U_CAPI int32_t U_EXPORT2
ucol_getRulesEx(const UCollator *coll, UColRuleOption delta, UChar *buffer, int32_t bufferLen) {
UErrorCode status = U_ZERO_ERROR;
int32_t len = 0;
int32_t UCAlen = 0;
const UChar* ucaRules = 0;
const UChar *rules = ucol_getRules(coll, &len);
if(delta == UCOL_FULL_RULES) {
/* take the UCA rules and append real rules at the end */
/* UCA rules will be probably coming from the root RB */
ucaRules = ures_getStringByKey(coll->rb,"UCARules",&UCAlen,&status);
/*
UResourceBundle* cresb = ures_getByKeyWithFallback(coll->rb, "collations", NULL, &status);
UResourceBundle* uca = ures_getByKeyWithFallback(cresb, "UCA", NULL, &status);
ucaRules = ures_getStringByKey(uca,"Sequence",&UCAlen,&status);
ures_close(uca);
ures_close(cresb);
*/
}
if(U_FAILURE(status)) {
return 0;
}
if(buffer!=0 && bufferLen>0){
*buffer=0;
if(UCAlen > 0) {
u_memcpy(buffer, ucaRules, uprv_min(UCAlen, bufferLen));
}
if(len > 0 && bufferLen > UCAlen) {
u_memcpy(buffer+UCAlen, rules, uprv_min(len, bufferLen-UCAlen));
}
}
return u_terminateUChars(buffer, bufferLen, len+UCAlen, &status);
}
/**
* Applies a new value to the text that serves as input at the current
* processing step. This value is identical to the original one when we begin
* the processing, but usually changes as the transformation progresses.
*
* @param pTransform A pointer to the <code>UBiDiTransform</code> structure.
* @param newSrc A pointer whose value is to be used as input text.
* @param newLength A length of the new text in <code>UChar</code>s.
* @param newSize A new source capacity in <code>UChar</code>s.
* @param pErrorCode Pointer to the error code value.
*/
static void
updateSrc(UBiDiTransform *pTransform, const UChar *newSrc, uint32_t newLength,
uint32_t newSize, UErrorCode *pErrorCode)
{
if (newSize < newLength) {
*pErrorCode = U_BUFFER_OVERFLOW_ERROR;
return;
}
if (newSize > pTransform->srcSize) {
newSize += 50; // allocate slightly more than needed right now
if (pTransform->src != NULL) {
uprv_free(pTransform->src);
pTransform->src = NULL;
}
pTransform->src = (UChar *)uprv_malloc(newSize * sizeof(UChar));
if (pTransform->src == NULL) {
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
//pTransform->srcLength = pTransform->srcSize = 0;
return;
}
pTransform->srcSize = newSize;
}
u_strncpy(pTransform->src, newSrc, newLength);
pTransform->srcLength = u_terminateUChars(pTransform->src,
pTransform->srcSize, newLength, pErrorCode);
}
static int32_t
_iterate(UCharIterator *src, UBool forward,
UChar *dest, int32_t destCapacity,
const Normalizer2 *n2,
UBool doNormalize, UBool *pNeededToNormalize,
UErrorCode *pErrorCode) {
if(U_FAILURE(*pErrorCode)) {
return 0;
}
if(destCapacity<0 || (dest==NULL && destCapacity>0) || src==NULL) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
if(pNeededToNormalize!=NULL) {
*pNeededToNormalize=FALSE;
}
if(!(forward ? src->hasNext(src) : src->hasPrevious(src))) {
return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
}
UnicodeString buffer;
UChar32 c;
if(forward) {
/* get one character and ignore its properties */
buffer.append(uiter_next32(src));
/* get all following characters until we see a boundary */
while((c=uiter_next32(src))>=0) {
if(n2->hasBoundaryBefore(c)) {
/* back out the latest movement to stop at the boundary */
src->move(src, -U16_LENGTH(c), UITER_CURRENT);
break;
} else {
buffer.append(c);
}
}
} else {
while((c=uiter_previous32(src))>=0) {
/* always write this character to the front of the buffer */
buffer.insert(0, c);
/* stop if this just-copied character is a boundary */
if(n2->hasBoundaryBefore(c)) {
break;
}
}
}
UnicodeString destString(dest, 0, destCapacity);
if(buffer.length()>0 && doNormalize) {
n2->normalize(buffer, destString, *pErrorCode).extract(dest, destCapacity, *pErrorCode);
if(pNeededToNormalize!=NULL && U_SUCCESS(*pErrorCode)) {
*pNeededToNormalize= destString!=buffer;
}
return destString.length();
} else {
/* just copy the source characters */
return buffer.extract(dest, destCapacity, *pErrorCode);
}
}
static int32_t
_getStringOrCopyKey(const char *path, const char *locale,
const char *tableKey,
const char* subTableKey,
const char *itemKey,
const char *substitute,
UChar *dest, int32_t destCapacity,
UErrorCode *pErrorCode) {
const UChar *s = NULL;
int32_t length = 0;
if(itemKey==NULL) {
/* top-level item: normal resource bundle access */
UResourceBundle *rb;
rb=ures_open(path, locale, pErrorCode);
if(U_SUCCESS(*pErrorCode)) {
s=ures_getStringByKey(rb, tableKey, &length, pErrorCode);
/* see comment about closing rb near "return item;" in _res_getTableStringWithFallback() */
ures_close(rb);
}
} else {
/* Language code should not be a number. If it is, set the error code. */
if (!uprv_strncmp(tableKey, "Languages", 9) && uprv_strtol(itemKey, NULL, 10)) {
*pErrorCode = U_MISSING_RESOURCE_ERROR;
} else {
/* second-level item, use special fallback */
s=uloc_getTableStringWithFallback(path, locale,
tableKey,
subTableKey,
itemKey,
&length,
pErrorCode);
}
}
if(U_SUCCESS(*pErrorCode)) {
int32_t copyLength=uprv_min(length, destCapacity);
if(copyLength>0 && s != NULL) {
u_memcpy(dest, s, copyLength);
}
} else {
/* no string from a resource bundle: convert the substitute */
length=(int32_t)uprv_strlen(substitute);
u_charsToUChars(substitute, dest, uprv_min(length, destCapacity));
*pErrorCode=U_USING_DEFAULT_WARNING;
}
return u_terminateUChars(dest, destCapacity, length, pErrorCode);
}
U_CAPI UChar* U_EXPORT2
u_strFromWCS(UChar *dest,
int32_t destCapacity,
int32_t *pDestLength,
const wchar_t *src,
int32_t srcLength,
UErrorCode *pErrorCode)
{
/* args check */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)){
return NULL;
}
if( (src==NULL && srcLength!=0) || srcLength < -1 ||
(destCapacity<0) || (dest == NULL && destCapacity > 0)
) {
*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
return NULL;
}
#ifdef U_WCHAR_IS_UTF16
/* wchar_t is UTF-16 just do a memcpy */
if(srcLength == -1){
srcLength = u_strlen(src);
}
if(0 < srcLength && srcLength <= destCapacity){
uprv_memcpy(dest,src,srcLength*U_SIZEOF_UCHAR);
}
if(pDestLength){
*pDestLength = srcLength;
}
u_terminateUChars(dest,destCapacity,srcLength,pErrorCode);
return dest;
#elif defined U_WCHAR_IS_UTF32
return u_strFromUTF32(dest, destCapacity, pDestLength,
(UChar32*)src, srcLength, pErrorCode);
#else
return _strFromWCS(dest,destCapacity,pDestLength,src,srcLength,pErrorCode);
#endif
}
static int32_t
_getDisplayNameForComponent(const char * locale,
const char * displayLocale,
UChar * dest, int32_t destCapacity,
UDisplayNameGetter * getter,
const char * tag,
UErrorCode * pErrorCode)
{
char localeBuffer[ULOC_FULLNAME_CAPACITY * 4];
int32_t length;
UErrorCode localStatus;
const char * root = NULL;
/* argument checking */
if (pErrorCode == NULL || U_FAILURE(*pErrorCode))
{
return 0;
}
if (destCapacity < 0 || (destCapacity > 0 && dest == NULL))
{
*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
localStatus = U_ZERO_ERROR;
length = (*getter)(locale, localeBuffer, sizeof(localeBuffer), &localStatus);
if (U_FAILURE(localStatus) || localStatus == U_STRING_NOT_TERMINATED_WARNING)
{
*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
if (length == 0)
{
return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
}
root = tag == _kCountries ? U_ICUDATA_REGION : U_ICUDATA_LANG;
return _getStringOrCopyKey(root, displayLocale,
tag, NULL, localeBuffer,
localeBuffer,
dest, destCapacity,
pErrorCode);
}
U_CAPI int32_t U_EXPORT2
ucnv_getDisplayName(const UConverter *cnv,
const char *displayLocale,
UChar *displayName, int32_t displayNameCapacity,
UErrorCode *pErrorCode) {
UResourceBundle *rb;
const UChar *name;
int32_t length;
UErrorCode localStatus = U_ZERO_ERROR;
/* check arguments */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
if(cnv==NULL || displayNameCapacity<0 || (displayNameCapacity>0 && displayName==NULL)) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/* open the resource bundle and get the display name string */
rb=ures_open(NULL, displayLocale, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
return 0;
}
/* use the internal name as the key */
name=ures_getStringByKey(rb, cnv->sharedData->staticData->name, &length, &localStatus);
ures_close(rb);
if(U_SUCCESS(localStatus)) {
/* copy the string */
if (*pErrorCode == U_ZERO_ERROR) {
*pErrorCode = localStatus;
}
u_memcpy(displayName, name, uprv_min(length, displayNameCapacity)*U_SIZEOF_UCHAR);
} else {
/* convert the internal name into a Unicode string */
length=(int32_t)uprv_strlen(cnv->sharedData->staticData->name);
u_charsToUChars(cnv->sharedData->staticData->name, displayName, uprv_min(length, displayNameCapacity));
}
return u_terminateUChars(displayName, displayNameCapacity, length, pErrorCode);
}
U_CAPI int32_t U_EXPORT2
uscript_getSampleString(UScriptCode script, UChar *dest, int32_t capacity, UErrorCode *pErrorCode) {
if(U_FAILURE(*pErrorCode)) { return 0; }
if(capacity < 0 || (capacity > 0 && dest == NULL)) {
*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
int32_t sampleChar = getScriptProps(script) & 0x1fffff;
int32_t length;
if(sampleChar == 0) {
length = 0;
} else {
length = U16_LENGTH(sampleChar);
if(length <= capacity) {
int32_t i = 0;
U16_APPEND_UNSAFE(dest, i, sampleChar);
}
}
return u_terminateUChars(dest, capacity, length, pErrorCode);
}
int32_t NamePrepTransform::map(const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
UBool allowUnassigned,
UParseError* /*parseError*/,
UErrorCode& status ){
if(U_FAILURE(status)){
return 0;
}
//check arguments
if(src==NULL || srcLength<-1 || (dest==NULL && destCapacity!=0)) {
status=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
UnicodeString rsource(src,srcLength);
// map the code points
// transliteration also performs NFKC
mapping->transliterate(rsource);
const UChar* buffer = rsource.getBuffer();
int32_t bufLen = rsource.length();
// check if unassigned
if(allowUnassigned == FALSE){
int32_t bufIndex=0;
UChar32 ch =0 ;
for(;bufIndex<bufLen;){
U16_NEXT(buffer, bufIndex, bufLen, ch);
if(unassigned.contains(ch)){
status = U_IDNA_UNASSIGNED_ERROR;
return 0;
}
}
}
// check if there is enough room in the output
if(bufLen < destCapacity){
uprv_memcpy(dest,buffer,bufLen*U_SIZEOF_UCHAR);
}
return u_terminateUChars(dest, destCapacity, bufLen, &status);
}
U_CAPI int32_t U_EXPORT2
ubidi_writeReverse(const UChar *src, int32_t srcLength,
UChar *dest, int32_t destSize,
uint16_t options,
UErrorCode *pErrorCode) {
int32_t destLength;
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
/* more error checking */
if( src==NULL || srcLength<-1 ||
destSize<0 || (destSize>0 && dest==NULL))
{
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/* do input and output overlap? */
if( dest!=NULL &&
((src>=dest && src<dest+destSize) ||
(dest>=src && dest<src+srcLength)))
{
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
if(srcLength==-1) {
srcLength=u_strlen(src);
}
if(srcLength>0) {
destLength=doWriteReverse(src, srcLength, dest, destSize, options, pErrorCode);
} else {
/* nothing to do */
destLength=0;
}
return u_terminateUChars(dest, destSize, destLength, pErrorCode);
}
U_CAPI int32_t U_EXPORT2
uloc_getDisplayKeywordValue( const char* locale,
const char* keyword,
const char* displayLocale,
UChar* dest,
int32_t destCapacity,
UErrorCode* status){
char keywordValue[ULOC_FULLNAME_CAPACITY*4];
int32_t capacity = ULOC_FULLNAME_CAPACITY*4;
int32_t keywordValueLen =0;
/* argument checking */
if(status==NULL || U_FAILURE(*status)) {
return 0;
}
if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
*status=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/* get the keyword value */
keywordValue[0]=0;
keywordValueLen = uloc_getKeywordValue(locale, keyword, keywordValue, capacity, status);
/*
* if the keyword is equal to currency .. then to get the display name
* we need to do the fallback ourselves
*/
if(uprv_stricmp(keyword, _kCurrency)==0){
int32_t dispNameLen = 0;
const UChar *dispName = NULL;
UResourceBundle *bundle = ures_open(U_ICUDATA_CURR, displayLocale, status);
UResourceBundle *currencies = ures_getByKey(bundle, _kCurrencies, NULL, status);
UResourceBundle *currency = ures_getByKeyWithFallback(currencies, keywordValue, NULL, status);
dispName = ures_getStringByIndex(currency, UCURRENCY_DISPLAY_NAME_INDEX, &dispNameLen, status);
/*close the bundles */
ures_close(currency);
ures_close(currencies);
ures_close(bundle);
if(U_FAILURE(*status)){
if(*status == U_MISSING_RESOURCE_ERROR){
/* we just want to write the value over if nothing is available */
*status = U_USING_DEFAULT_WARNING;
}else{
return 0;
}
}
/* now copy the dispName over if not NULL */
if(dispName != NULL){
if(dispNameLen <= destCapacity){
uprv_memcpy(dest, dispName, dispNameLen * U_SIZEOF_UCHAR);
return u_terminateUChars(dest, destCapacity, dispNameLen, status);
}else{
*status = U_BUFFER_OVERFLOW_ERROR;
return dispNameLen;
}
}else{
/* we have not found the display name for the value .. just copy over */
if(keywordValueLen <= destCapacity){
u_charsToUChars(keywordValue, dest, keywordValueLen);
return u_terminateUChars(dest, destCapacity, keywordValueLen, status);
}else{
*status = U_BUFFER_OVERFLOW_ERROR;
return keywordValueLen;
}
}
}else{
return _getStringOrCopyKey(U_ICUDATA_LANG, displayLocale,
_kTypes, keyword,
keywordValue,
keywordValue,
dest, destCapacity,
status);
}
}
/* Instead of having a separate pass for 'special' patterns, reintegrate the two
* so we don't get bitten by preflight bugs again. We can be reasonably efficient
* without two separate code paths, this code isn't that performance-critical.
*
* This code is general enough to deal with patterns that have a prefix or swap the
* language and remainder components, since we gave developers enough rope to do such
* things if they futz with the pattern data. But since we don't give them a way to
* specify a pattern for arbitrary combinations of components, there's not much use in
* that. I don't think our data includes such patterns, the only variable I know if is
* whether there is a space before the open paren, or not. Oh, and zh uses different
* chars than the standard open/close paren (which ja and ko use, btw).
*/
U_CAPI int32_t U_EXPORT2
uloc_getDisplayName(const char *locale,
const char *displayLocale,
UChar *dest, int32_t destCapacity,
UErrorCode *pErrorCode)
{
static const UChar defaultSeparator[9] = { 0x007b, 0x0030, 0x007d, 0x002c, 0x0020, 0x007b, 0x0031, 0x007d, 0x0000 }; /* "{0}, {1}" */
static const UChar sub0[4] = { 0x007b, 0x0030, 0x007d , 0x0000 } ; /* {0} */
static const UChar sub1[4] = { 0x007b, 0x0031, 0x007d , 0x0000 } ; /* {1} */
static const int32_t subLen = 3;
static const UChar defaultPattern[10] = {
0x007b, 0x0030, 0x007d, 0x0020, 0x0028, 0x007b, 0x0031, 0x007d, 0x0029, 0x0000
}; /* {0} ({1}) */
static const int32_t defaultPatLen = 9;
static const int32_t defaultSub0Pos = 0;
static const int32_t defaultSub1Pos = 5;
int32_t length; /* of formatted result */
const UChar *separator;
int32_t sepLen = 0;
const UChar *pattern;
int32_t patLen = 0;
int32_t sub0Pos, sub1Pos;
UChar formatOpenParen = 0x0028; // (
UChar formatReplaceOpenParen = 0x005B; // [
UChar formatCloseParen = 0x0029; // )
UChar formatReplaceCloseParen = 0x005D; // ]
UBool haveLang = TRUE; /* assume true, set false if we find we don't have
a lang component in the locale */
UBool haveRest = TRUE; /* assume true, set false if we find we don't have
any other component in the locale */
UBool retry = FALSE; /* set true if we need to retry, see below */
int32_t langi = 0; /* index of the language substitution (0 or 1), virtually always 0 */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
{
UErrorCode status = U_ZERO_ERROR;
UResourceBundle* locbundle=ures_open(U_ICUDATA_LANG, displayLocale, &status);
UResourceBundle* dspbundle=ures_getByKeyWithFallback(locbundle, _kLocaleDisplayPattern,
NULL, &status);
separator=ures_getStringByKeyWithFallback(dspbundle, _kSeparator, &sepLen, &status);
pattern=ures_getStringByKeyWithFallback(dspbundle, _kPattern, &patLen, &status);
ures_close(dspbundle);
ures_close(locbundle);
}
/* If we couldn't find any data, then use the defaults */
if(sepLen == 0) {
separator = defaultSeparator;
}
/* #10244: Even though separator is now a pattern, it is awkward to handle it as such
* here since we are trying to build the display string in place in the dest buffer,
* and to handle it as a pattern would entail having separate storage for the
* substrings that need to be combined (the first of which may be the result of
* previous such combinations). So for now we continue to treat the portion between
* {0} and {1} as a string to be appended when joining substrings, ignoring anything
* that is before {0} or after {1} (no existing separator pattern has any such thing).
* This is similar to how pattern is handled below.
*/
{
UChar *p0=u_strstr(separator, sub0);
UChar *p1=u_strstr(separator, sub1);
if (p0==NULL || p1==NULL || p1<p0) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
separator = (const UChar *)p0 + subLen;
sepLen = p1 - separator;
}
if(patLen==0 || (patLen==defaultPatLen && !u_strncmp(pattern, defaultPattern, patLen))) {
pattern=defaultPattern;
patLen=defaultPatLen;
sub0Pos=defaultSub0Pos;
sub1Pos=defaultSub1Pos;
// use default formatOpenParen etc. set above
} else { /* non-default pattern */
UChar *p0=u_strstr(pattern, sub0);
UChar *p1=u_strstr(pattern, sub1);
if (p0==NULL || p1==NULL) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
sub0Pos=p0-pattern;
sub1Pos=p1-pattern;
if (sub1Pos < sub0Pos) { /* a very odd pattern */
int32_t t=sub0Pos; sub0Pos=sub1Pos; sub1Pos=t;
langi=1;
}
if (u_strchr(pattern, 0xFF08) != NULL) {
formatOpenParen = 0xFF08; // fullwidth (
formatReplaceOpenParen = 0xFF3B; // fullwidth [
formatCloseParen = 0xFF09; // fullwidth )
formatReplaceCloseParen = 0xFF3D; // fullwidth ]
}
}
/* We loop here because there is one case in which after the first pass we could need to
* reextract the data. If there's initial padding before the first element, we put in
* the padding and then write that element. If it turns out there's no second element,
* we didn't need the padding. If we do need the data (no preflight), and the first element
* would have fit but for the padding, we need to reextract. In this case (only) we
* adjust the parameters so padding is not added, and repeat.
*/
do {
UChar* p=dest;
int32_t patPos=0; /* position in the pattern, used for non-substitution portions */
int32_t langLen=0; /* length of language substitution */
int32_t langPos=0; /* position in output of language substitution */
int32_t restLen=0; /* length of 'everything else' substitution */
int32_t restPos=0; /* position in output of 'everything else' substitution */
UEnumeration* kenum = NULL; /* keyword enumeration */
/* prefix of pattern, extremely likely to be empty */
if(sub0Pos) {
if(destCapacity >= sub0Pos) {
while (patPos < sub0Pos) {
*p++ = pattern[patPos++];
}
} else {
patPos=sub0Pos;
}
length=sub0Pos;
} else {
length=0;
}
for(int32_t subi=0,resti=0;subi<2;) { /* iterate through patterns 0 and 1*/
UBool subdone = FALSE; /* set true when ready to move to next substitution */
/* prep p and cap for calls to get display components, pin cap to 0 since
they complain if cap is negative */
int32_t cap=destCapacity-length;
if (cap <= 0) {
cap=0;
} else {
p=dest+length;
}
if (subi == langi) { /* {0}*/
if(haveLang) {
langPos=length;
langLen=uloc_getDisplayLanguage(locale, displayLocale, p, cap, pErrorCode);
length+=langLen;
haveLang=langLen>0;
}
subdone=TRUE;
} else { /* {1} */
if(!haveRest) {
subdone=TRUE;
} else {
int32_t len; /* length of component (plus other stuff) we just fetched */
switch(resti++) {
case 0:
restPos=length;
len=uloc_getDisplayScriptInContext(locale, displayLocale, p, cap, pErrorCode);
break;
case 1:
len=uloc_getDisplayCountry(locale, displayLocale, p, cap, pErrorCode);
break;
case 2:
len=uloc_getDisplayVariant(locale, displayLocale, p, cap, pErrorCode);
break;
case 3:
kenum = uloc_openKeywords(locale, pErrorCode);
/* fall through */
default: {
const char* kw=uenum_next(kenum, &len, pErrorCode);
if (kw == NULL) {
uenum_close(kenum);
len=0; /* mark that we didn't add a component */
subdone=TRUE;
} else {
/* incorporating this behavior into the loop made it even more complex,
so just special case it here */
len = uloc_getDisplayKeyword(kw, displayLocale, p, cap, pErrorCode);
if(len) {
if(len < cap) {
p[len]=0x3d; /* '=', assume we'll need it */
}
len+=1;
/* adjust for call to get keyword */
cap-=len;
if(cap <= 0) {
cap=0;
} else {
p+=len;
}
}
/* reset for call below */
if(*pErrorCode == U_BUFFER_OVERFLOW_ERROR) {
*pErrorCode=U_ZERO_ERROR;
}
int32_t vlen = uloc_getDisplayKeywordValue(locale, kw, displayLocale,
p, cap, pErrorCode);
if(len) {
if(vlen==0) {
--len; /* remove unneeded '=' */
}
/* restore cap and p to what they were at start */
cap=destCapacity-length;
if(cap <= 0) {
cap=0;
} else {
p=dest+length;
}
}
len+=vlen; /* total we added for key + '=' + value */
}
} break;
} /* end switch */
if (len>0) {
/* we addeed a component, so add separator and write it if there's room. */
if(len+sepLen<=cap) {
const UChar * plimit = p + len;
for (; p < plimit; p++) {
if (*p == formatOpenParen) {
*p = formatReplaceOpenParen;
} else if (*p == formatCloseParen) {
*p = formatReplaceCloseParen;
}
}
for(int32_t i=0;i<sepLen;++i) {
*p++=separator[i];
}
}
length+=len+sepLen;
} else if(subdone) {
/* remove separator if we added it */
if (length!=restPos) {
length-=sepLen;
}
restLen=length-restPos;
haveRest=restLen>0;
}
}
}
if(*pErrorCode == U_BUFFER_OVERFLOW_ERROR) {
*pErrorCode=U_ZERO_ERROR;
}
if(subdone) {
if(haveLang && haveRest) {
/* append internal portion of pattern, the first time,
or last portion of pattern the second time */
int32_t padLen;
patPos+=subLen;
padLen=(subi==0 ? sub1Pos : patLen)-patPos;
if(length+padLen < destCapacity) {
p=dest+length;
for(int32_t i=0;i<padLen;++i) {
*p++=pattern[patPos++];
}
} else {
patPos+=padLen;
}
length+=padLen;
} else if(subi==0) {
/* don't have first component, reset for second component */
sub0Pos=0;
length=0;
} else if(length>0) {
/* true length is the length of just the component we got. */
length=haveLang?langLen:restLen;
if(dest && sub0Pos!=0) {
if (sub0Pos+length<=destCapacity) {
/* first component not at start of result,
but we have full component in buffer. */
u_memmove(dest, dest+(haveLang?langPos:restPos), length);
} else {
/* would have fit, but didn't because of pattern prefix. */
sub0Pos=0; /* stops initial padding (and a second retry,
so we won't end up here again) */
retry=TRUE;
}
}
}
++subi; /* move on to next substitution */
}
}
} while(retry);
return u_terminateUChars(dest, destCapacity, length, pErrorCode);
}
U_CFUNC int32_t
u_strFromPunycode(const UChar *src, int32_t srcLength,
UChar *dest, int32_t destCapacity,
UBool *caseFlags,
UErrorCode *pErrorCode) {
int32_t n, destLength, i, bias, basicLength, j, in, oldi, w, k, digit, t,
destCPCount, firstSupplementaryIndex, cpLength;
UChar b;
/* argument checking */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
if(src==NULL || srcLength<-1 || (dest==NULL && destCapacity!=0)) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
if(srcLength==-1) {
srcLength=u_strlen(src);
}
/*
* Handle the basic code points:
* Let basicLength be the number of input code points
* before the last delimiter, or 0 if there is none,
* then copy the first basicLength code points to the output.
*
* The two following loops iterate backward.
*/
for(j=srcLength; j>0;) {
if(src[--j]==DELIMITER) {
break;
}
}
destLength=basicLength=destCPCount=j;
U_ASSERT(destLength>=0);
while(j>0) {
b=src[--j];
if(!IS_BASIC(b)) {
*pErrorCode=U_INVALID_CHAR_FOUND;
return 0;
}
if(j<destCapacity) {
dest[j]=(UChar)b;
if(caseFlags!=NULL) {
caseFlags[j]=IS_BASIC_UPPERCASE(b);
}
}
}
/* Initialize the state: */
n=INITIAL_N;
i=0;
bias=INITIAL_BIAS;
firstSupplementaryIndex=1000000000;
/*
* Main decoding loop:
* Start just after the last delimiter if any
* basic code points were copied; start at the beginning otherwise.
*/
for(in=basicLength>0 ? basicLength+1 : 0; in<srcLength; /* no op */) {
/*
* in is the index of the next character to be consumed, and
* destCPCount is the number of code points in the output array.
*
* Decode a generalized variable-length integer into delta,
* which gets added to i. The overflow checking is easier
* if we increase i as we go, then subtract off its starting
* value at the end to obtain delta.
*/
for(oldi=i, w=1, k=BASE; /* no condition */; k+=BASE) {
if(in>=srcLength) {
*pErrorCode=U_ILLEGAL_CHAR_FOUND;
return 0;
}
digit=basicToDigit[(uint8_t)src[in++]];
if(digit<0) {
*pErrorCode=U_INVALID_CHAR_FOUND;
return 0;
}
if(digit>(0x7fffffff-i)/w) {
/* integer overflow */
*pErrorCode=U_ILLEGAL_CHAR_FOUND;
return 0;
}
i+=digit*w;
/** RAM: comment out the old code for conformance with draft-ietf-idn-punycode-03.txt
t=k-bias;
if(t<TMIN) {
t=TMIN;
} else if(t>TMAX) {
t=TMAX;
}
*/
t=k-bias;
if(t<TMIN) {
t=TMIN;
} else if(k>=(bias+TMAX)) {
t=TMAX;
}
if(digit<t) {
break;
}
if(w>0x7fffffff/(BASE-t)) {
/* integer overflow */
*pErrorCode=U_ILLEGAL_CHAR_FOUND;
return 0;
}
w*=BASE-t;
}
/*
* Modification from sample code:
* Increments destCPCount here,
* where needed instead of in for() loop tail.
*/
++destCPCount;
bias=adaptBias(i-oldi, destCPCount, (UBool)(oldi==0));
/*
* i was supposed to wrap around from (incremented) destCPCount to 0,
* incrementing n each time, so we'll fix that now:
*/
if(i/destCPCount>(0x7fffffff-n)) {
/* integer overflow */
*pErrorCode=U_ILLEGAL_CHAR_FOUND;
return 0;
}
n+=i/destCPCount;
i%=destCPCount;
/* not needed for Punycode: */
/* if (decode_digit(n) <= BASE) return punycode_invalid_input; */
if(n>0x10ffff || U_IS_SURROGATE(n)) {
/* Unicode code point overflow */
*pErrorCode=U_ILLEGAL_CHAR_FOUND;
return 0;
}
/* Insert n at position i of the output: */
cpLength=U16_LENGTH(n);
if(dest!=NULL && ((destLength+cpLength)<=destCapacity)) {
int32_t codeUnitIndex;
/*
* Handle indexes when supplementary code points are present.
*
* In almost all cases, there will be only BMP code points before i
* and even in the entire string.
* This is handled with the same efficiency as with UTF-32.
*
* Only the rare cases with supplementary code points are handled
* more slowly - but not too bad since this is an insertion anyway.
*/
if(i<=firstSupplementaryIndex) {
codeUnitIndex=i;
if(cpLength>1) {
firstSupplementaryIndex=codeUnitIndex;
} else {
++firstSupplementaryIndex;
}
} else {
codeUnitIndex=firstSupplementaryIndex;
U16_FWD_N(dest, codeUnitIndex, destLength, i-codeUnitIndex);
}
/* use the UChar index codeUnitIndex instead of the code point index i */
if(codeUnitIndex<destLength) {
uprv_memmove(dest+codeUnitIndex+cpLength,
dest+codeUnitIndex,
(destLength-codeUnitIndex)*U_SIZEOF_UCHAR);
if(caseFlags!=NULL) {
uprv_memmove(caseFlags+codeUnitIndex+cpLength,
caseFlags+codeUnitIndex,
destLength-codeUnitIndex);
}
}
if(cpLength==1) {
/* BMP, insert one code unit */
dest[codeUnitIndex]=(UChar)n;
} else {
/* supplementary character, insert two code units */
dest[codeUnitIndex]=U16_LEAD(n);
dest[codeUnitIndex+1]=U16_TRAIL(n);
}
if(caseFlags!=NULL) {
/* Case of last character determines uppercase flag: */
caseFlags[codeUnitIndex]=IS_BASIC_UPPERCASE(src[in-1]);
if(cpLength==2) {
caseFlags[codeUnitIndex+1]=FALSE;
}
}
}
destLength+=cpLength;
U_ASSERT(destLength>=0);
++i;
}
return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
}
U_CFUNC int32_t
idnaref_toUnicode(const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UParseError* parseError,
UErrorCode* status){
if(status == NULL || U_FAILURE(*status)){
return 0;
}
if((src == NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
*status = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE], b3Stack[MAX_LABEL_BUFFER_SIZE];
//initialize pointers to stack buffers
UChar *b1 = b1Stack, *b2 = b2Stack, *b1Prime=NULL, *b3=b3Stack;
int32_t b1Len, b2Len, b1PrimeLen, b3Len,
b1Capacity = MAX_LABEL_BUFFER_SIZE,
b2Capacity = MAX_LABEL_BUFFER_SIZE,
b3Capacity = MAX_LABEL_BUFFER_SIZE,
reqLength=0;
// UParseError parseError;
NamePrepTransform* prep = TestIDNA::getInstance(*status);
b1Len = 0;
UBool* caseFlags = NULL;
//get the options
UBool allowUnassigned = (UBool)((options & IDNAREF_ALLOW_UNASSIGNED) != 0);
UBool useSTD3ASCIIRules = (UBool)((options & IDNAREF_USE_STD3_RULES) != 0);
UBool srcIsASCII = TRUE;
UBool srcIsLDH = TRUE;
int32_t failPos =0;
if(U_FAILURE(*status)){
goto CLEANUP;
}
// step 1: find out if all the codepoints in src are ASCII
if(srcLength==-1){
srcLength = 0;
for(;src[srcLength]!=0;){
if(src[srcLength]> 0x7f){
srcIsASCII = FALSE;
}if(prep->isLDHChar(src[srcLength])==FALSE){
// here we do not assemble surrogates
// since we know that LDH code points
// are in the ASCII range only
srcIsLDH = FALSE;
failPos = srcLength;
}
srcLength++;
}
}else{
for(int32_t j=0; j<srcLength; j++){
if(src[j]> 0x7f){
srcIsASCII = FALSE;
}else if(prep->isLDHChar(src[j])==FALSE){
// here we do not assemble surrogates
// since we know that LDH code points
// are in the ASCII range only
srcIsLDH = FALSE;
failPos = j;
}
}
}
if(srcIsASCII == FALSE){
// step 2: process the string
b1Len = prep->process(src,srcLength,b1,b1Capacity,allowUnassigned, parseError, *status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b1Len = prep->process(src,srcLength,b1, b1Len,allowUnassigned, parseError, *status);
}
//bail out on error
if(U_FAILURE(*status)){
goto CLEANUP;
}
}else{
// copy everything to b1
if(srcLength < b1Capacity){
uprv_memmove(b1,src, srcLength * U_SIZEOF_UCHAR);
}else{
/* we do not have enough room so grow the buffer*/
b1 = (UChar*) uprv_malloc(srcLength * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
uprv_memmove(b1,src, srcLength * U_SIZEOF_UCHAR);
}
b1Len = srcLength;
}
//step 3: verify ACE Prefix
if(startsWithPrefix(src,srcLength)){
//step 4: Remove the ACE Prefix
b1Prime = b1 + ACE_PREFIX_LENGTH;
b1PrimeLen = b1Len - ACE_PREFIX_LENGTH;
//step 5: Decode using punycode
b2Len = convertFromPuny(b1Prime,b1PrimeLen, b2, b2Capacity, *status);
//b2Len = u_strFromPunycode(b2, b2Capacity,b1Prime,b1PrimeLen, caseFlags, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
if(b2==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b2Len = convertFromPuny(b1Prime,b1PrimeLen, b2, b2Len, *status);
//b2Len = u_strFromPunycode(b2, b2Len,b1Prime,b1PrimeLen,caseFlags, status);
}
//step 6:Apply toASCII
b3Len = idnaref_toASCII(b2,b2Len,b3,b3Capacity,options,parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b3 = (UChar*) uprv_malloc(b3Len * U_SIZEOF_UCHAR);
if(b3==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b3Len = idnaref_toASCII(b2,b2Len,b3,b3Len, options, parseError, status);
}
//bail out on error
if(U_FAILURE(*status)){
goto CLEANUP;
}
//step 7: verify
if(compareCaseInsensitiveASCII(b1, b1Len, b3, b3Len) !=0){
*status = U_IDNA_VERIFICATION_ERROR;
goto CLEANUP;
}
//step 8: return output of step 5
reqLength = b2Len;
if(b2Len <= destCapacity) {
uprv_memmove(dest, b2, b2Len * U_SIZEOF_UCHAR);
}
}else{
// verify that STD3 ASCII rules are satisfied
if(useSTD3ASCIIRules == TRUE){
if( srcIsLDH == FALSE /* source contains some non-LDH characters */
|| src[0] == HYPHEN || src[srcLength-1] == HYPHEN){
*status = U_IDNA_STD3_ASCII_RULES_ERROR;
/* populate the parseError struct */
if(srcIsLDH==FALSE){
// failPos is always set the index of failure
uprv_syntaxError(src,failPos, srcLength,parseError);
}else if(src[0] == HYPHEN){
// fail position is 0
uprv_syntaxError(src,0,srcLength,parseError);
}else{
// the last index in the source is always length-1
uprv_syntaxError(src, (srcLength>0) ? srcLength-1 : srcLength, srcLength,parseError);
}
goto CLEANUP;
}
}
//copy the source to destination
if(srcLength <= destCapacity){
uprv_memmove(dest,src,srcLength * U_SIZEOF_UCHAR);
}
reqLength = srcLength;
}
CLEANUP:
if(b1 != b1Stack){
uprv_free(b1);
}
if(b2 != b2Stack){
uprv_free(b2);
}
uprv_free(caseFlags);
// delete prep;
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
/* helper function */
static UChar*
_strFromWCS( UChar *dest,
int32_t destCapacity,
int32_t *pDestLength,
const wchar_t *src,
int32_t srcLength,
UErrorCode *pErrorCode)
{
int32_t retVal =0, count =0 ;
UConverter* conv = NULL;
UChar* pTarget = NULL;
UChar* pTargetLimit = NULL;
UChar* target = NULL;
UChar uStack [_STACK_BUFFER_CAPACITY];
wchar_t wStack[_STACK_BUFFER_CAPACITY];
wchar_t* pWStack = wStack;
char cStack[_STACK_BUFFER_CAPACITY];
int32_t cStackCap = _STACK_BUFFER_CAPACITY;
char* pCSrc=cStack;
char* pCSave=pCSrc;
char* pCSrcLimit=NULL;
const wchar_t* pSrc = src;
const wchar_t* pSrcLimit = NULL;
if(srcLength ==-1){
/* if the wchar_t source is null terminated we can safely
* assume that there are no embedded nulls, this is a fast
* path for null terminated strings.
*/
for(;;){
/* convert wchars to chars */
retVal = uprv_wcstombs(pCSrc,src, cStackCap);
if(retVal == -1){
*pErrorCode = U_ILLEGAL_CHAR_FOUND;
goto cleanup;
}else if(retVal >= (cStackCap-1)){
/* Should rarely occur */
u_growAnyBufferFromStatic(cStack,(void**)&pCSrc,&cStackCap,
cStackCap * _BUFFER_CAPACITY_MULTIPLIER, 0, sizeof(char));
pCSave = pCSrc;
}else{
/* converted every thing */
pCSrc = pCSrc+retVal;
break;
}
}
}else{
/* here the source is not null terminated
* so it may have nulls embeded and we need to
* do some extra processing
*/
int32_t remaining =cStackCap;
pSrcLimit = src + srcLength;
for(;;){
register int32_t nulLen = 0;
/* find nulls in the string */
while(nulLen<srcLength && pSrc[nulLen++]!=0){
}
if((pSrc+nulLen) < pSrcLimit){
/* check if we have enough room in pCSrc */
if(remaining < (nulLen * MB_CUR_MAX)){
/* should rarely occur */
int32_t len = (pCSrc-pCSave);
pCSrc = pCSave;
/* we do not have enough room so grow the buffer*/
u_growAnyBufferFromStatic(cStack,(void**)&pCSrc,&cStackCap,
_BUFFER_CAPACITY_MULTIPLIER*cStackCap+(nulLen*MB_CUR_MAX),len,sizeof(char));
pCSave = pCSrc;
pCSrc = pCSave+len;
remaining = cStackCap-(pCSrc - pCSave);
}
/* we have found a null so convert the
* chunk from begining of non-null char to null
*/
retVal = uprv_wcstombs(pCSrc,pSrc,remaining);
if(retVal==-1){
/* an error occurred bail out */
*pErrorCode = U_ILLEGAL_CHAR_FOUND;
goto cleanup;
}
pCSrc += retVal+1 /* already null terminated */;
pSrc += nulLen; /* skip past the null */
srcLength-=nulLen; /* decrement the srcLength */
remaining -= (pCSrc-pCSave);
}else{
/* the source is not null terminated and we are
* end of source so we copy the source to a temp buffer
* null terminate it and convert wchar_ts to chars
*/
if(nulLen >= _STACK_BUFFER_CAPACITY){
/* Should rarely occcur */
/* allocate new buffer buffer */
pWStack =(wchar_t*) uprv_malloc(sizeof(wchar_t) * (nulLen + 1));
if(pWStack==NULL){
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
goto cleanup;
}
}
if(nulLen>0){
/* copy the contents to tempStack */
uprv_memcpy(pWStack,pSrc,nulLen*sizeof(wchar_t));
}
/* null terminate the tempBuffer */
pWStack[nulLen] =0 ;
if(remaining < (nulLen * MB_CUR_MAX)){
/* Should rarely occur */
int32_t len = (pCSrc-pCSave);
pCSrc = pCSave;
/* we do not have enough room so grow the buffer*/
u_growAnyBufferFromStatic(cStack,(void**)&pCSrc,&cStackCap,
cStackCap+(nulLen*MB_CUR_MAX),len,sizeof(char));
pCSave = pCSrc;
pCSrc = pCSave+len;
remaining = cStackCap-(pCSrc - pCSave);
}
/* convert to chars */
retVal = uprv_wcstombs(pCSrc,pWStack,remaining);
pCSrc += retVal;
pSrc += nulLen;
srcLength-=nulLen; /* decrement the srcLength */
break;
}
}
}
/* OK..now we have converted from wchar_ts to chars now
* convert chars to UChars
*/
pCSrcLimit = pCSrc;
pCSrc = pCSave;
pTarget = target= dest;
pTargetLimit = dest + destCapacity;
conv= u_getDefaultConverter(pErrorCode);
if(U_FAILURE(*pErrorCode)|| conv==NULL){
goto cleanup;
}
for(;;) {
*pErrorCode = U_ZERO_ERROR;
/* convert to stack buffer*/
ucnv_toUnicode(conv,&pTarget,pTargetLimit,(const char**)&pCSrc,pCSrcLimit,NULL,(UBool)(pCSrc==pCSrcLimit),pErrorCode);
/* increment count to number written to stack */
count+= pTarget - target;
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR){
target = uStack;
pTarget = uStack;
pTargetLimit = uStack + _STACK_BUFFER_CAPACITY;
} else {
break;
}
}
if(pDestLength){
*pDestLength =count;
}
u_terminateUChars(dest,destCapacity,count,pErrorCode);
cleanup:
if(cStack != pCSave){
uprv_free(pCSave);
}
if(wStack != pWStack){
uprv_free(pWStack);
}
u_releaseDefaultConverter(conv);
return dest;
}
U_CAPI int32_t U_EXPORT2
uidna_IDNToUnicode( const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UParseError* parseError,
UErrorCode* status){
if(status == NULL || U_FAILURE(*status)){
return 0;
}
if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
*status = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
int32_t reqLength = 0;
UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
if(U_FAILURE(*status)){
return 0;
}
//initialize pointers
UChar *delimiter = (UChar*)src;
UChar *labelStart = (UChar*)src;
UChar *currentDest = (UChar*) dest;
int32_t remainingLen = srcLength;
int32_t remainingDestCapacity = destCapacity;
int32_t labelLen = 0, labelReqLength = 0;
UBool done = FALSE;
for(;;){
labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done);
// The RFC states that
// <quote>
// ToUnicode never fails. If any step fails, then the original input
// is returned immediately in that step.
// </quote>
// _internal_toUnicode will copy the label.
/*if(labelLen==0 && done==FALSE){
*status = U_IDNA_ZERO_LENGTH_LABEL_ERROR;
break;
}*/
labelReqLength = _internal_toUnicode(labelStart, labelLen,
currentDest, remainingDestCapacity,
options, nameprep,
parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
*status = U_ZERO_ERROR; // reset error
remainingDestCapacity = 0;
}
if(U_FAILURE(*status)){
break;
}
reqLength +=labelReqLength;
// adjust the destination pointer
if(labelReqLength < remainingDestCapacity){
currentDest = currentDest + labelReqLength;
remainingDestCapacity -= labelReqLength;
}else{
// should never occur
remainingDestCapacity = 0;
}
if(done == TRUE){
break;
}
// add the label separator
// Unlike the ToASCII operation we don't normalize the label separators
if(remainingDestCapacity > 0){
*currentDest++ = *(labelStart + labelLen);
remainingDestCapacity--;
}
reqLength++;
labelStart = delimiter;
if(remainingLen >0 ){
remainingLen = (int32_t)(srcLength - (delimiter - src));
}
}
if(reqLength > MAX_DOMAIN_NAME_LENGTH){
*status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR;
}
usprep_close(nameprep);
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
int32_t umsg_autoQuoteApostrophe(const UChar* pattern,
int32_t patternLength,
UChar* dest,
int32_t destCapacity,
UErrorCode* ec)
{
int32_t state = STATE_INITIAL;
int32_t braceCount = 0;
int32_t len = 0;
if (ec == NULL || U_FAILURE(*ec)) {
return -1;
}
if (pattern == NULL || patternLength < -1 || (dest == NULL && destCapacity > 0)) {
*ec = U_ILLEGAL_ARGUMENT_ERROR;
return -1;
}
if (patternLength == -1) {
patternLength = u_strlen(pattern);
}
for (int i = 0; i < patternLength; ++i) {
UChar c = pattern[i];
switch (state) {
case STATE_INITIAL:
switch (c) {
case SINGLE_QUOTE:
state = STATE_SINGLE_QUOTE;
break;
case CURLY_BRACE_LEFT:
state = STATE_MSG_ELEMENT;
++braceCount;
break;
}
break;
case STATE_SINGLE_QUOTE:
switch (c) {
case SINGLE_QUOTE:
state = STATE_INITIAL;
break;
case CURLY_BRACE_LEFT:
case CURLY_BRACE_RIGHT:
state = STATE_IN_QUOTE;
break;
default:
MAppend(SINGLE_QUOTE);
state = STATE_INITIAL;
break;
}
break;
case STATE_IN_QUOTE:
switch (c) {
case SINGLE_QUOTE:
state = STATE_INITIAL;
break;
}
break;
case STATE_MSG_ELEMENT:
switch (c) {
case CURLY_BRACE_LEFT:
++braceCount;
break;
case CURLY_BRACE_RIGHT:
if (--braceCount == 0) {
state = STATE_INITIAL;
}
break;
}
break;
default: // Never happens.
break;
}
MAppend(c);
}
// End of scan
if (state == STATE_SINGLE_QUOTE || state == STATE_IN_QUOTE) {
MAppend(SINGLE_QUOTE);
}
return u_terminateUChars(dest, destCapacity, len, ec);
}
U_CAPI int32_t U_EXPORT2
uidna_IDNToASCII( const UChar *src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UParseError *parseError,
UErrorCode *status){
if(status == NULL || U_FAILURE(*status)){
return 0;
}
if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
*status = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
int32_t reqLength = 0;
UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
if(U_FAILURE(*status)){
return 0;
}
//initialize pointers
UChar *delimiter = (UChar*)src;
UChar *labelStart = (UChar*)src;
UChar *currentDest = (UChar*) dest;
int32_t remainingLen = srcLength;
int32_t remainingDestCapacity = destCapacity;
int32_t labelLen = 0, labelReqLength = 0;
UBool done = FALSE;
for(;;){
labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done);
labelReqLength = 0;
if(!(labelLen==0 && done)){// make sure this is not a root label separator.
labelReqLength = _internal_toASCII( labelStart, labelLen,
currentDest, remainingDestCapacity,
options, nameprep,
parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
*status = U_ZERO_ERROR; // reset error
remainingDestCapacity = 0;
}
}
if(U_FAILURE(*status)){
break;
}
reqLength +=labelReqLength;
// adjust the destination pointer
if(labelReqLength < remainingDestCapacity){
currentDest = currentDest + labelReqLength;
remainingDestCapacity -= labelReqLength;
}else{
// should never occur
remainingDestCapacity = 0;
}
if(done == TRUE){
break;
}
// add the label separator
if(remainingDestCapacity > 0){
*currentDest++ = FULL_STOP;
remainingDestCapacity--;
}
reqLength++;
labelStart = delimiter;
if(remainingLen >0 ){
remainingLen = (int32_t)(srcLength - (delimiter - src));
}
}
if(reqLength > MAX_DOMAIN_NAME_LENGTH){
*status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR;
}
usprep_close(nameprep);
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
static int32_t
_internal_toUnicode(const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UStringPrepProfile* nameprep,
UParseError* parseError,
UErrorCode* status)
{
//get the options
//UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0);
int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0;
// TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too.
UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE], b3Stack[MAX_LABEL_BUFFER_SIZE];
//initialize pointers to stack buffers
UChar *b1 = b1Stack, *b2 = b2Stack, *b1Prime=NULL, *b3=b3Stack;
int32_t b1Len, b2Len, b1PrimeLen, b3Len,
b1Capacity = MAX_LABEL_BUFFER_SIZE,
b2Capacity = MAX_LABEL_BUFFER_SIZE,
b3Capacity = MAX_LABEL_BUFFER_SIZE,
reqLength=0;
b1Len = 0;
UBool* caseFlags = NULL;
UBool srcIsASCII = TRUE;
/*UBool srcIsLDH = TRUE;
int32_t failPos =0;*/
// step 1: find out if all the codepoints in src are ASCII
if(srcLength==-1){
srcLength = 0;
for(;src[srcLength]!=0;){
if(src[srcLength]> 0x7f){
srcIsASCII = FALSE;
}/*else if(isLDHChar(src[srcLength])==FALSE){
// here we do not assemble surrogates
// since we know that LDH code points
// are in the ASCII range only
srcIsLDH = FALSE;
failPos = srcLength;
}*/
srcLength++;
}
}else if(srcLength > 0){
for(int32_t j=0; j<srcLength; j++){
if(src[j]> 0x7f){
srcIsASCII = FALSE;
}/*else if(isLDHChar(src[j])==FALSE){
// here we do not assemble surrogates
// since we know that LDH code points
// are in the ASCII range only
srcIsLDH = FALSE;
failPos = j;
}*/
}
}else{
return 0;
}
if(srcIsASCII == FALSE){
// step 2: process the string
b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status);
}
//bail out on error
if(U_FAILURE(*status)){
goto CLEANUP;
}
}else{
//just point src to b1
b1 = (UChar*) src;
b1Len = srcLength;
}
// The RFC states that
// <quote>
// ToUnicode never fails. If any step fails, then the original input
// is returned immediately in that step.
// </quote>
//step 3: verify ACE Prefix
if(startsWithPrefix(b1,b1Len)){
//step 4: Remove the ACE Prefix
b1Prime = b1 + ACE_PREFIX_LENGTH;
b1PrimeLen = b1Len - ACE_PREFIX_LENGTH;
//step 5: Decode using punycode
b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Capacity, caseFlags,status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
if(b2==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Len, caseFlags, status);
}
//step 6:Apply toASCII
b3Len = uidna_toASCII(b2, b2Len, b3, b3Capacity, options, parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b3 = (UChar*) uprv_malloc(b3Len * U_SIZEOF_UCHAR);
if(b3==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b3Len = uidna_toASCII(b2,b2Len,b3,b3Len,options,parseError, status);
}
//bail out on error
if(U_FAILURE(*status)){
goto CLEANUP;
}
//step 7: verify
if(compareCaseInsensitiveASCII(b1, b1Len, b3, b3Len) !=0){
// Cause the original to be returned.
*status = U_IDNA_VERIFICATION_ERROR;
goto CLEANUP;
}
//step 8: return output of step 5
reqLength = b2Len;
if(b2Len <= destCapacity) {
uprv_memmove(dest, b2, b2Len * U_SIZEOF_UCHAR);
}
}
else{
// See the start of this if statement for why this is commented out.
// verify that STD3 ASCII rules are satisfied
/*if(useSTD3ASCIIRules == TRUE){
if( srcIsLDH == FALSE // source contains some non-LDH characters
|| src[0] == HYPHEN || src[srcLength-1] == HYPHEN){
*status = U_IDNA_STD3_ASCII_RULES_ERROR;
// populate the parseError struct
if(srcIsLDH==FALSE){
// failPos is always set the index of failure
uprv_syntaxError(src,failPos, srcLength,parseError);
}else if(src[0] == HYPHEN){
// fail position is 0
uprv_syntaxError(src,0,srcLength,parseError);
}else{
// the last index in the source is always length-1
uprv_syntaxError(src, (srcLength>0) ? srcLength-1 : srcLength, srcLength,parseError);
}
goto CLEANUP;
}
}*/
// just return the source
//copy the source to destination
if(srcLength <= destCapacity){
uprv_memmove(dest,src,srcLength * U_SIZEOF_UCHAR);
}
reqLength = srcLength;
}
CLEANUP:
if(b1 != b1Stack && b1!=src){
uprv_free(b1);
}
if(b2 != b2Stack){
uprv_free(b2);
}
uprv_free(caseFlags);
// The RFC states that
// <quote>
// ToUnicode never fails. If any step fails, then the original input
// is returned immediately in that step.
// </quote>
// So if any step fails lets copy source to destination
if(U_FAILURE(*status)){
//copy the source to destination
if(dest && srcLength <= destCapacity){
// srcLength should have already been set earlier.
U_ASSERT(srcLength >= 0);
uprv_memmove(dest,src,srcLength * U_SIZEOF_UCHAR);
}
reqLength = srcLength;
*status = U_ZERO_ERROR;
}
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
static int32_t
_internal_toASCII(const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UStringPrepProfile* nameprep,
UParseError* parseError,
UErrorCode* status)
{
// TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too.
UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE];
//initialize pointers to stack buffers
UChar *b1 = b1Stack, *b2 = b2Stack;
int32_t b1Len=0, b2Len,
b1Capacity = MAX_LABEL_BUFFER_SIZE,
b2Capacity = MAX_LABEL_BUFFER_SIZE ,
reqLength=0;
int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0;
UBool* caseFlags = NULL;
// the source contains all ascii codepoints
UBool srcIsASCII = TRUE;
// assume the source contains all LDH codepoints
UBool srcIsLDH = TRUE;
int32_t j=0;
//get the options
UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0);
int32_t failPos = -1;
if(srcLength == -1){
srcLength = u_strlen(src);
}
if(srcLength > b1Capacity){
b1 = (UChar*) uprv_malloc(srcLength * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
b1Capacity = srcLength;
}
// step 1
for( j=0;j<srcLength;j++){
if(src[j] > 0x7F){
srcIsASCII = FALSE;
}
b1[b1Len++] = src[j];
}
// step 2 is performed only if the source contains non ASCII
if(srcIsASCII == FALSE){
// step 2
b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
// we do not have enough room so grow the buffer
if(b1 != b1Stack){
uprv_free(b1);
}
b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status);
}
}
// error bail out
if(U_FAILURE(*status)){
goto CLEANUP;
}
if(b1Len == 0){
*status = U_IDNA_ZERO_LENGTH_LABEL_ERROR;
goto CLEANUP;
}
// for step 3 & 4
srcIsASCII = TRUE;
for( j=0;j<b1Len;j++){
// check if output of usprep_prepare is all ASCII
if(b1[j] > 0x7F){
srcIsASCII = FALSE;
}else if(isLDHChar(b1[j])==FALSE){ // if the char is in ASCII range verify that it is an LDH character
srcIsLDH = FALSE;
failPos = j;
}
}
if(useSTD3ASCIIRules == TRUE){
// verify 3a and 3b
// 3(a) Verify the absence of non-LDH ASCII code points; that is, the
// absence of 0..2C, 2E..2F, 3A..40, 5B..60, and 7B..7F.
// 3(b) Verify the absence of leading and trailing hyphen-minus; that
// is, the absence of U+002D at the beginning and end of the
// sequence.
if( srcIsLDH == FALSE /* source at this point should not contain anyLDH characters */
|| b1[0] == HYPHEN || b1[b1Len-1] == HYPHEN){
*status = U_IDNA_STD3_ASCII_RULES_ERROR;
/* populate the parseError struct */
if(srcIsLDH==FALSE){
// failPos is always set the index of failure
uprv_syntaxError(b1,failPos, b1Len,parseError);
}else if(b1[0] == HYPHEN){
// fail position is 0
uprv_syntaxError(b1,0,b1Len,parseError);
}else{
// the last index in the source is always length-1
uprv_syntaxError(b1, (b1Len>0) ? b1Len-1 : b1Len, b1Len,parseError);
}
goto CLEANUP;
}
}
// Step 4: if the source is ASCII then proceed to step 8
if(srcIsASCII){
if(b1Len <= destCapacity){
uprv_memmove(dest, b1, b1Len * U_SIZEOF_UCHAR);
reqLength = b1Len;
}else{
reqLength = b1Len;
goto CLEANUP;
}
}else{
// step 5 : verify the sequence does not begin with ACE prefix
if(!startsWithPrefix(b1,b1Len)){
//step 6: encode the sequence with punycode
// do not preserve the case flags for now!
// TODO: Preserve the case while implementing the RFE
// caseFlags = (UBool*) uprv_malloc(b1Len * sizeof(UBool));
// uprv_memset(caseFlags,TRUE,b1Len);
b2Len = u_strToPunycode(b1,b1Len,b2,b2Capacity,caseFlags, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
if(b2 == NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b2Len = u_strToPunycode(b1,b1Len,b2,b2Len,caseFlags, status);
}
//error bail out
if(U_FAILURE(*status)){
goto CLEANUP;
}
// TODO : Reconsider while implementing the case preserve RFE
// convert all codepoints to lower case ASCII
// toASCIILower(b2,b2Len);
reqLength = b2Len+ACE_PREFIX_LENGTH;
if(reqLength > destCapacity){
*status = U_BUFFER_OVERFLOW_ERROR;
goto CLEANUP;
}
//Step 7: prepend the ACE prefix
uprv_memcpy(dest,ACE_PREFIX,ACE_PREFIX_LENGTH * U_SIZEOF_UCHAR);
//Step 6: copy the contents in b2 into dest
uprv_memcpy(dest+ACE_PREFIX_LENGTH, b2, b2Len * U_SIZEOF_UCHAR);
}else{
*status = U_IDNA_ACE_PREFIX_ERROR;
//position of failure is 0
uprv_syntaxError(b1,0,b1Len,parseError);
goto CLEANUP;
}
}
// step 8: verify the length of label
if(reqLength > MAX_LABEL_LENGTH){
*status = U_IDNA_LABEL_TOO_LONG_ERROR;
}
CLEANUP:
if(b1 != b1Stack){
uprv_free(b1);
}
if(b2 != b2Stack){
uprv_free(b2);
}
uprv_free(caseFlags);
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
U_CAPI int32_t U_EXPORT2
ubidi_writeReordered(UBiDi *pBiDi,
UChar *dest, int32_t destSize,
uint16_t options,
UErrorCode *pErrorCode) {
const UChar *text;
UChar *saveDest;
int32_t length, destCapacity;
int32_t run, runCount, logicalStart, runLength;
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
/* more error checking */
if( pBiDi==NULL ||
(text=pBiDi->text)==NULL || (length=pBiDi->length)<0 ||
destSize<0 || (destSize>0 && dest==NULL))
{
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/* do input and output overlap? */
if( dest!=NULL &&
((text>=dest && text<dest+destSize) ||
(dest>=text && dest<text+pBiDi->originalLength)))
{
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
if(length==0) {
/* nothing to do */
return u_terminateUChars(dest, destSize, 0, pErrorCode);
}
runCount=ubidi_countRuns(pBiDi, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
return 0;
}
/* destSize shrinks, later destination length=destCapacity-destSize */
saveDest=dest;
destCapacity=destSize;
/*
* Option "insert marks" implies UBIDI_INSERT_LRM_FOR_NUMERIC if the
* reordering mode (checked below) is appropriate.
*/
if(pBiDi->reorderingOptions & UBIDI_OPTION_INSERT_MARKS) {
options|=UBIDI_INSERT_LRM_FOR_NUMERIC;
options&=~UBIDI_REMOVE_BIDI_CONTROLS;
}
/*
* Option "remove controls" implies UBIDI_REMOVE_BIDI_CONTROLS
* and cancels UBIDI_INSERT_LRM_FOR_NUMERIC.
*/
if(pBiDi->reorderingOptions & UBIDI_OPTION_REMOVE_CONTROLS) {
options|=UBIDI_REMOVE_BIDI_CONTROLS;
options&=~UBIDI_INSERT_LRM_FOR_NUMERIC;
}
/*
* If we do not perform the "inverse BiDi" algorithm, then we
* don't need to insert any LRMs, and don't need to test for it.
*/
if((pBiDi->reorderingMode != UBIDI_REORDER_INVERSE_NUMBERS_AS_L) &&
(pBiDi->reorderingMode != UBIDI_REORDER_INVERSE_LIKE_DIRECT) &&
(pBiDi->reorderingMode != UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL) &&
(pBiDi->reorderingMode != UBIDI_REORDER_RUNS_ONLY)) {
options&=~UBIDI_INSERT_LRM_FOR_NUMERIC;
}
/*
* Iterate through all visual runs and copy the run text segments to
* the destination, according to the options.
*
* The tests for where to insert LRMs ignore the fact that there may be
* BN codes or non-BMP code points at the beginning and end of a run;
* they may insert LRMs unnecessarily but the tests are faster this way
* (this would have to be improved for UTF-8).
*
* Note that the only errors that are set by doWriteXY() are buffer overflow
* errors. Ignore them until the end, and continue for preflighting.
*/
if(!(options&UBIDI_OUTPUT_REVERSE)) {
/* forward output */
if(!(options&UBIDI_INSERT_LRM_FOR_NUMERIC)) {
/* do not insert BiDi controls */
for(run=0; run<runCount; ++run) {
if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength)) {
runLength=doWriteForward(text+logicalStart, runLength,
dest, destSize,
(uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
} else {
runLength=doWriteReverse(text+logicalStart, runLength,
dest, destSize,
options, pErrorCode);
}
dest+=runLength;
destSize-=runLength;
}
} else {
/* insert BiDi controls for "inverse BiDi" */
const DirProp *dirProps=pBiDi->dirProps;
const UChar *src;
UChar uc;
UBiDiDirection dir;
int32_t markFlag;
for(run=0; run<runCount; ++run) {
dir=ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength);
src=text+logicalStart;
/* check if something relevant in insertPoints */
markFlag=pBiDi->runs[run].insertRemove;
if(markFlag<0) { /* BiDi controls count */
markFlag=0;
}
if(UBIDI_LTR==dir) {
if((pBiDi->isInverse) &&
(/*run>0 &&*/ dirProps[logicalStart]!=L)) {
markFlag |= LRM_BEFORE;
}
if (markFlag & LRM_BEFORE) {
uc=LRM_CHAR;
}
else if (markFlag & RLM_BEFORE) {
uc=RLM_CHAR;
}
else uc=0;
if(uc) {
if(destSize>0) {
*dest++=uc;
}
--destSize;
}
runLength=doWriteForward(src, runLength,
dest, destSize,
(uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
dest+=runLength;
destSize-=runLength;
if((pBiDi->isInverse) &&
(/*run<runCount-1 &&*/ dirProps[logicalStart+runLength-1]!=L)) {
markFlag |= LRM_AFTER;
}
if (markFlag & LRM_AFTER) {
uc=LRM_CHAR;
}
else if (markFlag & RLM_AFTER) {
uc=RLM_CHAR;
}
else uc=0;
if(uc) {
if(destSize>0) {
*dest++=uc;
}
--destSize;
}
} else { /* RTL run */
if((pBiDi->isInverse) &&
(/*run>0 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart+runLength-1])))) {
markFlag |= RLM_BEFORE;
}
if (markFlag & LRM_BEFORE) {
uc=LRM_CHAR;
}
else if (markFlag & RLM_BEFORE) {
uc=RLM_CHAR;
}
else uc=0;
if(uc) {
if(destSize>0) {
*dest++=uc;
}
--destSize;
}
runLength=doWriteReverse(src, runLength,
dest, destSize,
options, pErrorCode);
dest+=runLength;
destSize-=runLength;
if((pBiDi->isInverse) &&
(/*run<runCount-1 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart])))) {
markFlag |= RLM_AFTER;
}
if (markFlag & LRM_AFTER) {
uc=LRM_CHAR;
}
else if (markFlag & RLM_AFTER) {
uc=RLM_CHAR;
}
else uc=0;
if(uc) {
if(destSize>0) {
*dest++=uc;
}
--destSize;
}
}
}
}
} else {
/* reverse output */
if(!(options&UBIDI_INSERT_LRM_FOR_NUMERIC)) {
/* do not insert BiDi controls */
for(run=runCount; --run>=0;) {
if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength)) {
runLength=doWriteReverse(text+logicalStart, runLength,
dest, destSize,
(uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
} else {
runLength=doWriteForward(text+logicalStart, runLength,
dest, destSize,
options, pErrorCode);
}
dest+=runLength;
destSize-=runLength;
}
} else {
/* insert BiDi controls for "inverse BiDi" */
const DirProp *dirProps=pBiDi->dirProps;
const UChar *src;
UBiDiDirection dir;
for(run=runCount; --run>=0;) {
/* reverse output */
dir=ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength);
src=text+logicalStart;
if(UBIDI_LTR==dir) {
if(/*run<runCount-1 &&*/ dirProps[logicalStart+runLength-1]!=L) {
if(destSize>0) {
*dest++=LRM_CHAR;
}
--destSize;
}
runLength=doWriteReverse(src, runLength,
dest, destSize,
(uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
dest+=runLength;
destSize-=runLength;
if(/*run>0 &&*/ dirProps[logicalStart]!=L) {
if(destSize>0) {
*dest++=LRM_CHAR;
}
--destSize;
}
} else {
if(/*run<runCount-1 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart]))) {
if(destSize>0) {
*dest++=RLM_CHAR;
}
--destSize;
}
runLength=doWriteForward(src, runLength,
dest, destSize,
options, pErrorCode);
dest+=runLength;
destSize-=runLength;
if(/*run>0 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart+runLength-1]))) {
if(destSize>0) {
*dest++=RLM_CHAR;
}
--destSize;
}
}
}
}
}
return u_terminateUChars(saveDest, destCapacity, destCapacity-destSize, pErrorCode);
}
static int32_t
caseMap(UChar *dest, int32_t destCapacity,
const UChar *src, int32_t srcLength,
UBreakIterator *titleIter,
const char *locale,
uint32_t options,
int32_t toWhichCase,
UErrorCode *pErrorCode) {
UChar buffer[300];
UChar *temp;
const UCaseProps *csp;
int32_t destLength;
UBool ownTitleIter;
/* check argument values */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
if( destCapacity<0 ||
(dest==NULL && destCapacity>0) ||
src==NULL ||
srcLength<-1
) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
csp=ucase_getSingleton(pErrorCode);
if(U_FAILURE(*pErrorCode)) {
return 0;
}
/* get the string length */
if(srcLength==-1) {
srcLength=u_strlen(src);
}
/* check for overlapping source and destination */
if( dest!=NULL &&
((src>=dest && src<(dest+destCapacity)) ||
(dest>=src && dest<(src+srcLength)))
) {
/* overlap: provide a temporary destination buffer and later copy the result */
if(destCapacity<=(sizeof(buffer)/U_SIZEOF_UCHAR)) {
/* the stack buffer is large enough */
temp=buffer;
} else {
/* allocate a buffer */
temp=(UChar *)uprv_malloc(destCapacity*U_SIZEOF_UCHAR);
if(temp==NULL) {
*pErrorCode=U_MEMORY_ALLOCATION_ERROR;
return 0;
}
}
} else {
temp=dest;
}
ownTitleIter=FALSE;
destLength=0;
if(toWhichCase==FOLD_CASE) {
destLength=ustr_foldCase(csp, temp, destCapacity, src, srcLength,
options, pErrorCode);
} else {
UCaseContext csc={ NULL };
int32_t locCache;
csc.p=(void *)src;
csc.limit=srcLength;
locCache=0;
/* the internal functions require locale!=NULL */
if(locale==NULL) {
locale=uloc_getDefault();
}
if(toWhichCase==TO_LOWER) {
destLength=_caseMap(csp, ucase_toFullLower,
temp, destCapacity,
src, &csc,
0, srcLength,
locale, &locCache, pErrorCode);
} else if(toWhichCase==TO_UPPER) {
destLength=_caseMap(csp, ucase_toFullUpper,
temp, destCapacity,
src, &csc,
0, srcLength,
locale, &locCache, pErrorCode);
} else /* if(toWhichCase==TO_TITLE) */ {
#if UCONFIG_NO_BREAK_ITERATION
*pErrorCode=U_UNSUPPORTED_ERROR;
#else
if(titleIter==NULL) {
titleIter=ubrk_open(UBRK_WORD, locale,
src, srcLength,
pErrorCode);
ownTitleIter=(UBool)U_SUCCESS(*pErrorCode);
}
if(U_SUCCESS(*pErrorCode)) {
destLength=_toTitle(csp, temp, destCapacity,
src, &csc, srcLength,
titleIter, locale, &locCache, pErrorCode);
}
#endif
}
}
if(temp!=dest) {
/* copy the result string to the destination buffer */
if(destLength>0) {
int32_t copyLength= destLength<=destCapacity ? destLength : destCapacity;
if(copyLength>0) {
uprv_memmove(dest, temp, copyLength*U_SIZEOF_UCHAR);
}
}
if(temp!=buffer) {
uprv_free(temp);
}
}
#if !UCONFIG_NO_BREAK_ITERATION
if(ownTitleIter) {
ubrk_close(titleIter);
}
#endif
return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
}
static int32_t
caseMap(const UCaseMap *csm,
UChar *dest, int32_t destCapacity,
const UChar *src, int32_t srcLength,
int32_t toWhichCase,
UErrorCode *pErrorCode) {
UChar buffer[300];
UChar *temp;
int32_t destLength;
/* check argument values */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
if( destCapacity<0 ||
(dest==NULL && destCapacity>0) ||
src==NULL ||
srcLength<-1
) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/* get the string length */
if(srcLength==-1) {
srcLength=u_strlen(src);
}
/* check for overlapping source and destination */
if( dest!=NULL &&
((src>=dest && src<(dest+destCapacity)) ||
(dest>=src && dest<(src+srcLength)))
) {
/* overlap: provide a temporary destination buffer and later copy the result */
if(destCapacity<=(sizeof(buffer)/U_SIZEOF_UCHAR)) {
/* the stack buffer is large enough */
temp=buffer;
} else {
/* allocate a buffer */
temp=(UChar *)uprv_malloc(destCapacity*U_SIZEOF_UCHAR);
if(temp==NULL) {
*pErrorCode=U_MEMORY_ALLOCATION_ERROR;
return 0;
}
}
} else {
temp=dest;
}
destLength=0;
if(toWhichCase==FOLD_CASE) {
destLength=ustr_foldCase(csm->csp, temp, destCapacity, src, srcLength,
csm->options, pErrorCode);
} else {
UCaseContext csc={ NULL };
csc.p=(void *)src;
csc.limit=srcLength;
if(toWhichCase==TO_LOWER) {
destLength=_caseMap(csm, ucase_toFullLower,
temp, destCapacity,
src, &csc,
0, srcLength,
pErrorCode);
} else if(toWhichCase==TO_UPPER) {
destLength=_caseMap(csm, ucase_toFullUpper,
temp, destCapacity,
src, &csc,
0, srcLength,
pErrorCode);
} else /* if(toWhichCase==TO_TITLE) */ {
#if UCONFIG_NO_BREAK_ITERATION
*pErrorCode=U_UNSUPPORTED_ERROR;
#else
/* UCaseMap is actually non-const in toTitle() APIs. */
destLength=_toTitle((UCaseMap *)csm, temp, destCapacity,
src, &csc, srcLength,
pErrorCode);
#endif
}
}
if(temp!=dest) {
/* copy the result string to the destination buffer */
if(destLength>0) {
int32_t copyLength= destLength<=destCapacity ? destLength : destCapacity;
if(copyLength>0) {
uprv_memmove(dest, temp, copyLength*U_SIZEOF_UCHAR);
}
}
if(temp!=buffer) {
uprv_free(temp);
}
}
return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
}
U_CFUNC int32_t
ustrcase_map(const UCaseMap *csm,
UChar *dest, int32_t destCapacity,
const UChar *src, int32_t srcLength,
UStringCaseMapper *stringCaseMapper,
UErrorCode *pErrorCode) {
UChar buffer[300];
UChar *temp;
int32_t destLength;
/* check argument values */
if(U_FAILURE(*pErrorCode)) {
return 0;
}
if( destCapacity<0 ||
(dest==NULL && destCapacity>0) ||
src==NULL ||
srcLength<-1
) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/* get the string length */
if(srcLength==-1) {
srcLength=u_strlen(src);
}
/* check for overlapping source and destination */
if( dest!=NULL &&
((src>=dest && src<(dest+destCapacity)) ||
(dest>=src && dest<(src+srcLength)))
) {
/* overlap: provide a temporary destination buffer and later copy the result */
if(destCapacity<=UPRV_LENGTHOF(buffer)) {
/* the stack buffer is large enough */
temp=buffer;
} else {
/* allocate a buffer */
temp=(UChar *)uprv_malloc(destCapacity*U_SIZEOF_UCHAR);
if(temp==NULL) {
*pErrorCode=U_MEMORY_ALLOCATION_ERROR;
return 0;
}
}
} else {
temp=dest;
}
destLength=stringCaseMapper(csm, temp, destCapacity, src, srcLength, pErrorCode);
if(temp!=dest) {
/* copy the result string to the destination buffer */
if(destLength>0) {
int32_t copyLength= destLength<=destCapacity ? destLength : destCapacity;
if(copyLength>0) {
uprv_memmove(dest, temp, copyLength*U_SIZEOF_UCHAR);
}
}
if(temp!=buffer) {
uprv_free(temp);
}
}
return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
}
int32_t NamePrepTransform::process( const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
UBool allowUnassigned,
UParseError* parseError,
UErrorCode& status ){
// check error status
if(U_FAILURE(status)){
return 0;
}
//check arguments
if(src==NULL || srcLength<-1 || (dest==NULL && destCapacity!=0)) {
status=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
UChar b1Stack[MAX_BUFFER_SIZE];
UChar *b1 = b1Stack;
int32_t b1Len,b1Capacity = MAX_BUFFER_SIZE;
int32_t b1Index = 0;
UCharDirection direction=U_CHAR_DIRECTION_COUNT, firstCharDir=U_CHAR_DIRECTION_COUNT;
UBool leftToRight=FALSE, rightToLeft=FALSE;
b1Len = map(src,srcLength, b1, b1Capacity,allowUnassigned,parseError, status);
if(status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
if(!u_growBufferFromStatic(b1Stack,&b1,&b1Capacity,b1Len,0)){
status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
status = U_ZERO_ERROR; // reset error
b1Len = map(src,srcLength, b1, b1Len,allowUnassigned, parseError, status);
}
if(U_FAILURE(status)){
goto CLEANUP;
}
for(; b1Index<b1Len; ){
UChar32 ch = 0;
U16_NEXT(b1, b1Index, b1Len, ch);
if(prohibited.contains(ch) && ch!=0x0020){
status = U_IDNA_PROHIBITED_ERROR;
goto CLEANUP;
}
direction = u_charDirection(ch);
if(firstCharDir==U_CHAR_DIRECTION_COUNT){
firstCharDir = direction;
}
if(direction == U_LEFT_TO_RIGHT){
leftToRight = TRUE;
}
if(direction == U_RIGHT_TO_LEFT || direction == U_RIGHT_TO_LEFT_ARABIC){
rightToLeft = TRUE;
}
}
// satisfy 2
if( leftToRight == TRUE && rightToLeft == TRUE){
status = U_IDNA_CHECK_BIDI_ERROR;
goto CLEANUP;
}
//satisfy 3
if( rightToLeft == TRUE &&
!((firstCharDir == U_RIGHT_TO_LEFT || firstCharDir == U_RIGHT_TO_LEFT_ARABIC) &&
(direction == U_RIGHT_TO_LEFT || direction == U_RIGHT_TO_LEFT_ARABIC))
){
status = U_IDNA_CHECK_BIDI_ERROR;
return FALSE;
}
if(b1Len <= destCapacity){
uprv_memmove(dest,b1, b1Len*U_SIZEOF_UCHAR);
}
CLEANUP:
if(b1!=b1Stack){
uprv_free(b1);
}
return u_terminateUChars(dest, destCapacity, b1Len, &status);
}
U_CAPI int32_t U_EXPORT2
ucurr_forLocale(const char* locale,
UChar* buff,
int32_t buffCapacity,
UErrorCode* ec)
{
int32_t resLen = 0;
const UChar* s = NULL;
if (ec != NULL && U_SUCCESS(*ec)) {
if ((buff && buffCapacity) || !buffCapacity) {
UErrorCode localStatus = U_ZERO_ERROR;
char id[ULOC_FULLNAME_CAPACITY];
if ((resLen = uloc_getKeywordValue(locale, "currency", id, ULOC_FULLNAME_CAPACITY, &localStatus))) {
// there is a currency keyword. Try to see if it's valid
if(buffCapacity > resLen) {
u_charsToUChars(id, buff, resLen);
}
} else {
// get country or country_variant in `id'
uint32_t variantType = idForLocale(locale, id, sizeof(id), ec);
if (U_FAILURE(*ec)) {
return 0;
}
#if !UCONFIG_NO_SERVICE
const UChar* result = CReg::get(id);
if (result) {
if(buffCapacity > u_strlen(result)) {
u_strcpy(buff, result);
}
return u_strlen(result);
}
#endif
// Remove variants, which is only needed for registration.
char *idDelim = strchr(id, VAR_DELIM);
if (idDelim) {
idDelim[0] = 0;
}
// Look up the CurrencyMap element in the root bundle.
UResourceBundle *rb = ures_openDirect(NULL, CURRENCY_DATA, &localStatus);
UResourceBundle *cm = ures_getByKey(rb, CURRENCY_MAP, rb, &localStatus);
UResourceBundle *countryArray = ures_getByKey(rb, id, cm, &localStatus);
UResourceBundle *currencyReq = ures_getByIndex(countryArray, 0, NULL, &localStatus);
s = ures_getStringByKey(currencyReq, "id", &resLen, &localStatus);
/*
Get the second item when PREEURO is requested, and this is a known Euro country.
If the requested variant is PREEURO, and this isn't a Euro country, assume
that the country changed over to the Euro in the future. This is probably
an old version of ICU that hasn't been updated yet. The latest currency is
probably correct.
*/
if (U_SUCCESS(localStatus)) {
if ((variantType & VARIANT_IS_PREEURO) && u_strcmp(s, EUR_STR) == 0) {
currencyReq = ures_getByIndex(countryArray, 1, currencyReq, &localStatus);
s = ures_getStringByKey(currencyReq, "id", &resLen, &localStatus);
}
else if ((variantType & VARIANT_IS_EURO)) {
s = EUR_STR;
}
}
ures_close(countryArray);
ures_close(currencyReq);
if ((U_FAILURE(localStatus)) && strchr(id, '_') != 0)
{
// We don't know about it. Check to see if we support the variant.
uloc_getParent(locale, id, sizeof(id), ec);
*ec = U_USING_FALLBACK_WARNING;
return ucurr_forLocale(id, buff, buffCapacity, ec);
}
else if (*ec == U_ZERO_ERROR || localStatus != U_ZERO_ERROR) {
// There is nothing to fallback to. Report the failure/warning if possible.
*ec = localStatus;
}
if (U_SUCCESS(*ec)) {
if(buffCapacity > resLen) {
u_strcpy(buff, s);
}
}
}
return u_terminateUChars(buff, buffCapacity, resLen, ec);
} else {
*ec = U_ILLEGAL_ARGUMENT_ERROR;
}
}
return resLen;
}
U_CFUNC int32_t
idnaref_toASCII(const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UParseError* parseError,
UErrorCode* status){
if(status == NULL || U_FAILURE(*status)){
return 0;
}
if((src == NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
*status = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE];
//initialize pointers to stack buffers
UChar *b1 = b1Stack, *b2 = b2Stack;
int32_t b1Len, b2Len,
b1Capacity = MAX_LABEL_BUFFER_SIZE,
b2Capacity = MAX_LABEL_BUFFER_SIZE ,
reqLength=0;
//get the options
UBool allowUnassigned = (UBool)((options & IDNAREF_ALLOW_UNASSIGNED) != 0);
UBool useSTD3ASCIIRules = (UBool)((options & IDNAREF_USE_STD3_RULES) != 0);
UBool* caseFlags = NULL;
// assume the source contains all ascii codepoints
UBool srcIsASCII = TRUE;
// assume the source contains all LDH codepoints
UBool srcIsLDH = TRUE;
int32_t j=0;
// UParseError parseError;
// step 2
NamePrepTransform* prep = TestIDNA::getInstance(*status);
if(U_FAILURE(*status)){
goto CLEANUP;
}
b1Len = prep->process(src,srcLength,b1, b1Capacity,allowUnassigned,parseError,*status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b1Len = prep->process(src,srcLength,b1, b1Len,allowUnassigned, parseError, *status);
}
// error bail out
if(U_FAILURE(*status)){
goto CLEANUP;
}
// step 3 & 4
for( j=0;j<b1Len;j++){
if(b1[j] > 0x7F){
srcIsASCII = FALSE;
}else if(prep->isLDHChar(b1[j])==FALSE){ // if the char is in ASCII range verify that it is an LDH character{
srcIsLDH = FALSE;
}
}
if(useSTD3ASCIIRules == TRUE){
// verify 3a and 3b
if( srcIsLDH == FALSE /* source contains some non-LDH characters */
|| b1[0] == HYPHEN || b1[b1Len-1] == HYPHEN){
*status = U_IDNA_STD3_ASCII_RULES_ERROR;
goto CLEANUP;
}
}
if(srcIsASCII){
if(b1Len <= destCapacity){
uprv_memmove(dest, b1, b1Len * U_SIZEOF_UCHAR);
reqLength = b1Len;
}else{
reqLength = b1Len;
goto CLEANUP;
}
}else{
// step 5 : verify the sequence does not begin with ACE prefix
if(!startsWithPrefix(b1,b1Len)){
//step 6: encode the sequence with punycode
//caseFlags = (UBool*) uprv_malloc(b1Len * sizeof(UBool));
b2Len = convertToPuny(b1,b1Len, b2,b2Capacity,*status);
//b2Len = u_strToPunycode(b2,b2Capacity,b1,b1Len, caseFlags, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
if(b2 == NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b2Len = convertToPuny(b1, b1Len, b2, b2Len, *status);
//b2Len = u_strToPunycode(b2,b2Len,b1,b1Len, caseFlags, status);
}
//error bail out
if(U_FAILURE(*status)){
goto CLEANUP;
}
reqLength = b2Len+ACE_PREFIX_LENGTH;
if(reqLength > destCapacity){
*status = U_BUFFER_OVERFLOW_ERROR;
goto CLEANUP;
}
//Step 7: prepend the ACE prefix
uprv_memcpy(dest,ACE_PREFIX,ACE_PREFIX_LENGTH * U_SIZEOF_UCHAR);
//Step 6: copy the contents in b2 into dest
uprv_memcpy(dest+ACE_PREFIX_LENGTH, b2, b2Len * U_SIZEOF_UCHAR);
}else{
*status = U_IDNA_ACE_PREFIX_ERROR;
goto CLEANUP;
}
}
if(reqLength > MAX_LABEL_LENGTH){
*status = U_IDNA_LABEL_TOO_LONG_ERROR;
}
CLEANUP:
if(b1 != b1Stack){
uprv_free(b1);
}
if(b2 != b2Stack){
uprv_free(b2);
}
uprv_free(caseFlags);
// delete prep;
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
U_CAPI int32_t U_EXPORT2
uloc_getDisplayName(const char * locale,
const char * displayLocale,
UChar * dest, int32_t destCapacity,
UErrorCode * pErrorCode)
{
int32_t length, length2, length3 = 0;
UBool hasLanguage, hasScript, hasCountry, hasVariant, hasKeywords;
UEnumeration * keywordEnum = NULL;
int32_t keywordCount = 0;
const char * keyword = NULL;
int32_t keywordLen = 0;
char keywordValue[256];
int32_t keywordValueLen = 0;
int32_t locSepLen = 0;
int32_t locPatLen = 0;
int32_t p0Len = 0;
int32_t defaultPatternLen = 9;
const UChar * dispLocSeparator;
const UChar * dispLocPattern;
static const UChar defaultSeparator[3] = { 0x002c, 0x0020 , 0x0000 }; /* comma + space */
static const UChar defaultPattern[10] = { 0x007b, 0x0030, 0x007d, 0x0020, 0x0028, 0x007b, 0x0031, 0x007d, 0x0029, 0x0000 }; /* {0} ({1}) */
static const UChar pat0[4] = { 0x007b, 0x0030, 0x007d , 0x0000 } ; /* {0} */
static const UChar pat1[4] = { 0x007b, 0x0031, 0x007d , 0x0000 } ; /* {1} */
UResourceBundle * bundle = NULL;
UResourceBundle * locdsppat = NULL;
UErrorCode status = U_ZERO_ERROR;
/* argument checking */
if (pErrorCode == NULL || U_FAILURE(*pErrorCode))
{
return 0;
}
if (destCapacity < 0 || (destCapacity > 0 && dest == NULL))
{
*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
bundle = ures_open(U_ICUDATA_LANG, displayLocale, &status);
locdsppat = ures_getByKeyWithFallback(bundle, _kLocaleDisplayPattern, NULL, &status);
dispLocSeparator = ures_getStringByKeyWithFallback(locdsppat, _kSeparator, &locSepLen, &status);
dispLocPattern = ures_getStringByKeyWithFallback(locdsppat, _kPattern, &locPatLen, &status);
/*close the bundles */
ures_close(locdsppat);
ures_close(bundle);
/* If we couldn't find any data, then use the defaults */
if (locSepLen == 0)
{
dispLocSeparator = defaultSeparator;
locSepLen = 2;
}
if (locPatLen == 0)
{
dispLocPattern = defaultPattern;
locPatLen = 9;
}
/*
* if there is a language, then write "language (country, variant)"
* otherwise write "country, variant"
*/
/* write the language */
length = uloc_getDisplayLanguage(locale, displayLocale,
dest, destCapacity,
pErrorCode);
hasLanguage = length > 0;
if (hasLanguage)
{
p0Len = length;
/* append " (" */
if (length < destCapacity)
{
dest[length] = 0x20;
}
++length;
if (length < destCapacity)
{
dest[length] = 0x28;
}
++length;
}
if (*pErrorCode == U_BUFFER_OVERFLOW_ERROR)
{
/* keep preflighting */
*pErrorCode = U_ZERO_ERROR;
}
/* append the script */
if (length < destCapacity)
{
length2 = uloc_getDisplayScript(locale, displayLocale,
dest + length, destCapacity - length,
pErrorCode);
}
else
{
length2 = uloc_getDisplayScript(locale, displayLocale,
NULL, 0,
pErrorCode);
}
hasScript = length2 > 0;
length += length2;
if (hasScript)
{
/* append separator */
if (length + locSepLen <= destCapacity)
{
u_memcpy(dest + length, dispLocSeparator, locSepLen);
}
length += locSepLen;
}
if (*pErrorCode == U_BUFFER_OVERFLOW_ERROR)
{
/* keep preflighting */
*pErrorCode = U_ZERO_ERROR;
}
/* append the country */
if (length < destCapacity)
{
length2 = uloc_getDisplayCountry(locale, displayLocale,
dest + length, destCapacity - length,
pErrorCode);
}
else
{
length2 = uloc_getDisplayCountry(locale, displayLocale,
NULL, 0,
pErrorCode);
}
hasCountry = length2 > 0;
length += length2;
if (hasCountry)
{
/* append separator */
if (length + locSepLen <= destCapacity)
{
u_memcpy(dest + length, dispLocSeparator, locSepLen);
}
length += locSepLen;
}
if (*pErrorCode == U_BUFFER_OVERFLOW_ERROR)
{
/* keep preflighting */
*pErrorCode = U_ZERO_ERROR;
}
/* append the variant */
if (length < destCapacity)
{
length2 = uloc_getDisplayVariant(locale, displayLocale,
dest + length, destCapacity - length,
pErrorCode);
}
else
{
length2 = uloc_getDisplayVariant(locale, displayLocale,
NULL, 0,
pErrorCode);
}
hasVariant = length2 > 0;
length += length2;
if (hasVariant)
{
/* append separator */
if (length + locSepLen <= destCapacity)
{
u_memcpy(dest + length, dispLocSeparator, locSepLen);
}
length += locSepLen;
}
keywordEnum = uloc_openKeywords(locale, pErrorCode);
for (keywordCount = uenum_count(keywordEnum, pErrorCode); keywordCount > 0 ; keywordCount--)
{
if (U_FAILURE(*pErrorCode))
{
break;
}
/* the uenum_next returns NUL terminated string */
keyword = uenum_next(keywordEnum, &keywordLen, pErrorCode);
if (length + length3 < destCapacity)
{
length3 += uloc_getDisplayKeyword(keyword, displayLocale, dest + length + length3, destCapacity - length - length3,
pErrorCode);
}
else
{
length3 += uloc_getDisplayKeyword(keyword, displayLocale, NULL, 0, pErrorCode);
}
if (*pErrorCode == U_BUFFER_OVERFLOW_ERROR)
{
/* keep preflighting */
*pErrorCode = U_ZERO_ERROR;
}
keywordValueLen = uloc_getKeywordValue(locale, keyword, keywordValue, 256, pErrorCode);
if (keywordValueLen)
{
if (length + length3 < destCapacity)
{
dest[length + length3] = 0x3D;
}
length3++;
if (length + length3 < destCapacity)
{
length3 += uloc_getDisplayKeywordValue(locale, keyword, displayLocale, dest + length + length3,
destCapacity - length - length3, pErrorCode);
}
else
{
length3 += uloc_getDisplayKeywordValue(locale, keyword, displayLocale, NULL, 0, pErrorCode);
}
if (*pErrorCode == U_BUFFER_OVERFLOW_ERROR)
{
/* keep preflighting */
*pErrorCode = U_ZERO_ERROR;
}
}
if (keywordCount > 1)
{
if (length + length3 + locSepLen <= destCapacity && keywordCount)
{
u_memcpy(dest + length + length3, dispLocSeparator, locSepLen);
length3 += locSepLen;
}
}
}
uenum_close(keywordEnum);
hasKeywords = length3 > 0;
length += length3;
if ((hasScript && !hasCountry)
|| ((hasScript || hasCountry) && !hasVariant && !hasKeywords)
|| ((hasScript || hasCountry || hasVariant) && !hasKeywords))
{
/* Remove separator */
length -= locSepLen;
}
else if (hasLanguage && !hasScript && !hasCountry && !hasVariant && !hasKeywords)
{
/* Remove " (" */
length -= 2;
}
if (hasLanguage && (hasScript || hasCountry || hasVariant || hasKeywords))
{
/* append ")" */
if (length < destCapacity)
{
dest[length] = 0x29;
}
++length;
/* If the localized display pattern is something other than the default pattern of "{0} ({1})", then
* then we need to do the formatting here. It would be easier to use a messageFormat to do this, but we
* can't since we don't have the APIs in the i18n library available to us at this point.
*/
if (locPatLen != defaultPatternLen ||
u_strcmp(dispLocPattern, defaultPattern)) /* Something other than the default pattern */
{
UChar * p0 = u_strstr(dispLocPattern, pat0);
UChar * p1 = u_strstr(dispLocPattern, pat1);
u_terminateUChars(dest, destCapacity, length, pErrorCode);
if (p0 != NULL && p1 != NULL) /* The pattern is well formed */
{
if (dest)
{
int32_t destLen = 0;
UChar * result = (UChar *)uprv_malloc((length + 1) * sizeof(UChar));
UChar * upos = (UChar *)dispLocPattern;
u_strcpy(result, dest);
dest[0] = 0;
while (*upos)
{
if (upos == p0) /* Handle {0} substitution */
{
u_strncat(dest, result, p0Len);
destLen += p0Len;
dest[destLen] = 0; /* Null terminate */
upos += 3;
}
else if (upos == p1) /* Handle {1} substitution */
{
UChar * p1Start = &result[p0Len + 2];
u_strncat(dest, p1Start, length - p0Len - 3);
destLen += (length - p0Len - 3);
dest[destLen] = 0; /* Null terminate */
upos += 3;
}
else /* Something from the pattern not {0} or {1} */
{
u_strncat(dest, upos, 1);
upos++;
destLen++;
dest[destLen] = 0; /* Null terminate */
}
}
length = destLen;
uprv_free(result);
}
}
}
}
if (*pErrorCode == U_BUFFER_OVERFLOW_ERROR)
{
/* keep preflighting */
*pErrorCode = U_ZERO_ERROR;
}
return u_terminateUChars(dest, destCapacity, length, pErrorCode);
}
U_CFUNC int32_t
idnaref_IDNToUnicode( const UChar* src, int32_t srcLength,
UChar* dest, int32_t destCapacity,
int32_t options,
UParseError* parseError,
UErrorCode* status){
if(status == NULL || U_FAILURE(*status)){
return 0;
}
if((src == NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
*status = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
int32_t reqLength = 0;
UBool done = FALSE;
NamePrepTransform* prep = TestIDNA::getInstance(*status);
//initialize pointers to stack buffers
UChar b1Stack[MAX_LABEL_BUFFER_SIZE];
UChar *b1 = b1Stack;
int32_t b1Len, labelLen;
UChar* delimiter = (UChar*)src;
UChar* labelStart = (UChar*)src;
int32_t remainingLen = srcLength;
int32_t b1Capacity = MAX_LABEL_BUFFER_SIZE;
//get the options
// UBool allowUnassigned = (UBool)((options & IDNAREF_ALLOW_UNASSIGNED) != 0);
// UBool useSTD3ASCIIRules = (UBool)((options & IDNAREF_USE_STD3_RULES) != 0);
if(U_FAILURE(*status)){
goto CLEANUP;
}
if(srcLength == -1){
for(;;){
if(*delimiter == 0){
break;
}
labelLen = getNextSeparator(labelStart, -1, prep, &delimiter, &done, status);
b1Len = idnaref_toUnicode(labelStart, labelLen, b1, b1Capacity,
options, parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b1Len = idnaref_toUnicode( labelStart, labelLen, b1, b1Len,
options, parseError, status);
}
if(U_FAILURE(*status)){
goto CLEANUP;
}
int32_t tempLen = (reqLength + b1Len );
// copy to dest
if( tempLen< destCapacity){
uprv_memmove(dest+reqLength, b1, b1Len * U_SIZEOF_UCHAR);
}
reqLength = tempLen;
// add the label separator
if(done == FALSE){
if(reqLength < destCapacity){
dest[reqLength] = FULL_STOP;
}
reqLength++;
}
labelStart = delimiter;
}
}else{
for(;;){
if(delimiter == src+srcLength){
break;
}
labelLen = getNextSeparator(labelStart, remainingLen, prep, &delimiter, &done, status);
b1Len = idnaref_toUnicode( labelStart,labelLen, b1, b1Capacity,
options, parseError, status);
if(*status == U_BUFFER_OVERFLOW_ERROR){
// redo processing of string
/* we do not have enough room so grow the buffer*/
b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
if(b1==NULL){
*status = U_MEMORY_ALLOCATION_ERROR;
goto CLEANUP;
}
*status = U_ZERO_ERROR; // reset error
b1Len = idnaref_toUnicode( labelStart, labelLen, b1, b1Len,
options, parseError, status);
}
if(U_FAILURE(*status)){
goto CLEANUP;
}
int32_t tempLen = (reqLength + b1Len );
// copy to dest
if( tempLen< destCapacity){
uprv_memmove(dest+reqLength, b1, b1Len * U_SIZEOF_UCHAR);
}
reqLength = tempLen;
// add the label separator
if(done == FALSE){
if(reqLength < destCapacity){
dest[reqLength] = FULL_STOP;
}
reqLength++;
}
labelStart = delimiter;
remainingLen = srcLength - (delimiter - src);
}
}
CLEANUP:
if(b1 != b1Stack){
uprv_free(b1);
}
// delete prep;
return u_terminateUChars(dest, destCapacity, reqLength, status);
}
U_CFUNC int32_t
u_strToPunycode(const UChar *src, int32_t srcLength,
UChar *dest, int32_t destCapacity,
const UBool *caseFlags,
UErrorCode *pErrorCode) {
int32_t cpBuffer[MAX_CP_COUNT];
int32_t n, delta, handledCPCount, basicLength, destLength, bias, j, m, q, k, t, srcCPCount;
UChar c, c2;
/* argument checking */
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return 0;
}
if(src==NULL || srcLength<-1 || (dest==NULL && destCapacity!=0)) {
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
/*
* Handle the basic code points and
* convert extended ones to UTF-32 in cpBuffer (caseFlag in sign bit):
*/
srcCPCount=destLength=0;
if(srcLength==-1) {
/* NUL-terminated input */
for(j=0; /* no condition */; ++j) {
if((c=src[j])==0) {
break;
}
if(srcCPCount==MAX_CP_COUNT) {
/* too many input code points */
*pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
if(IS_BASIC(c)) {
cpBuffer[srcCPCount++]=0;
if(destLength<destCapacity) {
dest[destLength]=
caseFlags!=NULL ?
asciiCaseMap((char)c, caseFlags[j]) :
(char)c;
}
++destLength;
} else {
n=(caseFlags!=NULL && caseFlags[j])<<31L;
if(U16_IS_SINGLE(c)) {
n|=c;
} else if(U16_IS_LEAD(c) && U16_IS_TRAIL(c2=src[j+1])) {
++j;
n|=(int32_t)U16_GET_SUPPLEMENTARY(c, c2);
} else {
/* error: unmatched surrogate */
*pErrorCode=U_INVALID_CHAR_FOUND;
return 0;
}
cpBuffer[srcCPCount++]=n;
}
}
} else {
/* length-specified input */
for(j=0; j<srcLength; ++j) {
if(srcCPCount==MAX_CP_COUNT) {
/* too many input code points */
*pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
c=src[j];
if(IS_BASIC(c)) {
cpBuffer[srcCPCount++]=0;
if(destLength<destCapacity) {
dest[destLength]=
caseFlags!=NULL ?
asciiCaseMap((char)c, caseFlags[j]) :
(char)c;
}
++destLength;
} else {
n=(caseFlags!=NULL && caseFlags[j])<<31L;
if(U16_IS_SINGLE(c)) {
n|=c;
} else if(U16_IS_LEAD(c) && (j+1)<srcLength && U16_IS_TRAIL(c2=src[j+1])) {
++j;
n|=(int32_t)U16_GET_SUPPLEMENTARY(c, c2);
} else {
/* error: unmatched surrogate */
*pErrorCode=U_INVALID_CHAR_FOUND;
return 0;
}
cpBuffer[srcCPCount++]=n;
}
}
}
/* Finish the basic string - if it is not empty - with a delimiter. */
basicLength=destLength;
if(basicLength>0) {
if(destLength<destCapacity) {
dest[destLength]=DELIMITER;
}
++destLength;
}
/*
* handledCPCount is the number of code points that have been handled
* basicLength is the number of basic code points
* destLength is the number of chars that have been output
*/
/* Initialize the state: */
n=INITIAL_N;
delta=0;
bias=INITIAL_BIAS;
/* Main encoding loop: */
for(handledCPCount=basicLength; handledCPCount<srcCPCount; /* no op */) {
/*
* All non-basic code points < n have been handled already.
* Find the next larger one:
*/
for(m=0x7fffffff, j=0; j<srcCPCount; ++j) {
q=cpBuffer[j]&0x7fffffff; /* remove case flag from the sign bit */
if(n<=q && q<m) {
m=q;
}
}
/*
* Increase delta enough to advance the decoder's
* <n,i> state to <m,0>, but guard against overflow:
*/
if(m-n>(0x7fffffff-MAX_CP_COUNT-delta)/(handledCPCount+1)) {
*pErrorCode=U_INTERNAL_PROGRAM_ERROR;
return 0;
}
delta+=(m-n)*(handledCPCount+1);
n=m;
/* Encode a sequence of same code points n */
for(j=0; j<srcCPCount; ++j) {
q=cpBuffer[j]&0x7fffffff; /* remove case flag from the sign bit */
if(q<n) {
++delta;
} else if(q==n) {
/* Represent delta as a generalized variable-length integer: */
for(q=delta, k=BASE; /* no condition */; k+=BASE) {
/** RAM: comment out the old code for conformance with draft-ietf-idn-punycode-03.txt
t=k-bias;
if(t<TMIN) {
t=TMIN;
} else if(t>TMAX) {
t=TMAX;
}
*/
t=k-bias;
if(t<TMIN) {
t=TMIN;
} else if(k>=(bias+TMAX)) {
t=TMAX;
}
if(q<t) {
break;
}
if(destLength<destCapacity) {
dest[destLength]=digitToBasic(t+(q-t)%(BASE-t), 0);
}
++destLength;
q=(q-t)/(BASE-t);
}
if(destLength<destCapacity) {
dest[destLength]=digitToBasic(q, (UBool)(cpBuffer[j]<0));
}
++destLength;
bias=adaptBias(delta, handledCPCount+1, (UBool)(handledCPCount==basicLength));
delta=0;
++handledCPCount;
}
}
++delta;
++n;
}
return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
}
