src/basic/utf8.c

   1 /* SPDX-License-Identifier: LGPL-2.1+ */
   2 /***
   3   This file is part of systemd.
   4
   5   Copyright 2008-2011 Kay Sievers
   6   Copyright 2012 Lennart Poettering
   7
   8   systemd is free software; you can redistribute it and/or modify it
   9   under the terms of the GNU Lesser General Public License as published by
  10   the Free Software Foundation; either version 2.1 of the License, or
  11   (at your option) any later version.
  12
  13   systemd is distributed in the hope that it will be useful, but
  14   WITHOUT ANY WARRANTY; without even the implied warranty of
  15   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  16   Lesser General Public License for more details.
  17
  18   You should have received a copy of the GNU Lesser General Public License
  19   along with systemd; If not, see <http://www.gnu.org/licenses/>.
  20 ***/
  21
  22 /* Parts of this file are based on the GLIB utf8 validation functions. The
  23  * original license text follows. */
  24
  25 /* gutf8.c - Operations on UTF-8 strings.
  26  *
  27  * Copyright (C) 1999 Tom Tromey
  28  * Copyright (C) 2000 Red Hat, Inc.
  29  *
  30  * This library is free software; you can redistribute it and/or
  31  * modify it under the terms of the GNU Library General Public
  32  * License as published by the Free Software Foundation; either
  33  * version 2 of the License, or (at your option) any later version.
  34  *
  35  * This library is distributed in the hope that it will be useful,
  36  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  37  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  38  * Library General Public License for more details.
  39  *
  40  * You should have received a copy of the GNU Library General Public
  41  * License along with this library; if not, write to the Free Software
  42  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
  43  */
  44
  45 #include <errno.h>
  46 #include <stdbool.h>
  47 #include <stdlib.h>
  48 #include <string.h>
  49
  50 #include "alloc-util.h"
  51 #include "hexdecoct.h"
  52 #include "macro.h"
  53 #include "utf8.h"
  54
  55 bool unichar_is_valid(char32_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(char32_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 < ' ' && !IN_SET(ch, '\t', '\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, char32_t *ret_unichar) {
 106         char32_t unichar;
 107         int len, i;
 108
 109         assert(str);
 110
 111         len = utf8_encoded_expected_len(str);
 112
 113         switch (len) {
 114         case 1:
 115                 *ret_unichar = (char32_t)str[0];
 116                 return 0;
 117         case 2:
 118                 unichar = str[0] & 0x1f;
 119                 break;
 120         case 3:
 121                 unichar = (char32_t)str[0] & 0x0f;
 122                 break;
 123         case 4:
 124                 unichar = (char32_t)str[0] & 0x07;
 125                 break;
 126         case 5:
 127                 unichar = (char32_t)str[0] & 0x03;
 128                 break;
 129         case 6:
 130                 unichar = (char32_t)str[0] & 0x01;
 131                 break;
 132         default:
 133                 return -EINVAL;
 134         }
 135
 136         for (i = 1; i < len; i++) {
 137                 if (((char32_t)str[i] & 0xc0) != 0x80)
 138                         return -EINVAL;
 139                 unichar <<= 6;
 140                 unichar |= (char32_t)str[i] & 0x3f;
 141         }
 142
 143         *ret_unichar = unichar;
 144
 145         return 0;
 146 }
 147
 148 bool utf8_is_printable_newline(const char* str, size_t length, bool newline) {
 149         const char *p;
 150
 151         assert(str);
 152
 153         for (p = str; length;) {
 154                 int encoded_len, r;
 155                 char32_t val;
 156
 157                 encoded_len = utf8_encoded_valid_unichar(p);
 158                 if (encoded_len < 0 ||
 159                     (size_t) encoded_len > length)
 160                         return false;
 161
 162                 r = utf8_encoded_to_unichar(p, &val);
 163                 if (r < 0 ||
 164                     unichar_is_control(val) ||
 165                     (!newline && val == '\n'))
 166                         return false;
 167
 168                 length -= encoded_len;
 169                 p += encoded_len;
 170         }
 171
 172         return true;
 173 }
 174
 175 const char *utf8_is_valid(const char *str) {
 176         const uint8_t *p;
 177
 178         assert(str);
 179
 180         for (p = (const uint8_t*) str; *p; ) {
 181                 int len;
 182
 183                 len = utf8_encoded_valid_unichar((const char *)p);
 184                 if (len < 0)
 185                         return NULL;
 186
 187                 p += len;
 188         }
 189
 190         return str;
 191 }
 192
 193 char *utf8_escape_invalid(const char *str) {
 194         char *p, *s;
 195
 196         assert(str);
 197
 198         p = s = malloc(strlen(str) * 4 + 1);
 199         if (!p)
 200                 return NULL;
 201
 202         while (*str) {
 203                 int len;
 204
 205                 len = utf8_encoded_valid_unichar(str);
 206                 if (len > 0) {
 207                         s = mempcpy(s, str, len);
 208                         str += len;
 209                 } else {
 210                         s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
 211                         str += 1;
 212                 }
 213         }
 214
 215         *s = '\0';
 216
 217         return p;
 218 }
 219
 220 char *utf8_escape_non_printable(const char *str) {
 221         char *p, *s;
 222
 223         assert(str);
 224
 225         p = s = malloc(strlen(str) * 4 + 1);
 226         if (!p)
 227                 return NULL;
 228
 229         while (*str) {
 230                 int len;
 231
 232                 len = utf8_encoded_valid_unichar(str);
 233                 if (len > 0) {
 234                         if (utf8_is_printable(str, len)) {
 235                                 s = mempcpy(s, str, len);
 236                                 str += len;
 237                         } else {
 238                                 while (len > 0) {
 239                                         *(s++) = '\\';
 240                                         *(s++) = 'x';
 241                                         *(s++) = hexchar((int) *str >> 4);
 242                                         *(s++) = hexchar((int) *str);
 243
 244                                         str += 1;
 245                                         len--;
 246                                 }
 247                         }
 248                 } else {
 249                         s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
 250                         str += 1;
 251                 }
 252         }
 253
 254         *s = '\0';
 255
 256         return p;
 257 }
 258
 259 char *ascii_is_valid(const char *str) {
 260         const char *p;
 261
 262         assert(str);
 263
 264         for (p = str; *p; p++)
 265                 if ((unsigned char) *p >= 128)
 266                         return NULL;
 267
 268         return (char*) str;
 269 }
 270
 271 /**
 272  * utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
 273  * @out_utf8: output buffer of at least 4 bytes or NULL
 274  * @g: UCS-4 character to encode
 275  *
 276  * This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
 277  * The length of the character is returned. It is not zero-terminated! If the
 278  * output buffer is NULL, only the length is returned.
 279  *
 280  * Returns: The length in bytes that the UTF-8 representation does or would
 281  *          occupy.
 282  */
 283 size_t utf8_encode_unichar(char *out_utf8, char32_t g) {
 284
 285         if (g < (1 << 7)) {
 286                 if (out_utf8)
 287                         out_utf8[0] = g & 0x7f;
 288                 return 1;
 289         } else if (g < (1 << 11)) {
 290                 if (out_utf8) {
 291                         out_utf8[0] = 0xc0 | ((g >> 6) & 0x1f);
 292                         out_utf8[1] = 0x80 | (g & 0x3f);
 293                 }
 294                 return 2;
 295         } else if (g < (1 << 16)) {
 296                 if (out_utf8) {
 297                         out_utf8[0] = 0xe0 | ((g >> 12) & 0x0f);
 298                         out_utf8[1] = 0x80 | ((g >> 6) & 0x3f);
 299                         out_utf8[2] = 0x80 | (g & 0x3f);
 300                 }
 301                 return 3;
 302         } else if (g < (1 << 21)) {
 303                 if (out_utf8) {
 304                         out_utf8[0] = 0xf0 | ((g >> 18) & 0x07);
 305                         out_utf8[1] = 0x80 | ((g >> 12) & 0x3f);
 306                         out_utf8[2] = 0x80 | ((g >> 6) & 0x3f);
 307                         out_utf8[3] = 0x80 | (g & 0x3f);
 308                 }
 309                 return 4;
 310         }
 311
 312         return 0;
 313 }
 314
 315 char *utf16_to_utf8(const void *s, size_t length) {
 316         const uint8_t *f;
 317         char *r, *t;
 318
 319         r = new(char, (length * 4 + 1) / 2 + 1);
 320         if (!r)
 321                 return NULL;
 322
 323         f = s;
 324         t = r;
 325
 326         while (f < (const uint8_t*) s + length) {
 327                 char16_t w1, w2;
 328
 329                 /* see RFC 2781 section 2.2 */
 330
 331                 w1 = f[1] << 8 | f[0];
 332                 f += 2;
 333
 334                 if (!utf16_is_surrogate(w1)) {
 335                         t += utf8_encode_unichar(t, w1);
 336
 337                         continue;
 338                 }
 339
 340                 if (utf16_is_trailing_surrogate(w1))
 341                         continue;
 342                 else if (f >= (const uint8_t*) s + length)
 343                         break;
 344
 345                 w2 = f[1] << 8 | f[0];
 346                 f += 2;
 347
 348                 if (!utf16_is_trailing_surrogate(w2)) {
 349                         f -= 2;
 350                         continue;
 351                 }
 352
 353                 t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));
 354         }
 355
 356         *t = 0;
 357         return r;
 358 }
 359
 360 /* expected size used to encode one unicode char */
 361 static int utf8_unichar_to_encoded_len(char32_t unichar) {
 362
 363         if (unichar < 0x80)
 364                 return 1;
 365         if (unichar < 0x800)
 366                 return 2;
 367         if (unichar < 0x10000)
 368                 return 3;
 369         if (unichar < 0x200000)
 370                 return 4;
 371         if (unichar < 0x4000000)
 372                 return 5;
 373
 374         return 6;
 375 }
 376
 377 /* validate one encoded unicode char and return its length */
 378 int utf8_encoded_valid_unichar(const char *str) {
 379         int len, i, r;
 380         char32_t unichar;
 381
 382         assert(str);
 383
 384         len = utf8_encoded_expected_len(str);
 385         if (len == 0)
 386                 return -EINVAL;
 387
 388         /* ascii is valid */
 389         if (len == 1)
 390                 return 1;
 391
 392         /* check if expected encoded chars are available */
 393         for (i = 0; i < len; i++)
 394                 if ((str[i] & 0x80) != 0x80)
 395                         return -EINVAL;
 396
 397         r = utf8_encoded_to_unichar(str, &unichar);
 398         if (r < 0)
 399                 return r;
 400
 401         /* check if encoded length matches encoded value */
 402         if (utf8_unichar_to_encoded_len(unichar) != len)
 403                 return -EINVAL;
 404
 405         /* check if value has valid range */
 406         if (!unichar_is_valid(unichar))
 407                 return -EINVAL;
 408
 409         return len;
 410 }