src/basic/utf8.c

   1 /* SPDX-License-Identifier: LGPL-2.1+ */
   2
   3 /* Parts of this file are based on the GLIB utf8 validation functions. The
   4  * original license text follows. */
   5
   6 /* gutf8.c - Operations on UTF-8 strings.
   7  *
   8  * Copyright (C) 1999 Tom Tromey
   9  * Copyright (C) 2000 Red Hat, Inc.
  10  *
  11  * This library is free software; you can redistribute it and/or
  12  * modify it under the terms of the GNU Library General Public
  13  * License as published by the Free Software Foundation; either
  14  * version 2 of the License, or (at your option) any later version.
  15  *
  16  * This library is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  19  * Library General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU Library General Public
  22  * License along with this library; if not, write to the Free Software
  23  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
  24  */
  25
  26 #include <errno.h>
  27 #include <stdbool.h>
  28 #include <stdlib.h>
  29 #include <string.h>
  30
  31 #include "alloc-util.h"
  32 #include "gunicode.h"
  33 #include "hexdecoct.h"
  34 #include "macro.h"
  35 #include "utf8.h"
  36
  37 bool unichar_is_valid(char32_t ch) {
  38
  39         if (ch >= 0x110000) /* End of unicode space */
  40                 return false;
  41         if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */
  42                 return false;
  43         if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */
  44                 return false;
  45         if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */
  46                 return false;
  47
  48         return true;
  49 }
  50
  51 static bool unichar_is_control(char32_t ch) {
  52
  53         /*
  54           0 to ' '-1 is the C0 range.
  55           DEL=0x7F, and DEL+1 to 0x9F is C1 range.
  56           '\t' is in C0 range, but more or less harmless and commonly used.
  57         */
  58
  59         return (ch < ' ' && !IN_SET(ch, '\t', '\n')) ||
  60                 (0x7F <= ch && ch <= 0x9F);
  61 }
  62
  63 /* count of characters used to encode one unicode char */
  64 static int utf8_encoded_expected_len(const char *str) {
  65         unsigned char c;
  66
  67         assert(str);
  68
  69         c = (unsigned char) str[0];
  70         if (c < 0x80)
  71                 return 1;
  72         if ((c & 0xe0) == 0xc0)
  73                 return 2;
  74         if ((c & 0xf0) == 0xe0)
  75                 return 3;
  76         if ((c & 0xf8) == 0xf0)
  77                 return 4;
  78         if ((c & 0xfc) == 0xf8)
  79                 return 5;
  80         if ((c & 0xfe) == 0xfc)
  81                 return 6;
  82
  83         return 0;
  84 }
  85
  86 /* decode one unicode char */
  87 int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) {
  88         char32_t unichar;
  89         int len, i;
  90
  91         assert(str);
  92
  93         len = utf8_encoded_expected_len(str);
  94
  95         switch (len) {
  96         case 1:
  97                 *ret_unichar = (char32_t)str[0];
  98                 return 0;
  99         case 2:
 100                 unichar = str[0] & 0x1f;
 101                 break;
 102         case 3:
 103                 unichar = (char32_t)str[0] & 0x0f;
 104                 break;
 105         case 4:
 106                 unichar = (char32_t)str[0] & 0x07;
 107                 break;
 108         case 5:
 109                 unichar = (char32_t)str[0] & 0x03;
 110                 break;
 111         case 6:
 112                 unichar = (char32_t)str[0] & 0x01;
 113                 break;
 114         default:
 115                 return -EINVAL;
 116         }
 117
 118         for (i = 1; i < len; i++) {
 119                 if (((char32_t)str[i] & 0xc0) != 0x80)
 120                         return -EINVAL;
 121                 unichar <<= 6;
 122                 unichar |= (char32_t)str[i] & 0x3f;
 123         }
 124
 125         *ret_unichar = unichar;
 126
 127         return 0;
 128 }
 129
 130 bool utf8_is_printable_newline(const char* str, size_t length, bool newline) {
 131         const char *p;
 132
 133         assert(str);
 134
 135         for (p = str; length;) {
 136                 int encoded_len, r;
 137                 char32_t val;
 138
 139                 encoded_len = utf8_encoded_valid_unichar(p);
 140                 if (encoded_len < 0 ||
 141                     (size_t) encoded_len > length)
 142                         return false;
 143
 144                 r = utf8_encoded_to_unichar(p, &val);
 145                 if (r < 0 ||
 146                     unichar_is_control(val) ||
 147                     (!newline && val == '\n'))
 148                         return false;
 149
 150                 length -= encoded_len;
 151                 p += encoded_len;
 152         }
 153
 154         return true;
 155 }
 156
 157 const char *utf8_is_valid(const char *str) {
 158         const uint8_t *p;
 159
 160         assert(str);
 161
 162         for (p = (const uint8_t*) str; *p; ) {
 163                 int len;
 164
 165                 len = utf8_encoded_valid_unichar((const char *)p);
 166                 if (len < 0)
 167                         return NULL;
 168
 169                 p += len;
 170         }
 171
 172         return str;
 173 }
 174
 175 char *utf8_escape_invalid(const char *str) {
 176         char *p, *s;
 177
 178         assert(str);
 179
 180         p = s = malloc(strlen(str) * 4 + 1);
 181         if (!p)
 182                 return NULL;
 183
 184         while (*str) {
 185                 int len;
 186
 187                 len = utf8_encoded_valid_unichar(str);
 188                 if (len > 0) {
 189                         s = mempcpy(s, str, len);
 190                         str += len;
 191                 } else {
 192                         s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
 193                         str += 1;
 194                 }
 195         }
 196
 197         *s = '\0';
 198
 199         return p;
 200 }
 201
 202 char *utf8_escape_non_printable(const char *str) {
 203         char *p, *s;
 204
 205         assert(str);
 206
 207         p = s = malloc(strlen(str) * 4 + 1);
 208         if (!p)
 209                 return NULL;
 210
 211         while (*str) {
 212                 int len;
 213
 214                 len = utf8_encoded_valid_unichar(str);
 215                 if (len > 0) {
 216                         if (utf8_is_printable(str, len)) {
 217                                 s = mempcpy(s, str, len);
 218                                 str += len;
 219                         } else {
 220                                 while (len > 0) {
 221                                         *(s++) = '\\';
 222                                         *(s++) = 'x';
 223                                         *(s++) = hexchar((int) *str >> 4);
 224                                         *(s++) = hexchar((int) *str);
 225
 226                                         str += 1;
 227                                         len--;
 228                                 }
 229                         }
 230                 } else {
 231                         s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
 232                         str += 1;
 233                 }
 234         }
 235
 236         *s = '\0';
 237
 238         return p;
 239 }
 240
 241 char *ascii_is_valid(const char *str) {
 242         const char *p;
 243
 244         /* Check whether the string consists of valid ASCII bytes,
 245          * i.e values between 0 and 127, inclusive. */
 246
 247         assert(str);
 248
 249         for (p = str; *p; p++)
 250                 if ((unsigned char) *p >= 128)
 251                         return NULL;
 252
 253         return (char*) str;
 254 }
 255
 256 char *ascii_is_valid_n(const char *str, size_t len) {
 257         size_t i;
 258
 259         /* Very similar to ascii_is_valid(), but checks exactly len
 260          * bytes and rejects any NULs in that range. */
 261
 262         assert(str);
 263
 264         for (i = 0; i < len; i++)
 265                 if ((unsigned char) str[i] >= 128 || str[i] == 0)
 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 }
 411
 412 size_t utf8_n_codepoints(const char *str) {
 413         size_t n = 0;
 414
 415         /* Returns the number of UTF-8 codepoints in this string, or (size_t) -1 if the string is not valid UTF-8. */
 416
 417         while (*str != 0) {
 418                 int k;
 419
 420                 k = utf8_encoded_valid_unichar(str);
 421                 if (k < 0)
 422                         return (size_t) -1;
 423
 424                 str += k;
 425                 n++;
 426         }
 427
 428         return n;
 429 }
 430
 431 size_t utf8_console_width(const char *str) {
 432         size_t n = 0;
 433
 434         /* Returns the approximate width a string will take on screen when printed on a character cell
 435          * terminal/console. */
 436
 437         while (*str != 0) {
 438                 char32_t c;
 439
 440                 if (utf8_encoded_to_unichar(str, &c) < 0)
 441                         return (size_t) -1;
 442
 443                 str = utf8_next_char(str);
 444
 445                 n += unichar_iswide(c) ? 2 : 1;
 446         }
 447
 448         return n;
 449 }