U_CAPI void U_EXPORT2
ucasemap_setLocale(UCaseMap *csm, const char *locale, UErrorCode *pErrorCode) {
if(U_FAILURE(*pErrorCode)) {
return;
}
if (locale != NULL && *locale == 0) {
csm->locale[0] = 0;
csm->caseLocale = UCASE_LOC_ROOT;
return;
}
int32_t length=uloc_getName(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode);
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR || length==sizeof(csm->locale)) {
*pErrorCode=U_ZERO_ERROR;
/* we only really need the language code for case mappings */
length=uloc_getLanguage(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode);
}
if(length==sizeof(csm->locale)) {
*pErrorCode=U_BUFFER_OVERFLOW_ERROR;
}
if(U_SUCCESS(*pErrorCode)) {
csm->caseLocale=UCASE_LOC_UNKNOWN;
csm->caseLocale = ucase_getCaseLocale(csm->locale);
} else {
csm->locale[0]=0;
csm->caseLocale = UCASE_LOC_ROOT;
}
}
/*
Function:
GetLocaleIso639LanguageTwoLetterName
Gets the language name for a locale (via uloc_getLanguage) and converts the result to UChars
*/
static UErrorCode GetLocaleIso639LanguageTwoLetterName(const char* locale, UChar* value, int32_t valueLength)
{
UErrorCode status = U_ZERO_ERROR, ignore = U_ZERO_ERROR;
int32_t length = uloc_getLanguage(locale, NULL, 0, &ignore) + 1;
char* buf = calloc(length, sizeof(char));
if (buf == NULL)
{
return U_MEMORY_ALLOCATION_ERROR;
}
uloc_getLanguage(locale, buf, length, &status);
u_charsToUChars_safe(buf, value, valueLength, &status);
free(buf);
return status;
}
/*
Function:
GetLocaleIso639LanguageName
Gets the language name for a locale (via uloc_getLanguage) and converts the result to UChars
*/
UErrorCode GetLocaleIso639LanguageName(const char* locale, UChar* value, int32_t valueLength)
{
UErrorCode status = U_ZERO_ERROR;
int32_t length = uloc_getLanguage(locale, nullptr, 0, &status);
std::vector<char> buf(length + 1, '\0');
status = U_ZERO_ERROR;
uloc_getLanguage(locale, buf.data(), length + 1, &status);
if (U_SUCCESS(status))
{
status = u_charsToUChars_safe(buf.data(), value, valueLength);
}
return status;
}
int32_t GetLocale(
const UChar* localeName, char* localeNameResult, int32_t localeNameResultLength, bool canonicalize, UErrorCode* err)
{
char localeNameTemp[ULOC_FULLNAME_CAPACITY] = {0};
int32_t localeLength;
// Convert ourselves instead of doing u_UCharsToChars as that function considers '@' a variant and stops.
for (int i = 0; i < ULOC_FULLNAME_CAPACITY - 1; i++)
{
UChar c = localeName[i];
if (c > (UChar)0x7F)
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
return ULOC_FULLNAME_CAPACITY;
}
localeNameTemp[i] = (char)c;
if (c == (UChar)0x0)
{
break;
}
}
if (canonicalize)
{
localeLength = uloc_canonicalize(localeNameTemp, localeNameResult, localeNameResultLength, err);
}
else
{
localeLength = uloc_getName(localeNameTemp, localeNameResult, localeNameResultLength, err);
}
if (U_SUCCESS(*err))
{
// Make sure the "language" part of the locale is reasonable (i.e. we can fetch it and it is within range).
// This mimics how the C++ ICU API determines if a locale is "bogus" or not.
char language[ULOC_LANG_CAPACITY];
uloc_getLanguage(localeNameTemp, language, ULOC_LANG_CAPACITY, err);
if (*err == U_STRING_NOT_TERMINATED_WARNING)
{
// ULOC_LANG_CAPACITY includes the null terminator, so if we couldn't extract the language with the null
// terminator, the language must be invalid.
*err = U_ILLEGAL_ARGUMENT_ERROR;
}
}
return localeLength;
}
// Implemented here because this calls uloc_addLikelySubtags().
U_CAPI UBool U_EXPORT2
uloc_isRightToLeft(const char *locale) {
UErrorCode errorCode = U_ZERO_ERROR;
char script[8];
int32_t scriptLength = uloc_getScript(locale, script, UPRV_LENGTHOF(script), &errorCode);
if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING ||
scriptLength == 0) {
// Fastpath: We know the likely scripts and their writing direction
// for some common languages.
errorCode = U_ZERO_ERROR;
char lang[8];
int32_t langLength = uloc_getLanguage(locale, lang, UPRV_LENGTHOF(lang), &errorCode);
if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING ||
langLength == 0) {
return FALSE;
}
const char* langPtr = uprv_strstr(LANG_DIR_STRING, lang);
if (langPtr != NULL) {
switch (langPtr[langLength]) {
case '-': return FALSE;
case '+': return TRUE;
default: break; // partial match of a longer code
}
}
// Otherwise, find the likely script.
errorCode = U_ZERO_ERROR;
char likely[ULOC_FULLNAME_CAPACITY];
(void)uloc_addLikelySubtags(locale, likely, UPRV_LENGTHOF(likely), &errorCode);
if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING) {
return FALSE;
}
scriptLength = uloc_getScript(likely, script, UPRV_LENGTHOF(script), &errorCode);
if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING ||
scriptLength == 0) {
return FALSE;
}
}
UScriptCode scriptCode = (UScriptCode)u_getPropertyValueEnum(UCHAR_SCRIPT, script);
return uscript_isRightToLeft(scriptCode);
}
U_CAPI void U_EXPORT2
ucasemap_setLocale(UCaseMap *csm, const char *locale, UErrorCode *pErrorCode) {
int32_t length;
if(U_FAILURE(*pErrorCode)) {
return;
}
length=uloc_getName(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode);
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR || length==sizeof(csm->locale)) {
*pErrorCode=U_ZERO_ERROR;
/* we only really need the language code for case mappings */
length=uloc_getLanguage(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode);
}
if(length==sizeof(csm->locale)) {
*pErrorCode=U_BUFFER_OVERFLOW_ERROR;
}
csm->locCache=0;
if(U_SUCCESS(*pErrorCode)) {
ucase_getCaseLocale(csm->locale, &csm->locCache);
} else {
csm->locale[0]=0;
}
}
/* {{{
* Gets the value from ICU
* common code shared by get_primary_language,get_script or get_region or get_variant
* result = 0 if error, 1 if successful , -1 if no value
*/
static char* get_icu_value_internal( char* loc_name , char* tag_name, int* result , int fromParseLocale)
{
char* tag_value = NULL;
int32_t tag_value_len = 512;
int singletonPos = 0;
char* mod_loc_name = NULL;
int grOffset = 0;
int32_t buflen = 512;
UErrorCode status = U_ZERO_ERROR;
if( tag_name != LOC_CANONICALIZE_TAG ){
/* Handle grandfathered languages */
grOffset = findOffset( LOC_GRANDFATHERED , loc_name );
if( grOffset >= 0 ){
if( strcmp(tag_name , LOC_LANG_TAG)==0 ){
tag_value = estrdup(loc_name);
return tag_value;
} else {
/* Since Grandfathered , no value , do nothing , retutn NULL */
return NULL;
}
}
if( fromParseLocale==1 ){
/* Handle singletons */
if( strcmp(tag_name , LOC_LANG_TAG)==0 ){
if( strlen(loc_name)>1 && (isIDPrefix(loc_name) ==1 ) ){
return loc_name;
}
}
singletonPos = getSingletonPos( loc_name );
if( singletonPos == 0){
/* singleton at start of script, region , variant etc.
* or invalid singleton at start of language */
return NULL;
} else if( singletonPos > 0 ){
/* singleton at some position except at start
* strip off the singleton and rest of the loc_name */
mod_loc_name = estrndup ( loc_name , singletonPos-1);
}
} /* end of if fromParse */
} /* end of if != LOC_CANONICAL_TAG */
if( mod_loc_name == NULL){
mod_loc_name = estrdup(loc_name );
}
/* Proceed to ICU */
do{
tag_value = erealloc( tag_value , buflen );
tag_value_len = buflen;
if( strcmp(tag_name , LOC_SCRIPT_TAG)==0 ){
buflen = uloc_getScript ( mod_loc_name ,tag_value , tag_value_len , &status);
}
if( strcmp(tag_name , LOC_LANG_TAG )==0 ){
buflen = uloc_getLanguage ( mod_loc_name ,tag_value , tag_value_len , &status);
}
if( strcmp(tag_name , LOC_REGION_TAG)==0 ){
buflen = uloc_getCountry ( mod_loc_name ,tag_value , tag_value_len , &status);
}
if( strcmp(tag_name , LOC_VARIANT_TAG)==0 ){
buflen = uloc_getVariant ( mod_loc_name ,tag_value , tag_value_len , &status);
}
if( strcmp(tag_name , LOC_CANONICALIZE_TAG)==0 ){
buflen = uloc_canonicalize ( mod_loc_name ,tag_value , tag_value_len , &status);
}
if( U_FAILURE( status ) ) {
if( status == U_BUFFER_OVERFLOW_ERROR ) {
status = U_ZERO_ERROR;
continue;
}
/* Error in retriving data */
*result = 0;
if( tag_value ){
efree( tag_value );
}
if( mod_loc_name ){
efree( mod_loc_name);
}
return NULL;
}
} while( buflen > tag_value_len );
if( buflen ==0 ){
/* No value found */
*result = -1;
if( tag_value ){
efree( tag_value );
}
if( mod_loc_name ){
efree( mod_loc_name);
}
return NULL;
} else {
*result = 1;
}
if( mod_loc_name ){
efree( mod_loc_name);
}
return tag_value;
}
/**
* Create a tag string from the supplied parameters. The lang, script and region
* parameters may be NULL pointers. If they are, their corresponding length parameters
* must be less than or equal to 0.
*
* If any of the language, script or region parameters are empty, and the alternateTags
* parameter is not NULL, it will be parsed for potential language, script and region tags
* to be used when constructing the new tag. If the alternateTags parameter is NULL, or
* it contains no language tag, the default tag for the unknown language is used.
*
* If the length of the new string exceeds the capacity of the output buffer,
* the function copies as many bytes to the output buffer as it can, and returns
* the error U_BUFFER_OVERFLOW_ERROR.
*
* If an illegal argument is provided, the function returns the error
* U_ILLEGAL_ARGUMENT_ERROR.
*
* Note that this function can return the warning U_STRING_NOT_TERMINATED_WARNING if
* the tag string fits in the output buffer, but the null terminator doesn't.
*
* @param lang The language tag to use.
* @param langLength The length of the language tag.
* @param script The script tag to use.
* @param scriptLength The length of the script tag.
* @param region The region tag to use.
* @param regionLength The length of the region tag.
* @param trailing Any trailing data to append to the new tag.
* @param trailingLength The length of the trailing data.
* @param alternateTags A string containing any alternate tags.
* @param tag The output buffer.
* @param tagCapacity The capacity of the output buffer.
* @param err A pointer to a UErrorCode for error reporting.
* @return The length of the tag string, which may be greater than tagCapacity, or -1 on error.
**/
static int32_t U_CALLCONV
createTagStringWithAlternates(
const char* lang,
int32_t langLength,
const char* script,
int32_t scriptLength,
const char* region,
int32_t regionLength,
const char* trailing,
int32_t trailingLength,
const char* alternateTags,
char* tag,
int32_t tagCapacity,
UErrorCode* err) {
if (U_FAILURE(*err)) {
goto error;
}
else if (tag == NULL ||
tagCapacity <= 0 ||
langLength >= ULOC_LANG_CAPACITY ||
scriptLength >= ULOC_SCRIPT_CAPACITY ||
regionLength >= ULOC_COUNTRY_CAPACITY) {
goto error;
}
else {
/**
* ULOC_FULLNAME_CAPACITY will provide enough capacity
* that we can build a string that contains the language,
* script and region code without worrying about overrunning
* the user-supplied buffer.
**/
char tagBuffer[ULOC_FULLNAME_CAPACITY];
int32_t tagLength = 0;
int32_t capacityRemaining = tagCapacity;
UBool regionAppended = FALSE;
if (langLength > 0) {
appendTag(
lang,
langLength,
tagBuffer,
&tagLength);
}
else if (alternateTags == NULL) {
/*
* Append the value for an unknown language, if
* we found no language.
*/
appendTag(
unknownLanguage,
(int32_t)uprv_strlen(unknownLanguage),
tagBuffer,
&tagLength);
}
else {
/*
* Parse the alternateTags string for the language.
*/
char alternateLang[ULOC_LANG_CAPACITY];
int32_t alternateLangLength = sizeof(alternateLang);
alternateLangLength =
uloc_getLanguage(
alternateTags,
alternateLang,
alternateLangLength,
err);
if(U_FAILURE(*err) ||
alternateLangLength >= ULOC_LANG_CAPACITY) {
goto error;
}
else if (alternateLangLength == 0) {
/*
* Append the value for an unknown language, if
* we found no language.
*/
appendTag(
unknownLanguage,
(int32_t)uprv_strlen(unknownLanguage),
tagBuffer,
&tagLength);
}
else {
appendTag(
alternateLang,
alternateLangLength,
tagBuffer,
&tagLength);
}
}
if (scriptLength > 0) {
appendTag(
script,
scriptLength,
tagBuffer,
&tagLength);
}
else if (alternateTags != NULL) {
/*
* Parse the alternateTags string for the script.
*/
char alternateScript[ULOC_SCRIPT_CAPACITY];
const int32_t alternateScriptLength =
uloc_getScript(
alternateTags,
alternateScript,
sizeof(alternateScript),
err);
if (U_FAILURE(*err) ||
alternateScriptLength >= ULOC_SCRIPT_CAPACITY) {
goto error;
}
else if (alternateScriptLength > 0) {
appendTag(
alternateScript,
alternateScriptLength,
tagBuffer,
&tagLength);
}
}
if (regionLength > 0) {
appendTag(
region,
regionLength,
tagBuffer,
&tagLength);
regionAppended = TRUE;
}
else if (alternateTags != NULL) {
/*
* Parse the alternateTags string for the region.
*/
char alternateRegion[ULOC_COUNTRY_CAPACITY];
const int32_t alternateRegionLength =
uloc_getCountry(
alternateTags,
alternateRegion,
sizeof(alternateRegion),
err);
if (U_FAILURE(*err) ||
alternateRegionLength >= ULOC_COUNTRY_CAPACITY) {
goto error;
}
else if (alternateRegionLength > 0) {
appendTag(
alternateRegion,
alternateRegionLength,
tagBuffer,
&tagLength);
regionAppended = TRUE;
}
}
{
const int32_t toCopy =
tagLength >= tagCapacity ? tagCapacity : tagLength;
/**
* Copy the partial tag from our internal buffer to the supplied
* target.
**/
uprv_memcpy(
tag,
tagBuffer,
toCopy);
capacityRemaining -= toCopy;
}
if (trailingLength > 0) {
if (*trailing != '@' && capacityRemaining > 0) {
tag[tagLength++] = '_';
--capacityRemaining;
if (capacityRemaining > 0 && !regionAppended) {
/* extra separator is required */
tag[tagLength++] = '_';
--capacityRemaining;
}
}
if (capacityRemaining > 0) {
/*
* Copy the trailing data into the supplied buffer. Use uprv_memmove, since we
* don't know if the user-supplied buffers overlap.
*/
const int32_t toCopy =
trailingLength >= capacityRemaining ? capacityRemaining : trailingLength;
uprv_memmove(
&tag[tagLength],
trailing,
toCopy);
}
}
tagLength += trailingLength;
return u_terminateChars(
tag,
tagCapacity,
tagLength,
err);
}
error:
/**
* An overflow indicates the locale ID passed in
* is ill-formed. If we got here, and there was
* no previous error, it's an implicit overflow.
**/
if (*err == U_BUFFER_OVERFLOW_ERROR ||
U_SUCCESS(*err)) {
*err = U_ILLEGAL_ARGUMENT_ERROR;
}
return -1;
}
U_CAPI int32_t U_EXPORT2
ucol_getShortDefinitionString(const UCollator *coll,
const char *locale,
char *dst,
int32_t capacity,
UErrorCode *status)
{
if(U_FAILURE(*status)) return 0;
if(coll->delegate != NULL) {
return ((icu::Collator*)coll->delegate)->internalGetShortDefinitionString(locale,dst,capacity,*status);
}
char buffer[internalBufferSize];
uprv_memset(buffer, 0, internalBufferSize*sizeof(char));
int32_t resultSize = 0;
char tempbuff[internalBufferSize];
char locBuff[internalBufferSize];
uprv_memset(buffer, 0, internalBufferSize*sizeof(char));
int32_t elementSize = 0;
UBool isAvailable = 0;
CollatorSpec s;
ucol_sit_initCollatorSpecs(&s);
if(!locale) {
locale = ucol_getLocaleByType(coll, ULOC_VALID_LOCALE, status);
}
elementSize = ucol_getFunctionalEquivalent(locBuff, internalBufferSize, "collation", locale, &isAvailable, status);
if(elementSize) {
// we should probably canonicalize here...
elementSize = uloc_getLanguage(locBuff, tempbuff, internalBufferSize, status);
appendShortStringElement(tempbuff, elementSize, buffer, &resultSize, /*capacity*/internalBufferSize, languageArg);
elementSize = uloc_getCountry(locBuff, tempbuff, internalBufferSize, status);
appendShortStringElement(tempbuff, elementSize, buffer, &resultSize, /*capacity*/internalBufferSize, regionArg);
elementSize = uloc_getScript(locBuff, tempbuff, internalBufferSize, status);
appendShortStringElement(tempbuff, elementSize, buffer, &resultSize, /*capacity*/internalBufferSize, scriptArg);
elementSize = uloc_getVariant(locBuff, tempbuff, internalBufferSize, status);
appendShortStringElement(tempbuff, elementSize, buffer, &resultSize, /*capacity*/internalBufferSize, variantArg);
elementSize = uloc_getKeywordValue(locBuff, "collation", tempbuff, internalBufferSize, status);
appendShortStringElement(tempbuff, elementSize, buffer, &resultSize, /*capacity*/internalBufferSize, keywordArg);
}
int32_t i = 0;
UColAttributeValue attribute = UCOL_DEFAULT;
for(i = 0; i < UCOL_SIT_ITEMS_COUNT; i++) {
if(options[i].action == _processCollatorOption) {
attribute = ucol_getAttributeOrDefault(coll, (UColAttribute)options[i].attr, status);
if(attribute != UCOL_DEFAULT) {
char letter = ucol_sit_attributeValueToLetter(attribute, status);
appendShortStringElement(&letter, 1,
buffer, &resultSize, /*capacity*/internalBufferSize, options[i].optionStart);
}
}
}
if(coll->variableTopValueisDefault == FALSE) {
//s.variableTopValue = ucol_getVariableTop(coll, status);
elementSize = T_CString_integerToString(tempbuff, coll->variableTopValue, 16);
appendShortStringElement(tempbuff, elementSize, buffer, &resultSize, capacity, variableTopValArg);
}
UParseError parseError;
return ucol_normalizeShortDefinitionString(buffer, dst, capacity, &parseError, status);
}
