static idn_result_t
converter_none_convfromucs4(idn_converter_t ctx, void *privdata,
const unsigned long *from, char *to, size_t tolen) {
assert(ctx != NULL && from != NULL && to != NULL);
return idn_ucs4_ucs4toutf8(from, to, tolen);
}
static idn_result_t
label_lencheck_nonace(idn_resconf_t ctx, labellist_t label) {
idn_converter_t idn_converter;
const unsigned long *from;
size_t to_length;
idn_result_t r;
char *buffer = NULL;
size_t buffer_length;
from = labellist_getname(label);
TRACE(("res lencheck(label=\"%s\")\n",
idn__debug_ucs4xstring(from, 50)));
buffer_length = idn_ucs4_strlen(from) * 4 + 16; /* 16 for margin */
idn_converter = idn_resconf_getidnconverter(ctx);
for (;;) {
void *new_buffer;
new_buffer = realloc(buffer, sizeof(*buffer) * buffer_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
buffer = (char *)new_buffer;
if (idn_converter != NULL) {
r = idn_converter_convfromucs4(idn_converter, from,
buffer, buffer_length);
} else {
r = idn_ucs4_ucs4toutf8(from, buffer, buffer_length);
}
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
buffer_length *= 2;
}
to_length = strlen(buffer);
if (to_length == 0 || to_length > MAX_LABEL_LENGTH) {
r = idn_invalid_length;
goto ret;
}
r = idn_success;
ret:
TRACE(("res lencheck(): %s\n", idn_result_tostring(r)));
if (idn_converter != NULL)
idn_converter_destroy(idn_converter);
free(buffer);
return (r);
}
static idn_result_t
converter_iconv_convfromucs4(idn_converter_t ctx, void *privdata,
const unsigned long *from, char *to,
size_t tolen) {
iconv_t ictx;
char *utf8 = NULL;
size_t utf8size = 256; /* large enough */
idn_result_t r;
size_t sz;
size_t inleft;
size_t outleft;
char *inbuf, *outbuf;
assert(ctx != NULL && from != NULL && to != NULL);
if (tolen <= 0) {
r = idn_buffer_overflow; /* need space for NUL */
goto ret;
}
/*
* UCS4 -> UTF-8 conversion.
*/
utf8 = (char *)malloc(utf8size);
if (utf8 == NULL) {
r = idn_nomemory;
goto ret;
}
try_again:
r = idn_ucs4_ucs4toutf8(from, utf8, utf8size);
if (r == idn_buffer_overflow) {
char *new_utf8;
utf8size *= 2;
new_utf8 = (char *)realloc(utf8, utf8size);
if (new_utf8 == NULL) {
r = idn_nomemory;
goto ret;
}
utf8 = new_utf8;
goto try_again;
} else if (r != idn_success) {
goto ret;
}
ictx = ((iconv_t *)privdata)[0];
/*
* Reset internal state.
*
* The following code should work according to the SUSv2 spec,
* but causes segmentation fault with Solaris 2.6.
* So.. a work-around.
*
* (void)iconv(ictx, (const char **)NULL, (size_t *)NULL,
* (char **)NULL, (size_t *)NULL);
*/
inleft = 0;
outbuf = NULL;
outleft = 0;
(void)iconv(ictx, (const char **)NULL, &inleft, &outbuf, &outleft);
inleft = strlen(utf8);
inbuf = utf8;
outleft = tolen - 1; /* reserve space for terminating NUL */
sz = iconv(ictx, (const char **)&inbuf, &inleft, &to, &outleft);
if (sz == (size_t)(-1) || inleft > 0) {
switch (errno) {
case EILSEQ:
case EINVAL:
/*
* We already checked the validity of the input
* string. So we assume a mapping error.
*/
r = idn_nomapping;
goto ret;
case E2BIG:
r = idn_buffer_overflow;
goto ret;
default:
WARNING(("iconv failed with errno %d\n", errno));
r = idn_failure;
goto ret;
}
}
/*
* For UTF-8 -> local conversion, append a sequence of
* state reset.
*/
inleft = 0;
sz = iconv(ictx, (const char **)NULL, &inleft, &to, &outleft);
if (sz == (size_t)(-1)) {
switch (errno) {
case EILSEQ:
case EINVAL:
r = idn_invalid_encoding;
goto ret;
case E2BIG:
r = idn_buffer_overflow;
goto ret;
default:
WARNING(("iconv failed with errno %d\n", errno));
r = idn_failure;
goto ret;
}
}
*to = '\0';
r = idn_success;
ret:
free(utf8);
return (r);
}
idn_result_t
idn_res_decodename2(idn_resconf_t ctx, idn_action_t actions, const char *from,
char *to, size_t tolen, const char *auxencoding) {
#ifdef WITHOUT_ICONV
return idn_failure;
#else /* WITHOUT_ICONV */
idn_result_t r;
idn_converter_t aux_converter = NULL;
unsigned long *buffer_ucs4 = NULL;
char *buffer_utf8 = NULL;
size_t buffer_length;
assert(ctx != NULL && from != NULL && to != NULL);
TRACE(("idn_res_decodename2(actions=%s, from=\"%s\", tolen=%d, "
"auxencoding=\"%s\")\n",
idn__res_actionstostring(actions),
idn__debug_xstring(from, 50), (int)tolen,
(auxencoding != NULL) ? auxencoding : "<null>"));
if (!initialized)
idn_res_initialize();
if (!enabled || actions == 0) {
r = copy_verbatim(from, to, tolen);
goto ret;
} else if (tolen <= 0) {
r = idn_buffer_overflow;
goto ret;
}
if (auxencoding == NULL ||
strcmp(auxencoding, IDN_UTF8_ENCODING_NAME) == 0 ||
strcmp(auxencoding, "UTF-8") == 0) {
return idn_res_decodename(ctx, actions, from, to, tolen);
}
/*
* Convert `from' to UCS4.
*/
r = idn_resconf_setauxidnconvertername(ctx, auxencoding,
IDN_CONVERTER_DELAYEDOPEN);
if (r != idn_success) {
goto ret;
}
aux_converter = idn_resconf_getauxidnconverter(ctx);
if (aux_converter == NULL) {
r = idn_failure;
goto ret;
}
buffer_length = tolen * 2;
for (;;) {
void *new_buffer;
new_buffer = realloc(buffer_ucs4,
sizeof(*buffer_ucs4) * buffer_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
buffer_ucs4 = (unsigned long *)new_buffer;
r = idn_converter_convtoucs4(aux_converter, from,
buffer_ucs4,
buffer_length);
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
buffer_length *= 2;
}
if (*buffer_ucs4 == '\0') {
if (tolen <= 0) {
r = idn_buffer_overflow;
goto ret;
}
*to = '\0';
r = idn_success;
goto ret;
}
/*
* Convert `buffer_ucs4' to UTF-8.
*/
buffer_length = tolen * 2;
for (;;) {
void *new_buffer;
new_buffer = realloc(buffer_utf8,
sizeof(*buffer_utf8) * buffer_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
buffer_utf8 = (char *)new_buffer;
r = idn_ucs4_ucs4toutf8(buffer_ucs4, buffer_utf8,
buffer_length);
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
buffer_length *= 2;
}
if (*buffer_utf8 == '\0') {
if (tolen <= 0) {
r = idn_buffer_overflow;
goto ret;
}
*to = '\0';
r = idn_success;
goto ret;
}
r = idn_res_decodename(ctx, actions, buffer_utf8, to, tolen);
ret:
if (r == idn_success) {
TRACE(("idn_res_decodename2(): success (to=\"%s\")\n",
idn__debug_xstring(to, 50)));
} else {
TRACE(("idn_res_decodename2(): %s\n", idn_result_tostring(r)));
}
free(buffer_ucs4);
free(buffer_utf8);
if (aux_converter != NULL)
idn_converter_destroy(aux_converter);
return (r);
#endif /* WITHOUT_ICONV */
}
idn_result_t
idn_res_decodename(idn_resconf_t ctx, idn_action_t actions, const char *from,
char *to, size_t tolen) {
idn_converter_t local_converter = NULL;
idn_converter_t idn_converter = NULL;
idn_delimitermap_t delimiter_mapper;
idn_result_t r;
labellist_t labels = NULL, l;
unsigned long *buffer = NULL;
unsigned long *saved_name = NULL;
size_t buffer_length;
int idn_is_ace;
assert(ctx != NULL && from != NULL && to != NULL);
TRACE(("idn_res_decodename(actions=%s, from=\"%s\", tolen=%d)\n",
idn__res_actionstostring(actions),
idn__debug_xstring(from, 50), (int)tolen));
if (actions & ~DECODE_MASK) {
WARNING(("idn_res_decodename: invalid actions 0x%x\n",
actions));
r = idn_invalid_action;
goto ret;
}
if (!initialized)
idn_res_initialize();
if (!enabled || actions == 0) {
r = copy_verbatim(from, to, tolen);
goto ret;
} else if (tolen <= 0) {
r = idn_buffer_overflow;
goto ret;
}
if (actions & IDN_DECODE_QUERY) {
#ifndef WITHOUT_ICONV
actions |= (IDN_DELIMMAP | IDN_MAP | IDN_NORMALIZE | \
IDN_PROHCHECK | IDN_BIDICHECK | IDN_IDNCONV | \
IDN_RTCHECK | IDN_LOCALCONV);
#else
actions |= (IDN_DELIMMAP | IDN_MAP | IDN_NORMALIZE | \
IDN_PROHCHECK | IDN_BIDICHECK | IDN_IDNCONV | \
IDN_RTCHECK);
#endif
}
/*
* Convert `from' to UCS4.
*/
local_converter = idn_resconf_getlocalconverter(ctx);
#ifndef WITHOUT_ICONV
if (local_converter == NULL) {
r = idn_invalid_name;
goto ret;
}
#endif
idn_converter = idn_resconf_getidnconverter(ctx);
if (idn_converter != NULL &&
idn_converter_isasciicompatible(idn_converter))
idn_is_ace = 1;
else
idn_is_ace = 0;
buffer_length = tolen * 2;
TRACE(("res idndecode(name=\"%s\")\n", idn__debug_xstring(from, 50)));
for (;;) {
void *new_buffer;
new_buffer = realloc(buffer, sizeof(*buffer) * buffer_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
buffer = (unsigned long *)new_buffer;
if ((actions & IDN_IDNCONV) &&
idn_converter != NULL && !idn_is_ace) {
r = idn_converter_convtoucs4(idn_converter, from,
buffer, buffer_length);
} else {
r = idn_ucs4_utf8toucs4(from, buffer, buffer_length);
}
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
buffer_length *= 2;
}
if (*buffer == '\0') {
if (tolen <= 0) {
r = idn_buffer_overflow;
goto ret;
}
*to = '\0';
r = idn_success;
goto ret;
}
/*
* Delimiter map.
*/
if (actions & IDN_DELIMMAP) {
TRACE(("res delimitermap(name=\"%s\")\n",
idn__debug_ucs4xstring(buffer, 50)));
delimiter_mapper = idn_resconf_getdelimitermap(ctx);
if (delimiter_mapper != NULL) {
r = idn_delimitermap_map(delimiter_mapper, buffer,
buffer, buffer_length);
idn_delimitermap_destroy(delimiter_mapper);
if (r != idn_success)
goto ret;
}
TRACE(("res delimitermap(): success (name=\"%s\")\n",
idn__debug_ucs4xstring(buffer, 50)));
}
/*
* Split the name into a list of labels.
*/
r = labellist_create(buffer, &labels);
if (r != idn_success)
goto ret;
/*
* Perform conversions and tests.
*/
for (l = labellist_tail(labels); l != NULL;
l = labellist_previous(l)) {
free(saved_name);
saved_name = NULL;
if (!idn__util_ucs4isasciirange(labellist_getname(l))) {
if (actions & IDN_MAP) {
r = label_map(ctx, l);
if (r != idn_success)
goto ret;
}
if (actions & IDN_NORMALIZE) {
r = label_normalize(ctx, l);
if (r != idn_success)
goto ret;
}
if (actions & IDN_PROHCHECK) {
r = label_prohcheck(ctx, l);
if (r == idn_prohibited) {
labellist_undo(l);
continue;
} else if (r != idn_success) {
goto ret;
}
}
if (actions & IDN_UNASCHECK) {
r = label_unascheck(ctx, l);
if (r == idn_prohibited) {
labellist_undo(l);
continue;
} else if (r != idn_success) {
goto ret;
}
}
if (actions & IDN_BIDICHECK) {
r = label_bidicheck(ctx, l);
if (r == idn_prohibited) {
labellist_undo(l);
continue;
} else if (r != idn_success) {
goto ret;
}
}
}
if ((actions & IDN_IDNCONV) && idn_is_ace) {
saved_name = idn_ucs4_strdup(labellist_getname(l));
if (saved_name == NULL) {
r = idn_nomemory;
goto ret;
}
r = label_idndecode(ctx, l);
if (r == idn_invalid_encoding) {
labellist_undo(l);
continue;
} else if (r != idn_success) {
goto ret;
}
}
if ((actions & IDN_RTCHECK) && saved_name != NULL) {
r = label_rtcheck(ctx, actions, l, saved_name);
if (r == idn_invalid_encoding) {
labellist_undo(l);
continue;
} else if (r != idn_success) {
goto ret;
}
}
#ifndef WITHOUT_ICONV
if (actions & IDN_LOCALCONV) {
r = label_localdecodecheck(ctx, l);
if (r != idn_success)
goto ret;
}
#endif
}
/*
* Concat a list of labels to a name.
*/
for (;;) {
void *new_buffer;
new_buffer = realloc(buffer, sizeof(*buffer) * buffer_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
buffer = (unsigned long *)new_buffer;
r = labellist_getnamelist(labels, buffer, buffer_length);
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
buffer_length *= 2;
}
if (actions & IDN_LOCALCONV) {
r = idn_converter_convfromucs4(local_converter, buffer, to,
tolen);
} else {
r = idn_ucs4_ucs4toutf8(buffer, to, tolen);
}
ret:
if (r == idn_success) {
TRACE(("idn_res_decodename(): success (to=\"%s\")\n",
idn__debug_xstring(to, 50)));
} else {
TRACE(("idn_res_decodename(): %s\n", idn_result_tostring(r)));
}
free(saved_name);
free(buffer);
if (local_converter != NULL)
idn_converter_destroy(local_converter);
if (idn_converter != NULL)
idn_converter_destroy(idn_converter);
if (labels != NULL)
labellist_destroy(labels);
return (r);
}
static idn_result_t
label_idndecode(idn_resconf_t ctx, labellist_t label) {
idn_converter_t idn_converter = NULL;
const unsigned long *from;
char *ascii_from = NULL;
unsigned long *to = NULL;
size_t from_length, to_length;
idn_result_t r;
from = labellist_getname(label);
from_length = idn_ucs4_strlen(from) + 1;
TRACE(("res idntoucs4(label=\"%s\")\n",
idn__debug_ucs4xstring(from, 50)));
idn_converter = idn_resconf_getidnconverter(ctx);
if (idn_converter == NULL) {
r = idn_success;
goto ret;
}
for (;;) {
char *new_buffer;
new_buffer = (char *) realloc(ascii_from, from_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
ascii_from = new_buffer;
r = idn_ucs4_ucs4toutf8(from, ascii_from, from_length);
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
from_length *= 2;
}
to = NULL;
to_length = from_length;
for (;;) {
unsigned long *new_buffer;
new_buffer = (unsigned long *)
realloc(to, sizeof(long) * to_length);
if (new_buffer == NULL) {
r = idn_nomemory;
goto ret;
}
to = new_buffer;
r = idn_converter_convtoucs4(idn_converter, ascii_from, to,
to_length);
if (r == idn_success)
break;
else if (r != idn_buffer_overflow)
goto ret;
to_length *= 2;
}
r = labellist_setname(label, to);
ret:
if (r == idn_success) {
TRACE(("res idntoucs4(): success (label=\"%s\")\n",
idn__debug_ucs4xstring(labellist_getname(label),
50)));
} else {
TRACE(("res idntoucs4(): %s\n", idn_result_tostring(r)));
}
if (idn_converter != NULL)
idn_converter_destroy(idn_converter);
free(to);
free(ascii_from);
return (r);
}
idn_result_t
idn__punycode_encode(idn_converter_t ctx, void *privdata,
const unsigned long *from, char *to, size_t tolen) {
char *to_org = to;
unsigned long cur_code, next_code, delta;
size_t prefixlen = strlen(IDN_PUNYCODE_PREFIX);
size_t fromlen;
size_t ucsdone;
size_t toidx;
int uidx, bias, first;
idn_result_t r;
assert(ctx != NULL);
TRACE(("idn__punycode_encode(from=\"%s\", tolen=%d)\n",
idn__debug_ucs4xstring(from, 50), (int)tolen));
if (*from == '\0') {
r = idn_ucs4_ucs4toutf8(from, to, tolen);
goto ret;
} else if (idn__util_ucs4haveaceprefix(from, IDN_PUNYCODE_PREFIX)) {
r = idn_prohibited;
goto ret;
}
if (tolen < prefixlen) {
r = idn_buffer_overflow;
goto ret;
}
memcpy(to, IDN_PUNYCODE_PREFIX, prefixlen);
to += prefixlen;
tolen -= prefixlen;
fromlen = idn_ucs4_strlen(from);
/*
* If the input string is too long (actually too long to be sane),
* return failure in order to prevent possible overflow.
*/
if (fromlen > PUNYCODE_MAXINPUT) {
ERROR(("idn__punycode_encode(): "
"the input string is too long to convert Punycode\n",
idn__debug_ucs4xstring(from, 50)));
r = idn_failure;
goto ret;
}
ucsdone = 0; /* number of characters processed */
toidx = 0;
/*
* First, pick up basic code points and copy them to 'to'.
*/
for (uidx = 0; uidx < fromlen; uidx++) {
if (from[uidx] < 0x80) {
if (toidx >= tolen) {
r = idn_buffer_overflow;
goto ret;
}
to[toidx++] = from[uidx];
ucsdone++;
}
}
/*
* If there are any basic code points, output a delimiter
* (hyphen-minus).
*/
if (toidx > 0) {
if (toidx >= tolen) {
r = idn_buffer_overflow;
goto ret;
}
to[toidx++] = '-';
to += toidx;
tolen -= toidx;
}
/*
* Then encode non-basic characters.
*/
first = 1;
cur_code = PUNYCODE_INITIAL_N;
bias = PUNYCODE_INITIAL_BIAS;
delta = 0;
while (ucsdone < fromlen) {
int limit = -1, rest;
/*
* Find the smallest code point equal to or greater
* than 'cur_code'. Also remember the index of the
* last occurence of the code point.
*/
for (next_code = MAX_UCS, uidx = fromlen - 1;
uidx >= 0; uidx--) {
if (from[uidx] >= cur_code && from[uidx] < next_code) {
next_code = from[uidx];
limit = uidx;
}
}
/* There must be such code point. */
assert(limit >= 0);
delta += (next_code - cur_code) * (ucsdone + 1);
cur_code = next_code;
/*
* Scan the input string again, and encode characters
* whose code point is 'cur_code'. Use 'limit' to avoid
* unnecessary scan.
*/
for (uidx = 0, rest = ucsdone; uidx <= limit; uidx++) {
if (from[uidx] < cur_code) {
delta++;
rest--;
} else if (from[uidx] == cur_code) {
int sz = punycode_putwc(to, tolen, delta, bias);
if (sz == 0) {
r = idn_buffer_overflow;
goto ret;
}
to += sz;
tolen -= sz;
ucsdone++;
bias = punycode_update_bias(delta, ucsdone,
first);
delta = 0;
first = 0;
}
}
delta += rest + 1;
cur_code++;
}
/*
* Terminate with NUL.
*/
if (tolen <= 0) {
r = idn_buffer_overflow;
goto ret;
}
*to = '\0';
r = idn_success;
ret:
if (r == idn_success) {
TRACE(("idn__punycode_encode(): succcess (to=\"%s\")\n",
idn__debug_xstring(to_org, 50)));
} else {
TRACE(("idn__punycode_encode(): %s\n", idn_result_tostring(r)));
}
return (r);
}
