src/basic/utf8.c

   1 /*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/
   2
   3 /***
   4   This file is part of systemd.
   5
   6   Copyright 2008-2011 Kay Sievers
   7   Copyright 2012 Lennart Poettering
   8
   9   systemd is free software; you can redistribute it and/or modify it
  10   under the terms of the GNU Lesser General Public License as published by
  11   the Free Software Foundation; either version 2.1 of the License, or
  12   (at your option) any later version.
  13
  14   systemd is distributed in the hope that it will be useful, but
  15   WITHOUT ANY WARRANTY; without even the implied warranty of
  16   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  17   Lesser General Public License for more details.
  18
  19   You should have received a copy of the GNU Lesser General Public License
  20   along with systemd; If not, see <http://www.gnu.org/licenses/>.
  21 ***/
  22
  23 /* Parts of this file are based on the GLIB utf8 validation functions. The
  24  * original license text follows. */
  25
  26 /* gutf8.c - Operations on UTF-8 strings.
  27  *
  28  * Copyright (C) 1999 Tom Tromey
  29  * Copyright (C) 2000 Red Hat, Inc.
  30  *
  31  * This library is free software; you can redistribute it and/or
  32  * modify it under the terms of the GNU Library General Public
  33  * License as published by the Free Software Foundation; either
  34  * version 2 of the License, or (at your option) any later version.
  35  *
  36  * This library is distributed in the hope that it will be useful,
  37  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  38  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  39  * Library General Public License for more details.
  40  *
  41  * You should have received a copy of the GNU Library General Public
  42  * License along with this library; if not, write to the Free Software
  43  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
  44  */
  45
  46 #include <errno.h>
  47 #include <stdlib.h>
  48 #include <inttypes.h>
  49 #include <string.h>
  50 #include <stdbool.h>
  51
  52 #include "utf8.h"
  53 #include "util.h"
  54
  55 bool unichar_is_valid(uint32_t ch) {
  56
  57         if (ch >= 0x110000) /* End of unicode space */
  58                 return false;
  59         if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */
  60                 return false;
  61         if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */
  62                 return false;
  63         if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */
  64                 return false;
  65
  66         return true;
  67 }
  68
  69 static bool unichar_is_control(uint32_t ch) {
  70
  71         /*
  72           0 to ' '-1 is the C0 range.
  73           DEL=0x7F, and DEL+1 to 0x9F is C1 range.
  74           '\t' is in C0 range, but more or less harmless and commonly used.
  75         */
  76
  77         return (ch < ' ' && ch != '\t' && ch != '\n') ||
  78                 (0x7F <= ch && ch <= 0x9F);
  79 }
  80
  81 /* count of characters used to encode one unicode char */
  82 static int utf8_encoded_expected_len(const char *str) {
  83         unsigned char c;
  84
  85         assert(str);
  86
  87         c = (unsigned char) str[0];
  88         if (c < 0x80)
  89                 return 1;
  90         if ((c & 0xe0) == 0xc0)
  91                 return 2;
  92         if ((c & 0xf0) == 0xe0)
  93                 return 3;
  94         if ((c & 0xf8) == 0xf0)
  95                 return 4;
  96         if ((c & 0xfc) == 0xf8)
  97                 return 5;
  98         if ((c & 0xfe) == 0xfc)
  99                 return 6;
 100
 101         return 0;
 102 }
 103
 104 /* decode one unicode char */
 105 int utf8_encoded_to_unichar(const char *str) {
 106         int unichar, len, i;
 107
 108         assert(str);
 109
 110         len = utf8_encoded_expected_len(str);
 111
 112         switch (len) {
 113         case 1:
 114                 return (int)str[0];
 115         case 2:
 116                 unichar = str[0] & 0x1f;
 117                 break;
 118         case 3:
 119                 unichar = (int)str[0] & 0x0f;
 120                 break;
 121         case 4:
 122                 unichar = (int)str[0] & 0x07;
 123                 break;
 124         case 5:
 125                 unichar = (int)str[0] & 0x03;
 126                 break;
 127         case 6:
 128                 unichar = (int)str[0] & 0x01;
 129                 break;
 130         default:
 131                 return -EINVAL;
 132         }
 133
 134         for (i = 1; i < len; i++) {
 135                 if (((int)str[i] & 0xc0) != 0x80)
 136                         return -EINVAL;
 137                 unichar <<= 6;
 138                 unichar |= (int)str[i] & 0x3f;
 139         }
 140
 141         return unichar;
 142 }
 143
 144 bool utf8_is_printable_newline(const char* str, size_t length, bool newline) {
 145         const char *p;
 146
 147         assert(str);
 148
 149         for (p = str; length;) {
 150                 int encoded_len, val;
 151
 152                 encoded_len = utf8_encoded_valid_unichar(p);
 153                 if (encoded_len < 0 ||
 154                     (size_t) encoded_len > length)
 155                         return false;
 156
 157                 val = utf8_encoded_to_unichar(p);
 158                 if (val < 0 ||
 159                     unichar_is_control(val) ||
 160                     (!newline && val == '\n'))
 161                         return false;
 162
 163                 length -= encoded_len;
 164                 p += encoded_len;
 165         }
 166
 167         return true;
 168 }
 169
 170 const char *utf8_is_valid(const char *str) {
 171         const uint8_t *p;
 172
 173         assert(str);
 174
 175         for (p = (const uint8_t*) str; *p; ) {
 176                 int len;
 177
 178                 len = utf8_encoded_valid_unichar((const char *)p);
 179                 if (len < 0)
 180                         return NULL;
 181
 182                 p += len;
 183         }
 184
 185         return str;
 186 }
 187
 188 char *utf8_escape_invalid(const char *str) {
 189         char *p, *s;
 190
 191         assert(str);
 192
 193         p = s = malloc(strlen(str) * 4 + 1);
 194         if (!p)
 195                 return NULL;
 196
 197         while (*str) {
 198                 int len;
 199
 200                 len = utf8_encoded_valid_unichar(str);
 201                 if (len > 0) {
 202                         s = mempcpy(s, str, len);
 203                         str += len;
 204                 } else {
 205                         s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
 206                         str += 1;
 207                 }
 208         }
 209
 210         *s = '\0';
 211
 212         return p;
 213 }
 214
 215 char *utf8_escape_non_printable(const char *str) {
 216         char *p, *s;
 217
 218         assert(str);
 219
 220         p = s = malloc(strlen(str) * 4 + 1);
 221         if (!p)
 222                 return NULL;
 223
 224         while (*str) {
 225                 int len;
 226
 227                 len = utf8_encoded_valid_unichar(str);
 228                 if (len > 0) {
 229                         if (utf8_is_printable(str, len)) {
 230                                 s = mempcpy(s, str, len);
 231                                 str += len;
 232                         } else {
 233                                 while (len > 0) {
 234                                         *(s++) = '\\';
 235                                         *(s++) = 'x';
 236                                         *(s++) = hexchar((int) *str >> 4);
 237                                         *(s++) = hexchar((int) *str);
 238
 239                                         str += 1;
 240                                         len --;
 241                                 }
 242                         }
 243                 } else {
 244                         s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
 245                         str += 1;
 246                 }
 247         }
 248
 249         *s = '\0';
 250
 251         return p;
 252 }
 253
 254 char *ascii_is_valid(const char *str) {
 255         const char *p;
 256
 257         assert(str);
 258
 259         for (p = str; *p; p++)
 260                 if ((unsigned char) *p >= 128)
 261                         return NULL;
 262
 263         return (char*) str;
 264 }
 265
 266 /**
 267  * utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
 268  * @out_utf8: output buffer of at least 4 bytes or NULL
 269  * @g: UCS-4 character to encode
 270  *
 271  * This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
 272  * The length of the character is returned. It is not zero-terminated! If the
 273  * output buffer is NULL, only the length is returned.
 274  *
 275  * Returns: The length in bytes that the UTF-8 representation does or would
 276  *          occupy.
 277  */
 278 size_t utf8_encode_unichar(char *out_utf8, uint32_t g) {
 279
 280         if (g < (1 << 7)) {
 281                 if (out_utf8)
 282                         out_utf8[0] = g & 0x7f;
 283                 return 1;
 284         } else if (g < (1 << 11)) {
 285                 if (out_utf8) {
 286                         out_utf8[0] = 0xc0 | ((g >> 6) & 0x1f);
 287                         out_utf8[1] = 0x80 | (g & 0x3f);
 288                 }
 289                 return 2;
 290         } else if (g < (1 << 16)) {
 291                 if (out_utf8) {
 292                         out_utf8[0] = 0xe0 | ((g >> 12) & 0x0f);
 293                         out_utf8[1] = 0x80 | ((g >> 6) & 0x3f);
 294                         out_utf8[2] = 0x80 | (g & 0x3f);
 295                 }
 296                 return 3;
 297         } else if (g < (1 << 21)) {
 298                 if (out_utf8) {
 299                         out_utf8[0] = 0xf0 | ((g >> 18) & 0x07);
 300                         out_utf8[1] = 0x80 | ((g >> 12) & 0x3f);
 301                         out_utf8[2] = 0x80 | ((g >> 6) & 0x3f);
 302                         out_utf8[3] = 0x80 | (g & 0x3f);
 303                 }
 304                 return 4;
 305         }
 306
 307         return 0;
 308 }
 309
 310 char *utf16_to_utf8(const void *s, size_t length) {
 311         const uint8_t *f;
 312         char *r, *t;
 313
 314         r = new(char, (length * 4 + 1) / 2 + 1);
 315         if (!r)
 316                 return NULL;
 317
 318         f = s;
 319         t = r;
 320
 321         while (f < (const uint8_t*) s + length) {
 322                 uint16_t w1, w2;
 323
 324                 /* see RFC 2781 section 2.2 */
 325
 326                 w1 = f[1] << 8 | f[0];
 327                 f += 2;
 328
 329                 if (!utf16_is_surrogate(w1)) {
 330                         t += utf8_encode_unichar(t, w1);
 331
 332                         continue;
 333                 }
 334
 335                 if (utf16_is_trailing_surrogate(w1))
 336                         continue;
 337                 else if (f >= (const uint8_t*) s + length)
 338                         break;
 339
 340                 w2 = f[1] << 8 | f[0];
 341                 f += 2;
 342
 343                 if (!utf16_is_trailing_surrogate(w2)) {
 344                         f -= 2;
 345                         continue;
 346                 }
 347
 348                 t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));
 349         }
 350
 351         *t = 0;
 352         return r;
 353 }
 354
 355 /* expected size used to encode one unicode char */
 356 static int utf8_unichar_to_encoded_len(int unichar) {
 357
 358         if (unichar < 0x80)
 359                 return 1;
 360         if (unichar < 0x800)
 361                 return 2;
 362         if (unichar < 0x10000)
 363                 return 3;
 364         if (unichar < 0x200000)
 365                 return 4;
 366         if (unichar < 0x4000000)
 367                 return 5;
 368
 369         return 6;
 370 }
 371
 372 /* validate one encoded unicode char and return its length */
 373 int utf8_encoded_valid_unichar(const char *str) {
 374         int len, unichar, i;
 375
 376         assert(str);
 377
 378         len = utf8_encoded_expected_len(str);
 379         if (len == 0)
 380                 return -EINVAL;
 381
 382         /* ascii is valid */
 383         if (len == 1)
 384                 return 1;
 385
 386         /* check if expected encoded chars are available */
 387         for (i = 0; i < len; i++)
 388                 if ((str[i] & 0x80) != 0x80)
 389                         return -EINVAL;
 390
 391         unichar = utf8_encoded_to_unichar(str);
 392
 393         /* check if encoded length matches encoded value */
 394         if (utf8_unichar_to_encoded_len(unichar) != len)
 395                 return -EINVAL;
 396
 397         /* check if value has valid range */
 398         if (!unichar_is_valid(unichar))
 399                 return -EINVAL;
 400
 401         return len;
 402 }