fs/ntfs/unistr.c

   1 /*
   2  * unistr.c - NTFS Unicode string handling. Part of the Linux-NTFS project.
   3  *
   4  * Copyright (c) 2001-2004 Anton Altaparmakov
   5  *
   6  * This program/include file is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU General Public License as published
   8  * by the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program/include file is distributed in the hope that it will be
  12  * useful, but WITHOUT ANY WARRANTY; without even the implied warranty
  13  * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * along with this program (in the main directory of the Linux-NTFS
  18  * distribution in the file COPYING); if not, write to the Free Software
  19  * Foundation,Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  20  */
  21
  22 #include "ntfs.h"
  23
  24 /*
  25  * IMPORTANT
  26  * =========
  27  *
  28  * All these routines assume that the Unicode characters are in little endian
  29  * encoding inside the strings!!!
  30  */
  31
  32 /*
  33  * This is used by the name collation functions to quickly determine what
  34  * characters are (in)valid.
  35  */
  36 static const u8 legal_ansi_char_array[0x40] = {
  37         0x00, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10,
  38         0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10,
  39
  40         0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10,
  41         0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10, 0x10,
  42
  43         0x17, 0x07, 0x18, 0x17, 0x17, 0x17, 0x17, 0x17,
  44         0x17, 0x17, 0x18, 0x16, 0x16, 0x17, 0x07, 0x00,
  45
  46         0x17, 0x17, 0x17, 0x17, 0x17, 0x17, 0x17, 0x17,
  47         0x17, 0x17, 0x04, 0x16, 0x18, 0x16, 0x18, 0x18,
  48 };
  49
  50 /**
  51  * ntfs_are_names_equal - compare two Unicode names for equality
  52  * @s1:                 name to compare to @s2
  53  * @s1_len:             length in Unicode characters of @s1
  54  * @s2:                 name to compare to @s1
  55  * @s2_len:             length in Unicode characters of @s2
  56  * @ic:                 ignore case bool
  57  * @upcase:             upcase table (only if @ic == IGNORE_CASE)
  58  * @upcase_size:        length in Unicode characters of @upcase (if present)
  59  *
  60  * Compare the names @s1 and @s2 and return TRUE (1) if the names are
  61  * identical, or FALSE (0) if they are not identical. If @ic is IGNORE_CASE,
  62  * the @upcase table is used to performa a case insensitive comparison.
  63  */
  64 BOOL ntfs_are_names_equal(const ntfschar *s1, size_t s1_len,
  65                 const ntfschar *s2, size_t s2_len, const IGNORE_CASE_BOOL ic,
  66                 const ntfschar *upcase, const u32 upcase_size)
  67 {
  68         if (s1_len != s2_len)
  69                 return FALSE;
  70         if (ic == CASE_SENSITIVE)
  71                 return !ntfs_ucsncmp(s1, s2, s1_len);
  72         return !ntfs_ucsncasecmp(s1, s2, s1_len, upcase, upcase_size);
  73 }
  74
  75 /**
  76  * ntfs_collate_names - collate two Unicode names
  77  * @name1:      first Unicode name to compare
  78  * @name2:      second Unicode name to compare
  79  * @err_val:    if @name1 contains an invalid character return this value
  80  * @ic:         either CASE_SENSITIVE or IGNORE_CASE
  81  * @upcase:     upcase table (ignored if @ic is CASE_SENSITIVE)
  82  * @upcase_len: upcase table size (ignored if @ic is CASE_SENSITIVE)
  83  *
  84  * ntfs_collate_names collates two Unicode names and returns:
  85  *
  86  *  -1 if the first name collates before the second one,
  87  *   0 if the names match,
  88  *   1 if the second name collates before the first one, or
  89  * @err_val if an invalid character is found in @name1 during the comparison.
  90  *
  91  * The following characters are considered invalid: '"', '*', '<', '>' and '?'.
  92  */
  93 int ntfs_collate_names(const ntfschar *name1, const u32 name1_len,
  94                 const ntfschar *name2, const u32 name2_len,
  95                 const int err_val, const IGNORE_CASE_BOOL ic,
  96                 const ntfschar *upcase, const u32 upcase_len)
  97 {
  98         u32 cnt, min_len;
  99         u16 c1, c2;
 100
 101         min_len = name1_len;
 102         if (name1_len > name2_len)
 103                 min_len = name2_len;
 104         for (cnt = 0; cnt < min_len; ++cnt) {
 105                 c1 = le16_to_cpu(*name1++);
 106                 c2 = le16_to_cpu(*name2++);
 107                 if (ic) {
 108                         if (c1 < upcase_len)
 109                                 c1 = le16_to_cpu(upcase[c1]);
 110                         if (c2 < upcase_len)
 111                                 c2 = le16_to_cpu(upcase[c2]);
 112                 }
 113                 if (c1 < 64 && legal_ansi_char_array[c1] & 8)
 114                         return err_val;
 115                 if (c1 < c2)
 116                         return -1;
 117                 if (c1 > c2)
 118                         return 1;
 119         }
 120         if (name1_len < name2_len)
 121                 return -1;
 122         if (name1_len == name2_len)
 123                 return 0;
 124         /* name1_len > name2_len */
 125         c1 = le16_to_cpu(*name1);
 126         if (c1 < 64 && legal_ansi_char_array[c1] & 8)
 127                 return err_val;
 128         return 1;
 129 }
 130
 131 /**
 132  * ntfs_ucsncmp - compare two little endian Unicode strings
 133  * @s1:         first string
 134  * @s2:         second string
 135  * @n:          maximum unicode characters to compare
 136  *
 137  * Compare the first @n characters of the Unicode strings @s1 and @s2,
 138  * The strings in little endian format and appropriate le16_to_cpu()
 139  * conversion is performed on non-little endian machines.
 140  *
 141  * The function returns an integer less than, equal to, or greater than zero
 142  * if @s1 (or the first @n Unicode characters thereof) is found, respectively,
 143  * to be less than, to match, or be greater than @s2.
 144  */
 145 int ntfs_ucsncmp(const ntfschar *s1, const ntfschar *s2, size_t n)
 146 {
 147         u16 c1, c2;
 148         size_t i;
 149
 150         for (i = 0; i < n; ++i) {
 151                 c1 = le16_to_cpu(s1[i]);
 152                 c2 = le16_to_cpu(s2[i]);
 153                 if (c1 < c2)
 154                         return -1;
 155                 if (c1 > c2)
 156                         return 1;
 157                 if (!c1)
 158                         break;
 159         }
 160         return 0;
 161 }
 162
 163 /**
 164  * ntfs_ucsncasecmp - compare two little endian Unicode strings, ignoring case
 165  * @s1:                 first string
 166  * @s2:                 second string
 167  * @n:                  maximum unicode characters to compare
 168  * @upcase:             upcase table
 169  * @upcase_size:        upcase table size in Unicode characters
 170  *
 171  * Compare the first @n characters of the Unicode strings @s1 and @s2,
 172  * ignoring case. The strings in little endian format and appropriate
 173  * le16_to_cpu() conversion is performed on non-little endian machines.
 174  *
 175  * Each character is uppercased using the @upcase table before the comparison.
 176  *
 177  * The function returns an integer less than, equal to, or greater than zero
 178  * if @s1 (or the first @n Unicode characters thereof) is found, respectively,
 179  * to be less than, to match, or be greater than @s2.
 180  */
 181 int ntfs_ucsncasecmp(const ntfschar *s1, const ntfschar *s2, size_t n,
 182                 const ntfschar *upcase, const u32 upcase_size)
 183 {
 184         size_t i;
 185         u16 c1, c2;
 186
 187         for (i = 0; i < n; ++i) {
 188                 if ((c1 = le16_to_cpu(s1[i])) < upcase_size)
 189                         c1 = le16_to_cpu(upcase[c1]);
 190                 if ((c2 = le16_to_cpu(s2[i])) < upcase_size)
 191                         c2 = le16_to_cpu(upcase[c2]);
 192                 if (c1 < c2)
 193                         return -1;
 194                 if (c1 > c2)
 195                         return 1;
 196                 if (!c1)
 197                         break;
 198         }
 199         return 0;
 200 }
 201
 202 void ntfs_upcase_name(ntfschar *name, u32 name_len, const ntfschar *upcase,
 203                 const u32 upcase_len)
 204 {
 205         u32 i;
 206         u16 u;
 207
 208         for (i = 0; i < name_len; i++)
 209                 if ((u = le16_to_cpu(name[i])) < upcase_len)
 210                         name[i] = upcase[u];
 211 }
 212
 213 void ntfs_file_upcase_value(FILE_NAME_ATTR *file_name_attr,
 214                 const ntfschar *upcase, const u32 upcase_len)
 215 {
 216         ntfs_upcase_name((ntfschar*)&file_name_attr->file_name,
 217                         file_name_attr->file_name_length, upcase, upcase_len);
 218 }
 219
 220 int ntfs_file_compare_values(FILE_NAME_ATTR *file_name_attr1,
 221                 FILE_NAME_ATTR *file_name_attr2,
 222                 const int err_val, const IGNORE_CASE_BOOL ic,
 223                 const ntfschar *upcase, const u32 upcase_len)
 224 {
 225         return ntfs_collate_names((ntfschar*)&file_name_attr1->file_name,
 226                         file_name_attr1->file_name_length,
 227                         (ntfschar*)&file_name_attr2->file_name,
 228                         file_name_attr2->file_name_length,
 229                         err_val, ic, upcase, upcase_len);
 230 }
 231
 232 /**
 233  * ntfs_nlstoucs - convert NLS string to little endian Unicode string
 234  * @vol:        ntfs volume which we are working with
 235  * @ins:        input NLS string buffer
 236  * @ins_len:    length of input string in bytes
 237  * @outs:       on return contains the allocated output Unicode string buffer
 238  *
 239  * Convert the input string @ins, which is in whatever format the loaded NLS
 240  * map dictates, into a little endian, 2-byte Unicode string.
 241  *
 242  * This function allocates the string and the caller is responsible for
 243  * calling kmem_cache_free(ntfs_name_cache, @outs); when finished with it.
 244  *
 245  * On success the function returns the number of Unicode characters written to
 246  * the output string *@outs (>= 0), not counting the terminating Unicode NULL
 247  * character. *@outs is set to the allocated output string buffer.
 248  *
 249  * On error, a negative number corresponding to the error code is returned. In
 250  * that case the output string is not allocated. Both *@outs and *@outs_len
 251  * are then undefined.
 252  *
 253  * This might look a bit odd due to fast path optimization...
 254  */
 255 int ntfs_nlstoucs(const ntfs_volume *vol, const char *ins,
 256                 const int ins_len, ntfschar **outs)
 257 {
 258         struct nls_table *nls = vol->nls_map;
 259         ntfschar *ucs;
 260         wchar_t wc;
 261         int i, o, wc_len;
 262
 263         /* We don't trust outside sources. */
 264         if (ins) {
 265                 ucs = (ntfschar*)kmem_cache_alloc(ntfs_name_cache, SLAB_NOFS);
 266                 if (ucs) {
 267                         for (i = o = 0; i < ins_len; i += wc_len) {
 268                                 wc_len = nls->char2uni(ins + i, ins_len - i,
 269                                                 &wc);
 270                                 if (wc_len >= 0) {
 271                                         if (wc) {
 272                                                 ucs[o++] = cpu_to_le16(wc);
 273                                                 continue;
 274                                         } /* else (!wc) */
 275                                         break;
 276                                 } /* else (wc_len < 0) */
 277                                 goto conversion_err;
 278                         }
 279                         ucs[o] = 0;
 280                         *outs = ucs;
 281                         return o;
 282                 } /* else (!ucs) */
 283                 ntfs_error(vol->sb, "Failed to allocate name from "
 284                                 "ntfs_name_cache!");
 285                 return -ENOMEM;
 286         } /* else (!ins) */
 287         ntfs_error(NULL, "Received NULL pointer.");
 288         return -EINVAL;
 289 conversion_err:
 290         ntfs_error(vol->sb, "Name using character set %s contains characters "
 291                         "that cannot be converted to Unicode.", nls->charset);
 292         kmem_cache_free(ntfs_name_cache, ucs);
 293         return -EILSEQ;
 294 }
 295
 296 /**
 297  * ntfs_ucstonls - convert little endian Unicode string to NLS string
 298  * @vol:        ntfs volume which we are working with
 299  * @ins:        input Unicode string buffer
 300  * @ins_len:    length of input string in Unicode characters
 301  * @outs:       on return contains the (allocated) output NLS string buffer
 302  * @outs_len:   length of output string buffer in bytes
 303  *
 304  * Convert the input little endian, 2-byte Unicode string @ins, of length
 305  * @ins_len into the string format dictated by the loaded NLS.
 306  *
 307  * If *@outs is NULL, this function allocates the string and the caller is
 308  * responsible for calling kfree(*@outs); when finished with it. In this case
 309  * @outs_len is ignored and can be 0.
 310  *
 311  * On success the function returns the number of bytes written to the output
 312  * string *@outs (>= 0), not counting the terminating NULL byte. If the output
 313  * string buffer was allocated, *@outs is set to it.
 314  *
 315  * On error, a negative number corresponding to the error code is returned. In
 316  * that case the output string is not allocated. The contents of *@outs are
 317  * then undefined.
 318  *
 319  * This might look a bit odd due to fast path optimization...
 320  */
 321 int ntfs_ucstonls(const ntfs_volume *vol, const ntfschar *ins,
 322                 const int ins_len, unsigned char **outs, int outs_len)
 323 {
 324         struct nls_table *nls = vol->nls_map;
 325         unsigned char *ns;
 326         int i, o, ns_len, wc;
 327
 328         /* We don't trust outside sources. */
 329         if (ins) {
 330                 ns = *outs;
 331                 ns_len = outs_len;
 332                 if (ns && !ns_len) {
 333                         wc = -ENAMETOOLONG;
 334                         goto conversion_err;
 335                 }
 336                 if (!ns) {
 337                         ns_len = ins_len * NLS_MAX_CHARSET_SIZE;
 338                         ns = (unsigned char*)kmalloc(ns_len + 1, GFP_NOFS);
 339                         if (!ns)
 340                                 goto mem_err_out;
 341                 }
 342                 for (i = o = 0; i < ins_len; i++) {
 343 retry:                  wc = nls->uni2char(le16_to_cpu(ins[i]), ns + o,
 344                                         ns_len - o);
 345                         if (wc > 0) {
 346                                 o += wc;
 347                                 continue;
 348                         } else if (!wc)
 349                                 break;
 350                         else if (wc == -ENAMETOOLONG && ns != *outs) {
 351                                 unsigned char *tc;
 352                                 /* Grow in multiples of 64 bytes. */
 353                                 tc = (unsigned char*)kmalloc((ns_len + 64) &
 354                                                 ~63, GFP_NOFS);
 355                                 if (tc) {
 356                                         memcpy(tc, ns, ns_len);
 357                                         ns_len = ((ns_len + 64) & ~63) - 1;
 358                                         kfree(ns);
 359                                         ns = tc;
 360                                         goto retry;
 361                                 } /* No memory so goto conversion_error; */
 362                         } /* wc < 0, real error. */
 363                         goto conversion_err;
 364                 }
 365                 ns[o] = 0;
 366                 *outs = ns;
 367                 return o;
 368         } /* else (!ins) */
 369         ntfs_error(vol->sb, "Received NULL pointer.");
 370         return -EINVAL;
 371 conversion_err:
 372         ntfs_error(vol->sb, "Unicode name contains characters that cannot be "
 373                         "converted to character set %s.", nls->charset);
 374         if (ns != *outs)
 375                 kfree(ns);
 376         if (wc != -ENAMETOOLONG)
 377                 wc = -EILSEQ;
 378         return wc;
 379 mem_err_out:
 380         ntfs_error(vol->sb, "Failed to allocate name!");
 381         return -ENOMEM;
 382 }