1 /* idna.c Convert to or from IDN strings.
2 * Copyright (C) 2002, 2003, 2004, 2011 Simon Josefsson
4 * This file is part of GNU Libidn.
6 * GNU Libidn 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.
11 * GNU Libidn 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.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with GNU Libidn; if not, see <http://www.gnu.org/licenses/>.
26 #include <stringprep.h>
32 #define DOTP(c) ((c) == 0x002E || (c) == 0x3002 || \
33 (c) == 0xFF0E || (c) == 0xFF61)
39 * @in: input array with unicode code points.
40 * @inlen: length of input array with unicode code points.
41 * @out: output zero terminated string that must have room for at
42 * least 63 characters plus the terminating zero.
43 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
45 * The ToASCII operation takes a sequence of Unicode code points that make
46 * up one label and transforms it into a sequence of code points in the
47 * ASCII range (0..7F). If ToASCII succeeds, the original sequence and the
48 * resulting sequence are equivalent labels.
50 * It is important to note that the ToASCII operation can fail. ToASCII
51 * fails if any step of it fails. If any step of the ToASCII operation
52 * fails on any label in a domain name, that domain name MUST NOT be used
53 * as an internationalized domain name. The method for deadling with this
54 * failure is application-specific.
56 * The inputs to ToASCII are a sequence of code points, the AllowUnassigned
57 * flag, and the UseSTD3ASCIIRules flag. The output of ToASCII is either a
58 * sequence of ASCII code points or a failure condition.
60 * ToASCII never alters a sequence of code points that are all in the ASCII
61 * range to begin with (although it could fail). Applying the ToASCII
62 * operation multiple times has exactly the same effect as applying it just
65 * Return value: Returns 0 on success, or an error code.
68 idna_to_ascii_4i (const uint32_t * in, size_t inlen, char *out, int flags)
71 uint32_t *src; /* XXX don't need to copy data? */
75 * ToASCII consists of the following steps:
77 * 1. If all code points in the sequence are in the ASCII range (0..7F)
78 * then skip to step 3.
86 for (i = 0; i < inlen; i++)
91 src = malloc (sizeof (in[0]) * (inlen + 1));
93 return IDNA_MALLOC_ERROR;
95 memcpy (src, in, sizeof (in[0]) * inlen);
103 * 2. Perform the steps specified in [NAMEPREP] and fail if there is
104 * an error. The AllowUnassigned flag is used in [NAMEPREP].
110 p = stringprep_ucs4_to_utf8 (in, inlen, NULL, NULL);
112 return IDNA_MALLOC_ERROR;
119 len = 2 * len + 10; /* XXX better guess? */
120 newp = realloc (p, len);
124 return IDNA_MALLOC_ERROR;
128 if (flags & IDNA_ALLOW_UNASSIGNED)
129 rc = stringprep_nameprep (p, len);
131 rc = stringprep_nameprep_no_unassigned (p, len);
133 while (rc == STRINGPREP_TOO_SMALL_BUFFER);
135 if (rc != STRINGPREP_OK)
138 return IDNA_STRINGPREP_ERROR;
141 src = stringprep_utf8_to_ucs4 (p, -1, NULL);
148 * 3. If the UseSTD3ASCIIRules flag is set, then perform these checks:
150 * (a) Verify the absence of non-LDH ASCII code points; that is,
151 * the absence of 0..2C, 2E..2F, 3A..40, 5B..60, and 7B..7F.
153 * (b) Verify the absence of leading and trailing hyphen-minus;
154 * that is, the absence of U+002D at the beginning and end of
158 if (flags & IDNA_USE_STD3_ASCII_RULES)
162 for (i = 0; src[i]; i++)
163 if (src[i] <= 0x2C || src[i] == 0x2E || src[i] == 0x2F ||
164 (src[i] >= 0x3A && src[i] <= 0x40) ||
165 (src[i] >= 0x5B && src[i] <= 0x60) ||
166 (src[i] >= 0x7B && src[i] <= 0x7F))
169 return IDNA_CONTAINS_NON_LDH;
172 if (src[0] == 0x002D || (i > 0 && src[i - 1] == 0x002D))
175 return IDNA_CONTAINS_MINUS;
180 * 4. If all code points in the sequence are in the ASCII range
181 * (0..7F), then skip to step 8.
189 for (i = 0; src[i]; i++)
193 /* copy string to output buffer if we are about to skip to step8 */
204 * 5. Verify that the sequence does NOT begin with the ACE prefix.
213 for (i = 0; match && i < strlen (IDNA_ACE_PREFIX); i++)
214 if (((uint32_t) IDNA_ACE_PREFIX[i] & 0xFF) != src[i])
219 return IDNA_CONTAINS_ACE_PREFIX;
224 * 6. Encode the sequence using the encoding algorithm in [PUNYCODE]
225 * and fail if there is an error.
227 for (len = 0; src[len]; len++)
230 outlen = 63 - strlen (IDNA_ACE_PREFIX);
231 rc = punycode_encode (len, src, NULL,
232 &outlen, &out[strlen (IDNA_ACE_PREFIX)]);
233 if (rc != PUNYCODE_SUCCESS)
236 return IDNA_PUNYCODE_ERROR;
238 out[strlen (IDNA_ACE_PREFIX) + outlen] = '\0';
241 * 7. Prepend the ACE prefix.
244 memcpy (out, IDNA_ACE_PREFIX, strlen (IDNA_ACE_PREFIX));
247 * 8. Verify that the number of code points is in the range 1 to 63
248 * inclusive (0 is excluded).
253 if (strlen (out) < 1 || strlen (out) > 63)
254 return IDNA_INVALID_LENGTH;
259 /* ToUnicode(). May realloc() utf8in. */
261 idna_to_unicode_internal (char *utf8in,
262 uint32_t * out, size_t * outlen, int flags)
266 size_t utf8len = strlen (utf8in) + 1;
270 * ToUnicode consists of the following steps:
272 * 1. If the sequence contains any code points outside the ASCII range
273 * (0..7F) then proceed to step 2, otherwise skip to step 3.
281 for (i = 0; utf8in[i]; i++)
282 if (utf8in[i] & ~0x7F)
289 * 2. Perform the steps specified in [NAMEPREP] and fail if there is an
290 * error. (If step 3 of ToASCII is also performed here, it will not
291 * affect the overall behavior of ToUnicode, but it is not
292 * necessary.) The AllowUnassigned flag is used in [NAMEPREP].
296 char *newp = realloc (utf8in, utf8len + addlen);
300 return IDNA_MALLOC_ERROR;
303 if (flags & IDNA_ALLOW_UNASSIGNED)
304 rc = stringprep_nameprep (utf8in, utf8len + addlen);
306 rc = stringprep_nameprep_no_unassigned (utf8in, utf8len + addlen);
309 while (rc == STRINGPREP_TOO_SMALL_BUFFER);
311 if (rc != STRINGPREP_OK)
314 return IDNA_STRINGPREP_ERROR;
317 /* 3. Verify that the sequence begins with the ACE prefix, and save a
318 * copy of the sequence.
322 if (memcmp (IDNA_ACE_PREFIX, utf8in, strlen (IDNA_ACE_PREFIX)) != 0)
325 return IDNA_NO_ACE_PREFIX;
328 /* 4. Remove the ACE prefix.
331 memmove (utf8in, &utf8in[strlen (IDNA_ACE_PREFIX)],
332 strlen (utf8in) - strlen (IDNA_ACE_PREFIX) + 1);
334 /* 5. Decode the sequence using the decoding algorithm in [PUNYCODE]
335 * and fail if there is an error. Save a copy of the result of
339 (*outlen)--; /* reserve one for the zero */
341 rc = punycode_decode (strlen (utf8in), utf8in, outlen, out, NULL);
342 if (rc != PUNYCODE_SUCCESS)
345 return IDNA_PUNYCODE_ERROR;
348 out[*outlen] = 0; /* add zero */
353 rc = idna_to_ascii_4i (out, *outlen, tmpout, flags);
354 if (rc != IDNA_SUCCESS)
360 /* 7. Verify that the result of step 6 matches the saved copy from
361 * step 3, using a case-insensitive ASCII comparison.
364 if (strcasecmp (utf8in, tmpout + strlen (IDNA_ACE_PREFIX)) != 0)
367 return IDNA_ROUNDTRIP_VERIFY_ERROR;
370 /* 8. Return the saved copy from step 5.
378 * idna_to_unicode_44i
379 * @in: input array with unicode code points.
380 * @inlen: length of input array with unicode code points.
381 * @out: output array with unicode code points.
382 * @outlen: on input, maximum size of output array with unicode code points,
383 * on exit, actual size of output array with unicode code points.
384 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
386 * The ToUnicode operation takes a sequence of Unicode code points
387 * that make up one label and returns a sequence of Unicode code
388 * points. If the input sequence is a label in ACE form, then the
389 * result is an equivalent internationalized label that is not in ACE
390 * form, otherwise the original sequence is returned unaltered.
392 * ToUnicode never fails. If any step fails, then the original input
393 * sequence is returned immediately in that step.
395 * The Punycode decoder can never output more code points than it
396 * inputs, but Nameprep can, and therefore ToUnicode can. Note that
397 * the number of octets needed to represent a sequence of code points
398 * depends on the particular character encoding used.
400 * The inputs to ToUnicode are a sequence of code points, the
401 * AllowUnassigned flag, and the UseSTD3ASCIIRules flag. The output of
402 * ToUnicode is always a sequence of Unicode code points.
404 * Return value: Returns error condition, but it must only be used for
405 * debugging purposes. The output buffer is always
406 * guaranteed to contain the correct data according to
407 * the specification (sans malloc induced errors). NB!
408 * This means that you normally ignore the return code
409 * from this function, as checking it means breaking the
413 idna_to_unicode_44i (const uint32_t * in, size_t inlen,
414 uint32_t * out, size_t * outlen, int flags)
417 size_t outlensave = *outlen;
420 p = stringprep_ucs4_to_utf8 (in, inlen, NULL, NULL);
422 return IDNA_MALLOC_ERROR;
424 rc = idna_to_unicode_internal (p, out, outlen, flags);
425 if (rc != IDNA_SUCCESS)
427 memcpy (out, in, sizeof (in[0]) * (inlen < outlensave ?
428 inlen : outlensave));
432 /* p is freed in idna_to_unicode_internal. */
437 /* Wrappers that handle several labels */
441 * @input: zero terminated input Unicode string.
442 * @output: pointer to newly allocated output string.
443 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
445 * Convert UCS-4 domain name to ASCII string. The domain name may
446 * contain several labels, separated by dots. The output buffer must
447 * be deallocated by the caller.
449 * Return value: Returns IDNA_SUCCESS on success, or error code.
452 idna_to_ascii_4z (const uint32_t * input, char **output, int flags)
454 const uint32_t *start = input;
455 const uint32_t *end = input;
460 /* 1) Whenever dots are used as label separators, the following
461 characters MUST be recognized as dots: U+002E (full stop),
462 U+3002 (ideographic full stop), U+FF0E (fullwidth full stop),
463 U+FF61 (halfwidth ideographic full stop). */
467 /* Handle implicit zero-length root label. */
468 *output = malloc (1);
470 return IDNA_MALLOC_ERROR;
471 strcpy (*output, "");
475 if (DOTP (input[0]) && input[1] == 0)
477 /* Handle explicit zero-length root label. */
478 *output = malloc (2);
480 return IDNA_MALLOC_ERROR;
481 strcpy (*output, ".");
490 for (; *end && !DOTP (*end); end++)
493 if (*end == '\0' && start == end)
495 /* Handle explicit zero-length root label. */
500 rc = idna_to_ascii_4i (start, end - start, buf, flags);
501 if (rc != IDNA_SUCCESS)
507 char *newp = realloc (out, strlen (out) + 1 + strlen (buf) + 1);
511 return IDNA_MALLOC_ERROR;
519 out = (char *) malloc (strlen (buf) + 1);
521 return IDNA_MALLOC_ERROR;
536 * @input: zero terminated input UTF-8 string.
537 * @output: pointer to newly allocated output string.
538 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
540 * Convert UTF-8 domain name to ASCII string. The domain name may
541 * contain several labels, separated by dots. The output buffer must
542 * be deallocated by the caller.
544 * Return value: Returns IDNA_SUCCESS on success, or error code.
547 idna_to_ascii_8z (const char *input, char **output, int flags)
553 ucs4 = stringprep_utf8_to_ucs4 (input, -1, &ucs4len);
555 return IDNA_ICONV_ERROR;
557 rc = idna_to_ascii_4z (ucs4, output, flags);
567 * @input: zero terminated input UTF-8 string.
568 * @output: pointer to newly allocated output string.
569 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
571 * Convert domain name in the locale's encoding to ASCII string. The
572 * domain name may contain several labels, separated by dots. The
573 * output buffer must be deallocated by the caller.
575 * Return value: Returns IDNA_SUCCESS on success, or error code.
578 idna_to_ascii_lz (const char *input, char **output, int flags)
583 utf8 = stringprep_locale_to_utf8 (input);
585 return IDNA_ICONV_ERROR;
587 rc = idna_to_ascii_8z (utf8, output, flags);
595 * idna_to_unicode_4z4z:
596 * @input: zero-terminated Unicode string.
597 * @output: pointer to newly allocated output Unicode string.
598 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
600 * Convert possibly ACE encoded domain name in UCS-4 format into a
601 * UCS-4 string. The domain name may contain several labels,
602 * separated by dots. The output buffer must be deallocated by the
605 * Return value: Returns IDNA_SUCCESS on success, or error code.
608 idna_to_unicode_4z4z (const uint32_t * input, uint32_t ** output, int flags)
610 const uint32_t *start = input;
611 const uint32_t *end = input;
614 uint32_t *out = NULL;
623 for (; *end && !DOTP (*end); end++)
626 buflen = end - start;
627 buf = malloc (sizeof (buf[0]) * (buflen + 1));
629 return IDNA_MALLOC_ERROR;
631 idna_to_unicode_44i (start, end - start, buf, &buflen, flags);
632 /* don't check return value as per specification! */
636 uint32_t *newp = realloc (out,
638 * (outlen + 1 + buflen + 1));
643 return IDNA_MALLOC_ERROR;
646 out[outlen++] = 0x002E; /* '.' (full stop) */
647 memcpy (out + outlen, buf, sizeof (buf[0]) * buflen);
669 * idna_to_unicode_8z4z:
670 * @input: zero-terminated UTF-8 string.
671 * @output: pointer to newly allocated output Unicode string.
672 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
674 * Convert possibly ACE encoded domain name in UTF-8 format into a
675 * UCS-4 string. The domain name may contain several labels,
676 * separated by dots. The output buffer must be deallocated by the
679 * Return value: Returns IDNA_SUCCESS on success, or error code.
682 idna_to_unicode_8z4z (const char *input, uint32_t ** output, int flags)
688 ucs4 = stringprep_utf8_to_ucs4 (input, -1, &ucs4len);
690 return IDNA_ICONV_ERROR;
692 rc = idna_to_unicode_4z4z (ucs4, output, flags);
699 * idna_to_unicode_8z8z:
700 * @input: zero-terminated UTF-8 string.
701 * @output: pointer to newly allocated output UTF-8 string.
702 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
704 * Convert possibly ACE encoded domain name in UTF-8 format into a
705 * UTF-8 string. The domain name may contain several labels,
706 * separated by dots. The output buffer must be deallocated by the
709 * Return value: Returns IDNA_SUCCESS on success, or error code.
712 idna_to_unicode_8z8z (const char *input, char **output, int flags)
717 rc = idna_to_unicode_8z4z (input, &ucs4, flags);
718 *output = stringprep_ucs4_to_utf8 (ucs4, -1, NULL, NULL);
722 return IDNA_ICONV_ERROR;
728 * idna_to_unicode_8zlz:
729 * @input: zero-terminated UTF-8 string.
730 * @output: pointer to newly allocated output string encoded in the
731 * current locale's character set.
732 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
734 * Convert possibly ACE encoded domain name in UTF-8 format into a
735 * string encoded in the current locale's character set. The domain
736 * name may contain several labels, separated by dots. The output
737 * buffer must be deallocated by the caller.
739 * Return value: Returns IDNA_SUCCESS on success, or error code.
742 idna_to_unicode_8zlz (const char *input, char **output, int flags)
747 rc = idna_to_unicode_8z8z (input, &utf8, flags);
748 *output = stringprep_utf8_to_locale (utf8);
752 return IDNA_ICONV_ERROR;
758 * idna_to_unicode_lzlz:
759 * @input: zero-terminated string encoded in the current locale's
761 * @output: pointer to newly allocated output string encoded in the
762 * current locale's character set.
763 * @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
765 * Convert possibly ACE encoded domain name in the locale's character
766 * set into a string encoded in the current locale's character set.
767 * The domain name may contain several labels, separated by dots. The
768 * output buffer must be deallocated by the caller.
770 * Return value: Returns IDNA_SUCCESS on success, or error code.
773 idna_to_unicode_lzlz (const char *input, char **output, int flags)
778 utf8 = stringprep_locale_to_utf8 (input);
780 return IDNA_ICONV_ERROR;
782 rc = idna_to_unicode_8zlz (utf8, output, flags);
791 * The IANA allocated prefix to use for IDNA. "xn--"
796 * @IDNA_SUCCESS: Successful operation. This value is guaranteed to
797 * always be zero, the remaining ones are only guaranteed to hold
798 * non-zero values, for logical comparison purposes.
799 * @IDNA_STRINGPREP_ERROR: Error during string preparation.
800 * @IDNA_PUNYCODE_ERROR: Error during punycode operation.
801 * @IDNA_CONTAINS_NON_LDH: For IDNA_USE_STD3_ASCII_RULES, indicate that
802 * the string contains non-LDH ASCII characters.
803 * @IDNA_CONTAINS_MINUS: For IDNA_USE_STD3_ASCII_RULES, indicate that
804 * the string contains a leading or trailing hyphen-minus (U+002D).
805 * @IDNA_INVALID_LENGTH: The final output string is not within the
806 * (inclusive) range 1 to 63 characters.
807 * @IDNA_NO_ACE_PREFIX: The string does not contain the ACE prefix
809 * @IDNA_ROUNDTRIP_VERIFY_ERROR: The ToASCII operation on output
810 * string does not equal the input.
811 * @IDNA_CONTAINS_ACE_PREFIX: The input contains the ACE prefix (for
813 * @IDNA_ICONV_ERROR: Could not convert string in locale encoding.
814 * @IDNA_MALLOC_ERROR: Could not allocate buffer (this is typically a
816 * @IDNA_DLOPEN_ERROR: Could not dlopen the libcidn DSO (only used
817 * internally in libc).
819 * Enumerated return codes of idna_to_ascii_4i(),
820 * idna_to_unicode_44i() functions (and functions derived from those
821 * functions). The value 0 is guaranteed to always correspond to
828 * @IDNA_ALLOW_UNASSIGNED: Don't reject strings containing unassigned
829 * Unicode code points.
830 * @IDNA_USE_STD3_ASCII_RULES: Validate strings according to STD3
831 * rules (i.e., normal host name rules).
833 * Flags to pass to idna_to_ascii_4i(), idna_to_unicode_44i() etc.