/*
* BNF in RFC5646
*
* Language-Tag = langtag ; normal language tags
* / privateuse ; private use tag
* / grandfathered ; grandfathered tags
*
*
* langtag = language
* ["-" script]
* ["-" region]
* *("-" variant)
* *("-" extension)
* ["-" privateuse]
*
* language = 2*3ALPHA ; shortest ISO 639 code
* ["-" extlang] ; sometimes followed by
* ; extended language subtags
* / 4ALPHA ; or reserved for future use
* / 5*8ALPHA ; or registered language subtag
*
* extlang = 3ALPHA ; selected ISO 639 codes
* *2("-" 3ALPHA) ; permanently reserved
*
* script = 4ALPHA ; ISO 15924 code
*
* region = 2ALPHA ; ISO 3166-1 code
* / 3DIGIT ; UN M.49 code
*
* variant = 5*8alphanum ; registered variants
* / (DIGIT 3alphanum)
*
* extension = singleton 1*("-" (2*8alphanum))
*
* ; Single alphanumerics
* ; "x" reserved for private use
* singleton = DIGIT ; 0 - 9
* / %x41-57 ; A - W
* / %x59-5A ; Y - Z
* / %x61-77 ; a - w
* / %x79-7A ; y - z
*
* privateuse = "x" 1*("-" (1*8alphanum))
*
*/
public static LanguageTag parse(String languageTag, ParseStatus sts) {
if (sts == null) {
sts = new ParseStatus();
} else {
sts.reset();
}
StringTokenIterator itr;
// Check if the tag is grandfathered
String[] gfmap = GRANDFATHERED.get(LocaleUtils.toLowerString(languageTag));
if (gfmap != null) {
// use preferred mapping
itr = new StringTokenIterator(gfmap[1], SEP);
} else {
itr = new StringTokenIterator(languageTag, SEP);
}
LanguageTag tag = new LanguageTag();
// langtag must start with either language or privateuse
if (tag.parseLanguage(itr, sts)) {
tag.parseExtlangs(itr, sts);
tag.parseScript(itr, sts);
tag.parseRegion(itr, sts);
tag.parseVariants(itr, sts);
tag.parseExtensions(itr, sts);
}
tag.parsePrivateuse(itr, sts);
if (!itr.isDone() && !sts.isError()) {
String s = itr.current();
sts.errorIndex = itr.currentStart();
if (s.length() == 0) {
sts.errorMsg = "Empty subtag";
} else {
sts.errorMsg = "Invalid subtag: " + s;
}
}
return tag;
}
public static LanguageTag parseLocale(BaseLocale baseLocale, LocaleExtensions localeExtensions) {
LanguageTag tag = new LanguageTag();
String language = baseLocale.getLanguage();
String script = baseLocale.getScript();
String region = baseLocale.getRegion();
String variant = baseLocale.getVariant();
boolean hasSubtag = false;
String privuseVar = null; // store ill-formed variant subtags
if (isLanguage(language)) {
// Convert a deprecated language code to its new code
if (language.equals("iw")) {
language = "he";
} else if (language.equals("ji")) {
language = "yi";
} else if (language.equals("in")) {
language = "id";
}
tag.language = language;
}
if (isScript(script)) {
tag.script = canonicalizeScript(script);
hasSubtag = true;
}
if (isRegion(region)) {
tag.region = canonicalizeRegion(region);
hasSubtag = true;
}
// Special handling for no_NO_NY - use nn_NO for language tag
if (tag.language.equals("no") && tag.region.equals("NO") && variant.equals("NY")) {
tag.language = "nn";
variant = "";
}
if (variant.length() > 0) {
List<String> variants = null;
StringTokenIterator varitr = new StringTokenIterator(variant, BaseLocale.SEP);
while (!varitr.isDone()) {
String var = varitr.current();
if (!isVariant(var)) {
break;
}
if (variants == null) {
variants = new ArrayList<>();
}
variants.add(var); // Do not canonicalize!
varitr.next();
}
if (variants != null) {
tag.variants = variants;
hasSubtag = true;
}
if (!varitr.isDone()) {
// ill-formed variant subtags
StringJoiner sj = new StringJoiner(SEP);
while (!varitr.isDone()) {
String prvv = varitr.current();
if (!isPrivateuseSubtag(prvv)) {
// cannot use private use subtag - truncated
break;
}
sj.add(prvv);
varitr.next();
}
if (sj.length() > 0) {
privuseVar = sj.toString();
}
}
}
List<String> extensions = null;
String privateuse = null;
if (localeExtensions != null) {
Set<Character> locextKeys = localeExtensions.getKeys();
for (Character locextKey : locextKeys) {
Extension ext = localeExtensions.getExtension(locextKey);
if (isPrivateusePrefixChar(locextKey)) {
privateuse = ext.getValue();
} else {
if (extensions == null) {
extensions = new ArrayList<>();
}
extensions.add(locextKey.toString() + SEP + ext.getValue());
}
}
}
if (extensions != null) {
tag.extensions = extensions;
hasSubtag = true;
}
// append ill-formed variant subtags to private use
if (privuseVar != null) {
if (privateuse == null) {
privateuse = PRIVUSE_VARIANT_PREFIX + SEP + privuseVar;
} else {
privateuse =
privateuse
+ SEP
+ PRIVUSE_VARIANT_PREFIX
+ SEP
+ privuseVar.replace(BaseLocale.SEP, SEP);
}
}
if (privateuse != null) {
tag.privateuse = privateuse;
}
if (tag.language.length() == 0 && (hasSubtag || privateuse == null)) {
// use lang "und" when 1) no language is available AND
// 2) any of other subtags other than private use are available or
// no private use tag is available
tag.language = UNDETERMINED;
}
return tag;
}