uschar *
string_domain_utf8_to_alabel(const uschar * utf8, uschar ** err)
{
uschar * s1, * s;
int rc;
#ifdef SUPPORT_I18N_2008
/* Avoid lowercasing plain-ascii domains */
if (!string_is_utf8(utf8))
return string_copy(utf8);
/* Only lowercase is accepted by the library call. A pity since we lose
any mixed-case annotation. This does not really matter for a domain. */
{
uschar c;
for (s1 = s = US utf8; (c = *s1); s1++) if (!(c & 0x80) && isupper(c))
{
s = string_copy(utf8);
for (s1 = s + (s1 - utf8); (c = *s1); s1++) if (!(c & 0x80) && isupper(c))
*s1 = tolower(c);
break;
}
}
if ((rc = idn2_lookup_u8((const uint8_t *) s, &s1, IDN2_NFC_INPUT)) != IDN2_OK)
{
if (err) *err = US idn2_strerror(rc);
return NULL;
}
#else
s = US stringprep_utf8_nfkc_normalize(CCS utf8, -1);
if ( (rc = idna_to_ascii_8z(CCS s, CSS &s1, IDNA_ALLOW_UNASSIGNED))
!= IDNA_SUCCESS)
{
free(s);
if (err) *err = US idna_strerror(rc);
return NULL;
}
free(s);
#endif
s = string_copy(s1);
free(s1);
return s;
}
const char *wget_str_to_ascii(const char *src)
{
#ifdef WITH_LIBIDN2
if (wget_str_needs_encoding(src)) {
char *asc = NULL;
int rc;
#ifdef WITH_LIBUNISTRING
uint8_t *lower, resbuf[256];
size_t len = sizeof(resbuf) - 1; // leave space for additional \0 byte
// we need a conversion to lowercase
lower = u8_tolower((uint8_t *)src, u8_strlen((uint8_t *)src), 0, UNINORM_NFKC, resbuf, &len);
if (!lower) {
error_printf("u8_tolower(%s) failed (%d)\n", src, errno);
return src;
}
// u8_tolower() does not terminate the result string
if (lower == resbuf) {
lower[len]=0;
} else {
uint8_t *tmp = lower;
lower = (uint8_t *)wget_strmemdup((char *)lower, len);
xfree(tmp);
}
if ((rc = idn2_lookup_u8(lower, (uint8_t **)&asc, 0)) == IDN2_OK) {
debug_printf("idn2 '%s' -> '%s'\n", src, asc);
src = asc;
} else
error_printf(_("toASCII(%s) failed (%d): %s\n"), lower, rc, idn2_strerror(rc));
if (lower != resbuf)
xfree(lower);
#else
if ((rc = idn2_lookup_u8((uint8_t *)src, (uint8_t **)&asc, 0)) == IDN2_OK) {
debug_printf("idn2 '%s' -> '%s'\n", src, asc);
src = asc;
} else
error_printf(_("toASCII(%s) failed (%d): %s\n"), src, rc, idn2_strerror(rc));
#endif
}
#elif WITH_LIBIDN
if (wget_str_needs_encoding(src)) {
char *asc = NULL;
int rc;
if (_utf8_is_valid(src)) {
// idna_to_ascii_8z() automatically converts UTF-8 to lowercase
if ((rc = idna_to_ascii_8z(src, &asc, IDNA_USE_STD3_ASCII_RULES)) == IDNA_SUCCESS) {
// debug_printf("toASCII '%s' -> '%s'\n", src, asc);
src = asc;
} else
error_printf(_("toASCII failed (%d): %s\n"), rc, idna_strerror(rc));
}
else
error_printf(_("Invalid UTF-8 sequence not converted: '%s'\n"), src);
}
#else
if (wget_str_needs_encoding(src)) {
error_printf(_("toASCII not available: '%s'\n"), src);
}
#endif
return src;
}
