/**
* idna_to_unicode_8z8z:
* @input: zero-terminated UTF-8 string.
* @output: pointer to newly allocated output UTF-8 string.
* @flags: IDNA flags, e.g. IDNA_ALLOW_UNASSIGNED or IDNA_USE_STD3_ASCII_RULES.
*
* Convert possibly ACE encoded domain name in UTF-8 format into a
* UTF-8 string. The domain name may contain several labels,
* separated by dots. The output buffer must be deallocated by the
* caller.
*
* Return value: Returns IDNA_SUCCESS on success, or error code.
**/
int
idna_to_unicode_8z8z (const char *input, char **output, int flags)
{
uint32_t *ucs4;
int rc;
rc = idna_to_unicode_8z4z (input, &ucs4, flags);
*output = stringprep_ucs4_to_utf8 (ucs4, -1, NULL, NULL);
free (ucs4);
if (!*output)
return IDNA_ICONV_ERROR;
return rc;
}
int
main (int argc, char *argv[])
{
struct gengetopt_args_info args_info;
char readbuf[BUFSIZ];
char *p, *r;
uint32_t *q;
unsigned cmdn = 0;
int rc;
setlocale (LC_ALL, "");
bindtextdomain (PACKAGE, LOCALEDIR);
textdomain (PACKAGE);
if (cmdline_parser (argc, argv, &args_info) != 0)
return 1;
if (!args_info.stringprep_given &&
!args_info.punycode_encode_given && !args_info.punycode_decode_given &&
!args_info.idna_to_ascii_given && !args_info.idna_to_unicode_given)
args_info.idna_to_ascii_given = 1;
if ((args_info.stringprep_given ? 1 : 0) +
(args_info.punycode_encode_given ? 1 : 0) +
(args_info.punycode_decode_given ? 1 : 0) +
(args_info.idna_to_ascii_given ? 1 : 0) +
(args_info.idna_to_unicode_given ? 1 : 0) != 1)
{
fprintf (stderr,
_("%s: Only one of -s, -e, -d, -a or -u can be specified.\n"),
argv[0]);
cmdline_parser_print_help ();
return 1;
}
if (!args_info.quiet_given)
fprintf (stderr, "%s %s\n" GREETING, PACKAGE, VERSION);
if (args_info.debug_given)
fprintf (stderr, _("Charset `%s'.\n"), stringprep_locale_charset ());
if (!args_info.quiet_given && args_info.inputs_num == 0)
fprintf (stderr, _("Type each input string on a line by itself, "
"terminated by a newline character.\n"));
do
{
if (cmdn < args_info.inputs_num)
{
strncpy (readbuf, args_info.inputs[cmdn++], BUFSIZ - 1);
readbuf[BUFSIZ - 1] = '\0';
}
else if (fgets (readbuf, BUFSIZ, stdin) == NULL)
{
sprintf (readbuf, _("%s: fgets() failed: "), argv[0]);
if (!feof (stdin))
perror (readbuf);
return 1;
}
if (readbuf[strlen (readbuf) - 1] == '\n')
readbuf[strlen (readbuf) - 1] = '\0';
if (args_info.stringprep_given)
{
p = stringprep_locale_to_utf8 (readbuf);
if (!p)
{
fprintf (stderr, _("%s: could not convert from %s to UTF-8.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
q = stringprep_utf8_to_ucs4 (p, -1, NULL);
if (!q)
{
free (p);
fprintf (stderr,
_("%s: could not convert from UTF-8 to UCS-4.\n"),
argv[0]);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; q[i]; i++)
fprintf (stderr, _("input[%d] = U+%04x\n"), i, q[i]);
}
free (q);
rc = stringprep_profile (p, &r,
args_info.profile_given ?
args_info.profile_arg : "Nameprep", 0);
free (p);
if (rc != STRINGPREP_OK)
{
fprintf (stderr,
_("%s: stringprep_profile() failed with error %d.\n"),
argv[0], rc);
return 1;
}
q = stringprep_utf8_to_ucs4 (r, -1, NULL);
if (!q)
{
free (r);
fprintf (stderr,
_("%s: could not convert from UTF-8 to UCS-4.\n"),
argv[0]);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; q[i]; i++)
fprintf (stderr, _("output[%d] = U+%04x\n"), i, q[i]);
}
free (q);
p = stringprep_utf8_to_locale (r);
free (r);
if (!p)
{
fprintf (stderr, _("%s: could not convert from UTF-8 to %s.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
fprintf (stdout, "%s\n", p);
free (p);
}
if (args_info.punycode_encode_given)
{
size_t len, len2;
p = stringprep_locale_to_utf8 (readbuf);
if (!p)
{
fprintf (stderr, _("%s: could not convert from %s to UTF-8.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
q = stringprep_utf8_to_ucs4 (p, -1, &len);
free (p);
if (!q)
{
fprintf (stderr,
_("%s: could not convert from UTF-8 to UCS-4.\n"),
argv[0]);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; i < len; i++)
fprintf (stderr, _("input[%d] = U+%04x\n"), i, q[i]);
}
len2 = BUFSIZ;
rc = punycode_encode (len, q, NULL, &len2, readbuf);
free (q);
if (rc != PUNYCODE_SUCCESS)
{
fprintf (stderr,
_("%s: punycode_encode() failed with error %d.\n"),
argv[0], rc);
return 1;
}
readbuf[len2] = '\0';
p = stringprep_utf8_to_locale (readbuf);
if (!p)
{
fprintf (stderr, _("%s: could not convert from UTF-8 to %s.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
fprintf (stdout, "%s\n", p);
free (p);
}
if (args_info.punycode_decode_given)
{
size_t len;
len = BUFSIZ;
q = (uint32_t *) malloc (len * sizeof (q[0]));
if (!q)
{
sprintf (readbuf, _("%s: malloc() failed: "), argv[0]);
perror (readbuf);
return 1;
}
rc = punycode_decode (strlen (readbuf), readbuf, &len, q, NULL);
if (rc != PUNYCODE_SUCCESS)
{
free (q);
fprintf (stderr,
_("%s: punycode_decode() failed with error %d.\n"),
argv[0], rc);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; i < len; i++)
fprintf (stderr, _("output[%d] = U+%04x\n"), i, q[i]);
}
q[len] = 0;
r = stringprep_ucs4_to_utf8 (q, -1, NULL, NULL);
free (q);
if (!r)
{
fprintf (stderr,
_("%s: could not convert from UCS-4 to UTF-8.\n"),
argv[0]);
return 1;
}
p = stringprep_utf8_to_locale (r);
free (r);
if (!r)
{
fprintf (stderr, _("%s: could not convert from UTF-8 to %s.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
fprintf (stdout, "%s\n", p);
free (p);
}
if (args_info.idna_to_ascii_given)
{
p = stringprep_locale_to_utf8 (readbuf);
if (!p)
{
fprintf (stderr, _("%s: could not convert from %s to UTF-8.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
q = stringprep_utf8_to_ucs4 (p, -1, NULL);
free (p);
if (!q)
{
fprintf (stderr,
_("%s: could not convert from UCS-4 to UTF-8.\n"),
argv[0]);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; q[i]; i++)
fprintf (stderr, _("input[%d] = U+%04x\n"), i, q[i]);
}
rc = idna_to_ascii_4z (q, &p,
(args_info.allow_unassigned_given ?
IDNA_ALLOW_UNASSIGNED : 0) |
(args_info.usestd3asciirules_given ?
IDNA_USE_STD3_ASCII_RULES : 0));
free (q);
if (rc != IDNA_SUCCESS)
{
fprintf (stderr, _("%s: idna_to_ascii_4z() failed "
"with error %d.\n"), argv[0], rc);
return 1;
}
#ifdef WITH_TLD
if (args_info.tld_flag)
{
size_t errpos;
rc = idna_to_unicode_8z4z (p, &q,
(args_info.allow_unassigned_given ?
IDNA_ALLOW_UNASSIGNED : 0) |
(args_info.usestd3asciirules_given ?
IDNA_USE_STD3_ASCII_RULES : 0));
if (rc != IDNA_SUCCESS)
{
fprintf (stderr, _("%s: TLD idna_to_unicode_8z8z() failed "
"with error %d.\n"), argv[0], rc);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; q[i]; i++)
fprintf (stderr, _("tld[%d] = U+%04x\n"), i, q[i]);
}
rc = tld_check_4z (q, &errpos, NULL);
if (rc == TLD_INVALID)
{
fprintf (stderr, _("%s: string rejected by TLD test "
"(Unicode position %d)\n"), argv[0],
errpos);
free (q);
return 1;
}
if (rc != TLD_SUCCESS)
{
fprintf (stderr,
_("%s: tld_check_4z failed with error %d.\n"),
argv[0], rc);
free (q);
return 1;
}
free (r);
}
#endif
if (args_info.debug_given)
{
size_t i;
for (i = 0; p[i]; i++)
fprintf (stderr, _("output[%d] = U+%04x\n"), i, p[i]);
}
fprintf (stdout, "%s\n", p);
free (p);
}
if (args_info.idna_to_unicode_given)
{
p = stringprep_locale_to_utf8 (readbuf);
if (!p)
{
fprintf (stderr, _("%s: could not convert from %s to UTF-8.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
q = stringprep_utf8_to_ucs4 (p, -1, NULL);
if (!q)
{
free (p);
fprintf (stderr,
_("%s: could not convert from UCS-4 to UTF-8.\n"),
argv[0]);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; q[i]; i++)
fprintf (stderr, _("input[%d] = U+%04x\n"), i, q[i]);
}
free (q);
rc = idna_to_unicode_8z4z (p, &q,
(args_info.allow_unassigned_given ?
IDNA_ALLOW_UNASSIGNED : 0) |
(args_info.usestd3asciirules_given ?
IDNA_USE_STD3_ASCII_RULES : 0));
free (p);
if (rc != IDNA_SUCCESS)
{
fprintf (stderr, _("%s: idna_to_unicode_8z4z() "
"failed with error %d.\n"), argv[0], rc);
return 1;
}
if (args_info.debug_given)
{
size_t i;
for (i = 0; q[i]; i++)
fprintf (stderr, _("output[%d] = U+%04x\n"), i, q[i]);
}
#ifdef WITH_TLD
if (args_info.tld_flag)
{
size_t errpos;
rc = tld_check_4z (q, &errpos, NULL);
if (rc == TLD_INVALID)
{
fprintf (stderr, _("%s: string rejected by TLD test "
"(Unicode position %d)\n"), argv[0],
errpos);
free (q);
return 1;
}
if (rc != TLD_SUCCESS)
{
fprintf (stderr,
_("%s: tld_check_4z failed with error %d.\n"),
argv[0], rc);
free (q);
return 1;
}
}
#endif
r = stringprep_ucs4_to_utf8 (q, -1, NULL, NULL);
free (q);
if (!r)
{
fprintf (stderr,
_("%s: could not convert from UTF-8 to UCS-4.\n"),
argv[0]);
return 1;
}
p = stringprep_utf8_to_locale (r);
free (r);
if (!r)
{
fprintf (stderr, _("%s: could not convert from UTF-8 to %s.\n"),
argv[0], stringprep_locale_charset ());
return 1;
}
fprintf (stdout, "%s\n", p);
free (p);
}
}
while (!feof (stdin) && !ferror (stdin) && (args_info.inputs_num == 0 ||
cmdn < args_info.inputs_num));
return 0;
}
int
main (int argc, char *argv[])
{
char buf[BUFSIZ];
char *p;
uint32_t *r;
int rc;
size_t errpos, i;
printf ("Input domain encoded as `%s': ", stringprep_locale_charset ());
fflush (stdout);
fgets (buf, BUFSIZ, stdin);
buf[strlen (buf) - 1] = '\0';
printf ("Read string (length %d): ", strlen (buf));
for (i = 0; i < strlen (buf); i++)
printf ("%02x ", buf[i] & 0xFF);
printf ("\n");
p = stringprep_locale_to_utf8 (buf);
if (p)
{
strcpy (buf, p);
free (p);
}
else
printf ("Could not convert string to UTF-8, continuing anyway...\n");
rc = idna_to_ascii_8z (buf, &p, 0);
if (rc != IDNA_SUCCESS)
{
printf ("idna_to_ascii_8z failed (%d): %s\n", rc, idna_strerror (rc));
return 2;
}
printf ("ToASCII string (length %d): %s\n", strlen (p), p);
rc = idna_to_unicode_8z4z (p, &r, 0);
free (p);
if (rc != IDNA_SUCCESS)
{
printf ("idna_to_unicode_8z4z failed (%d): %s\n",
rc, idna_strerror (rc));
return 2;
}
printf ("ToUnicode string: ");
for (i = 0; r[i]; i++)
printf ("U+%04x ", r[i]);
printf ("\n");
rc = tld_check_4z (r, &errpos, NULL);
free (r);
if (rc == TLD_INVALID)
{
printf ("Domain rejected by TLD check, Unicode position %d\n", errpos);
return 1;
}
else if (rc != TLD_SUCCESS)
{
printf ("tld_check_4z() failed (%d): %s\n", rc, tld_strerror (rc));
return 2;
}
printf ("Domain accepted by TLD check\n");
return 0;
}
