stringprep.c

   1 /* stringprep.c library implementing IETF stringprep specification
   2  * Copyright (C) 2002  Simon Josefsson
   3  *
   4  * This file is part of libstringprep.
   5  *
   6  * Libstringprep is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * Libstringprep is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with libstringprep; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  19  *
  20  */
  21
  22 #include "internal.h"
  23
  24 #define DBG 0
  25
  26 static int
  27 stringprep_find_character_in_table (unsigned long ucs4,
  28                                     Stringprep_table_element * table)
  29 {
  30   int i;
  31
  32   for (i = 0; table[i].start; i++)
  33     if (ucs4 >= table[i].start &&
  34         ucs4 <= (table[i].end ? table[i].end : table[i].start))
  35       return i;
  36
  37   return -1;
  38 }
  39
  40 static int
  41 stringprep_find_string_in_table (unsigned long *ucs4,
  42                                  int ucs4len,
  43                                  int *tablepos,
  44                                  Stringprep_table_element * table)
  45 {
  46   int j, pos;
  47
  48   for (j = 0; j < ucs4len; j++)
  49     if ((pos = stringprep_find_character_in_table (ucs4[j], table)) != -1)
  50       {
  51         if (tablepos)
  52           *tablepos = pos;
  53         return j;
  54       }
  55
  56   return -1;
  57 }
  58
  59 static int
  60 stringprep_apply_table_to_string (unsigned long *ucs4,
  61                                   int *ucs4len,
  62                                   int maxucs4len,
  63                                   Stringprep_table_element * table)
  64 {
  65   int i;
  66   int pos, maplen;
  67
  68   while ((pos = stringprep_find_string_in_table
  69           (ucs4, *ucs4len, &i, table)) != -1)
  70     {
  71 #if DBG
  72       printf ("hit %06lx %06lx-%06lx\n", ucs4[pos], table[i].start,
  73               table[i].end);
  74 #endif
  75
  76       for (maplen = 0; table[i].map[maplen]; maplen++)
  77         ;
  78
  79       if (*ucs4len - 1 + maplen >= maxucs4len)
  80         return STRINGPREP_TOO_SMALL_BUFFER;
  81
  82       memmove (&ucs4[pos + maplen], &ucs4[pos + 1],
  83                *ucs4len * sizeof (unsigned long) - (&ucs4[pos + 1] - ucs4));
  84       memcpy (&ucs4[pos], table[i].map, sizeof (unsigned long) * maplen);
  85       *ucs4len = *ucs4len - 1 + maplen;
  86     }
  87
  88   return STRINGPREP_OK;
  89 }
  90
  91 #define INVERTED(x) ((x) & ((~0UL) >> 1))
  92 #define UNAPPLICAPLEFLAGS(flags, profileflags) \
  93   ((!INVERTED(profileflags) && !(profileflags & flags) && profileflags) || \
  94    ( INVERTED(profileflags) && (profileflags & flags)))
  95
  96 int
  97 stringprep (char *in, int maxlen, int flags, Stringprep_profile * profile)
  98 {
  99   int i;
 100   int rc;
 101   char *p = 0;
 102   unsigned long *q = 0;
 103   unsigned long *ucs4;
 104   int ucs4len;
 105   int maxucs4len;
 106
 107   ucs4 = stringprep_utf8_to_ucs4 (in, -1, &ucs4len);
 108   maxucs4len = 4 * ucs4len + 10;
 109   ucs4 = realloc (ucs4, 1 + maxucs4len * sizeof (unsigned long));
 110   if (!ucs4)
 111     {
 112       rc = STRINGPREP_MALLOC_ERROR;
 113       goto done;
 114     }
 115
 116 #if DBG
 117   {
 118     int j;
 119     printf ("\t;;    in: ");
 120     for (j = 0; j < ucs4len; j++)
 121       printf ("U+%06lx ", ucs4[j]);
 122     printf ("\n");
 123   }
 124 #endif
 125
 126   for (i = 0; profile[i].operation; i++)
 127     {
 128       switch (profile[i].operation)
 129         {
 130         case STRINGPREP_NFKC:
 131           if (UNAPPLICAPLEFLAGS (flags, profile[i].flags))
 132             break;
 133           if (flags & STRINGPREP_NO_NFKC && !profile[i].flags)
 134             {
 135               /* Profile requires NFKC, but callee asked for no NFKC. */
 136               rc = STRINGPREP_FLAG_ERROR;
 137               goto done;
 138             }
 139
 140           q = stringprep_ucs4_nfkc_normalize (ucs4, ucs4len);
 141
 142           if (!q)
 143             {
 144               rc = STRINGPREP_NFKC_FAILED;
 145               goto done;
 146             }
 147
 148           for (ucs4len = 0; ucs4[ucs4len]; ucs4len++)
 149             ;
 150
 151           free (ucs4);
 152           ucs4 = q;
 153           q = 0;
 154           break;
 155
 156         case STRINGPREP_PROHIBIT_TABLE:
 157           if (stringprep_find_string_in_table
 158               (ucs4, ucs4len, NULL, profile[i].table) != -1)
 159             {
 160               rc = STRINGPREP_CONTAINS_PROHIBITED;
 161               goto done;
 162             }
 163           break;
 164
 165         case STRINGPREP_UNASSIGNED_TABLE:
 166           if (UNAPPLICAPLEFLAGS (flags, profile[i].flags))
 167             break;
 168           if (flags & STRINGPREP_NO_UNASSIGNED)
 169             if (stringprep_find_string_in_table
 170                 (ucs4, ucs4len, NULL, profile[i].table) != -1)
 171               {
 172                 rc = STRINGPREP_CONTAINS_UNASSIGNED;
 173                 goto done;
 174               }
 175           break;
 176
 177         case STRINGPREP_MAP_TABLE:
 178           if (UNAPPLICAPLEFLAGS (flags, profile[i].flags))
 179             break;
 180           rc = stringprep_apply_table_to_string
 181             (ucs4, &ucs4len, maxucs4len, profile[i].table);
 182           if (rc != STRINGPREP_OK)
 183             goto done;
 184           break;
 185
 186         case STRINGPREP_BIDI_PROHIBIT_TABLE:
 187         case STRINGPREP_BIDI_RAL_TABLE:
 188         case STRINGPREP_BIDI_L_TABLE:
 189           break;
 190
 191         case STRINGPREP_BIDI:
 192           {
 193             int done_prohibited = 0;
 194             int done_ral = 0;
 195             int done_l = 0;
 196             int contains_ral = -1;
 197             int contains_l = -1;
 198             int j;
 199
 200             for (j = 0; profile[j].operation; j++)
 201               if (profile[j].operation == STRINGPREP_BIDI_PROHIBIT_TABLE)
 202                 {
 203                   done_prohibited = 1;
 204                   if (stringprep_find_string_in_table
 205                       (ucs4, ucs4len, NULL, profile[j].table) != -1)
 206                     {
 207                       rc = STRINGPREP_BIDI_CONTAINS_PROHIBITED;
 208                       goto done;
 209                     }
 210                 }
 211               else if (profile[j].operation == STRINGPREP_BIDI_RAL_TABLE)
 212                 {
 213                   done_ral = 1;
 214                   if (stringprep_find_string_in_table
 215                       (ucs4, ucs4len, NULL, profile[j].table) != -1)
 216                     contains_ral = j;
 217                 }
 218               else if (profile[j].operation == STRINGPREP_BIDI_L_TABLE)
 219                 {
 220                   done_l = 1;
 221                   if (stringprep_find_string_in_table
 222                       (ucs4, ucs4len, NULL, profile[j].table) != -1)
 223                     contains_l = j;
 224                 }
 225
 226             if (!done_prohibited || !done_ral || !done_l)
 227               {
 228                 rc = STRINGPREP_PROFILE_ERROR;
 229                 goto done;
 230               }
 231
 232             if (contains_ral != -1 && contains_l != -1)
 233               {
 234                 rc = STRINGPREP_BIDI_BOTH_L_AND_RAL;
 235                 goto done;
 236               }
 237
 238             if (contains_ral != -1)
 239               if (!(stringprep_find_character_in_table
 240                     (ucs4[0], profile[contains_ral].table) != -1 &&
 241                     stringprep_find_character_in_table
 242                     (ucs4[ucs4len - 1], profile[contains_ral].table) != -1))
 243                 {
 244                   rc = STRINGPREP_BIDI_LEADTRAIL_NOT_RAL;
 245                   goto done;
 246                 }
 247           }
 248           break;
 249
 250         default:
 251           rc = STRINGPREP_PROFILE_ERROR;
 252           goto done;
 253           break;
 254         }
 255 #if DBG
 256       {
 257         int j;
 258         printf ("\t;; %5s: ", profile[i].name ? profile[i].name : "<NONAME>");
 259         for (j = 0; j < ucs4len; j++)
 260           printf ("U+%06lx ", ucs4[j]);
 261         printf ("\n");
 262       }
 263 #endif
 264     }
 265
 266   p = stringprep_ucs4_to_utf8 (ucs4, ucs4len, 0, 0);
 267
 268   if (strlen (p) >= maxlen)
 269     {
 270       rc = STRINGPREP_TOO_SMALL_BUFFER;
 271       goto done;
 272     }
 273
 274   strcpy (in, p);               /* flawfinder: ignore */
 275
 276   rc = STRINGPREP_OK;
 277
 278 done:
 279   if (p)
 280     free (p);
 281   if (q)
 282     free (q);
 283   if (ucs4)
 284     free (ucs4);
 285   return rc;
 286 }