly unaligned buffer. Return the copied smb_ucs2_t */
  25 #define COPY_UCS2_CHAR(dest,src) (((unsigned char *)(dest))[0] = ((const unsigned char *)(src))[0],\
  26                                 ((unsigned char *)(dest))[1] = ((const unsigned char *)(src))[1], (dest))
  27
  28
  29 /* return an ascii version of a ucs2 character */
  30 #define UCS2_TO_CHAR(c) (((c) >> UCS2_SHIFT) & 0xff)
  31
  32 static int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len);
  33
  34 /*******************************************************************
  35  Count the number of two-byte pairs in a UTF16 string.
  36 ********************************************************************/
  37
  38 size_t strlen_w(const smb_ucs2_t *src)
  39 {
  40         size_t len;
  41         smb_ucs2_t c;
  42
  43         for(len = 0; *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
  44                 ;
  45         }
  46
  47         return len;
  48 }
  49
  50 /*******************************************************************
  51  Count up to max number of characters in a smb_ucs2_t string.
  52 ********************************************************************/
  53
  54 size_t strnlen_w(const smb_ucs2_t *src, size_t max)
  55 {
  56         size_t len;
  57         smb_ucs2_t c;
  58
  59         for(len = 0; (len < max) && *(COPY_UCS2_CHAR(&c,src)); src++, len++) {
  60                 ;
  61         }
  62
  63         return len;
  64 }
  65
  66 /*******************************************************************
  67  Wide strchr().
  68 ********************************************************************/
  69
  70 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
  71 {
  72         smb_ucs2_t cp;
  73         while (*(COPY_UCS2_CHAR(&cp,s))) {
  74                 if (c == cp) {
  75                         return discard_const_p(smb_ucs2_t, s);
  76                 }
  77                 s++;
  78         }
  79         if (c == cp) {
  80                 return discard_const_p(smb_ucs2_t, s);
  81         }
  82
  83         return NULL;
  84 }
  85
  86 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c)
  87 {
  88         return strchr_w(s, UCS2_CHAR(c));
  89 }
  90
  91 /*******************************************************************
  92  Wide strrchr().
  93 ********************************************************************/
  94
  95 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c)
  96 {
  97         smb_ucs2_t cp;
  98         const smb_ucs2_t *p = s;
  99         int len = strlen_w(s);
 100
 101         if (len == 0) {
 102                 return NULL;
 103         }
 104         p += (len - 1);
 105         do {
 106                 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
 107                         return discard_const_p(smb_ucs2_t, p);
 108                 }
 109         } while (p-- != s);
 110         return NULL;
 111 }
 112
 113 /*******************************************************************
 114  Wide version of strrchr that returns after doing strrchr 'n' times.
 115 ********************************************************************/
 116
 117 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n)
 118 {
 119         smb_ucs2_t cp;
 120         const smb_ucs2_t *p = s;
 121         int len = strlen_w(s);
 122
 123         if (len == 0 || !n) {
 124                 return NULL;
 125         }
 126         p += (len - 1);
 127         do {
 128                 if (c == *(COPY_UCS2_CHAR(&cp,p))) {
 129                         n--;
 130                 }
 131
 132                 if (!n) {
 133                         return discard_const_p(smb_ucs2_t, p);
 134                 }
 135         } while (p-- != s);
 136         return NULL;
 137 }
 138
 139 /*******************************************************************
 140  Wide strstr().
 141 ********************************************************************/
 142
 143 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins)
 144 {
 145         const smb_ucs2_t *r;
 146         size_t inslen;
 147
 148         if (!s || !*s || !ins || !*ins) {
 149                 return NULL;
 150         }
 151
 152         inslen = strlen_w(ins);
 153         r = s;
 154
 155         while ((r = strchr_w(r, *ins))) {
 156                 if (strncmp_w(r, ins, inslen) == 0) {
 157                         return discard_const_p(smb_ucs2_t, r);
 158                 }
 159                 r++;
 160         }
 161
 162         return NULL;
 163 }
 164
 165 /*******************************************************************
 166  Convert a string to lower case.
 167  return True if any char is converted
 168
 169  This is unsafe for any string involving a UTF16 character
 170 ********************************************************************/
 171
 172 bool strlower_w(smb_ucs2_t *s)
 173 {
 174         smb_ucs2_t cp;
 175         bool ret = false;
 176
 177         while (*(COPY_UCS2_CHAR(&cp,s))) {
 178                 smb_ucs2_t v = tolower_m(cp);
 179                 if (v != cp) {
 180                         COPY_UCS2_CHAR(s,&v);
 181                         ret = true;
 182                 }
 183                 s++;
 184         }
 185         return ret;
 186 }
 187
 188 /*******************************************************************
 189  Convert a string to upper case.
 190  return True if any char is converted
 191
 192  This is unsafe for any string involving a UTF16 character
 193 ********************************************************************/
 194
 195 bool strupper_w(smb_ucs2_t *s)
 196 {
 197         smb_ucs2_t cp;
 198         bool ret = false;
 199         while (*(COPY_UCS2_CHAR(&cp,s))) {
 200                 smb_ucs2_t v = toupper_m(cp);
 201                 if (v != cp) {
 202                         COPY_UCS2_CHAR(s,&v);
 203                         ret = true;
 204                 }
 205                 s++;
 206         }
 207         return ret;
 208 }
 209
 210 static int strncmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len)
 211 {
 212         smb_ucs2_t cpa, cpb;
 213         size_t n = 0;
 214
 215         while ((n < len) && (*(COPY_UCS2_CHAR(&cpb,b))) && (*(COPY_UCS2_CHAR(&cpa,a)) == cpb)) {
 216                 a++;
 217                 b++;
 218                 n++;
 219         }
 220         return (len - n)?(*(COPY_UCS2_CHAR(&cpa,a)) - *(COPY_UCS2_CHAR(&cpb,b))):0;
 221 }
 222
 223 /*
 224   The *_wa() functions take a combination of 7 bit ascii
 225   and wide characters They are used so that you can use string
 226   functions combining C string constants with ucs2 strings
 227
 228   The char* arguments must NOT be multibyte - to be completely sure
 229   of this only pass string constants */
 230
 231 int strcmp_wa(const smb_ucs2_t *a, const char *b)
 232 {
 233         smb_ucs2_t cp = 0;
 234
 235         while (*b && *(COPY_UCS2_CHAR(&cp,a)) == UCS2_CHAR(*b)) {
 236                 a++;
 237                 b++;
 238         }
 239         return (*(COPY_UCS2_CHAR(&cp,a)) - UCS2_CHAR(*b));
 240 }
 241
 242 smb_ucs2_t toupper_w(smb_ucs2_t v)
 243 {
 244         smb_ucs2_t ret;
 245         /* LE to native. */
 246         codepoint_t cp = SVAL(&v,0);
 247         cp = toupper_m(cp);
 248         /* native to LE. */
 249         SSVAL(&ret,0,cp);
 250         return ret;
 251 }