/**
* Gets the character extended type
*
* @param ch character to be tested
* @return extended type it is associated with
*/
private static int getType(int ch) {
if (UCharacterUtility.isNonCharacter(ch)) {
// not a character we return a invalid category count
return NON_CHARACTER_;
}
int result = UCharacter.getType(ch);
if (result == UCharacterCategory.SURROGATE) {
if (ch <= UTF16.LEAD_SURROGATE_MAX_VALUE) {
result = LEAD_SURROGATE_;
} else {
result = TRAIL_SURROGATE_;
}
}
return result;
}
public static CasingType from(String s) {
if (s == null || s.length() == 0) {
return other;
}
int cp;
// Look for the first meaningful character in the string to determine case.
for (int i = 0; i < s.length(); i += Character.charCount(cp)) {
cp = s.codePointAt(i);
// used to skip the placeholders, but works better to have them be 'other'
// if (cp == '{') {
// if (placeholder.reset(s).region(i,s.length()).lookingAt()) {
// i = placeholder.end() - 1; // skip
// continue;
// }
// }
int type = UCharacter.getType(cp);
switch (type) {
case UCharacter.LOWERCASE_LETTER:
return lowercase;
case UCharacter.UPPERCASE_LETTER:
case UCharacter.TITLECASE_LETTER:
return titlecase;
// for other letters / numbers / symbols, return other
case UCharacter.OTHER_LETTER:
case UCharacter.DECIMAL_DIGIT_NUMBER:
case UCharacter.LETTER_NUMBER:
case UCharacter.OTHER_NUMBER:
case UCharacter.MATH_SYMBOL:
case UCharacter.CURRENCY_SYMBOL:
case UCharacter.MODIFIER_SYMBOL:
case UCharacter.OTHER_SYMBOL:
return other;
// ignore everything else (whitespace, punctuation, etc) and keep going
}
}
return other;
}
/**
* Set an identifier to analyze. Afterwards, call methods like getScripts()
*
* @param identifier the identifier to analyze
* @return self
* @internal
* @deprecated This API is ICU internal only.
*/
@Deprecated
public IdentifierInfo setIdentifier(String identifier) {
this.identifier = identifier;
clear();
BitSet scriptsForCP = new BitSet();
int cp;
for (int i = 0; i < identifier.length(); i += Character.charCount(i)) {
cp = Character.codePointAt(identifier, i);
// Store a representative character for each kind of decimal digit
if (UCharacter.getType(cp) == UCharacterCategory.DECIMAL_DIGIT_NUMBER) {
// Just store the zero character as a representative for comparison. Unicode guarantees it
// is cp - value
numerics.add(cp - UCharacter.getNumericValue(cp));
}
UScript.getScriptExtensions(cp, scriptsForCP);
scriptsForCP.clear(UScript.COMMON);
scriptsForCP.clear(UScript.INHERITED);
// if (temp.cardinality() == 0) {
// // HACK for older version of ICU
// requiredScripts.set(UScript.getScript(cp));
// } else
switch (scriptsForCP.cardinality()) {
case 0:
break;
case 1:
// Single script, record it.
requiredScripts.or(scriptsForCP);
break;
default:
if (!requiredScripts.intersects(scriptsForCP) && scriptSetSet.add(scriptsForCP)) {
scriptsForCP = new BitSet();
}
break;
}
}
// Now make a final pass through to remove alternates that came before singles.
// [Kana], [Kana Hira] => [Kana]
// This is relatively infrequent, so doesn't have to be optimized.
// We also compute any commonalities among the alternates.
if (scriptSetSet.size() > 0) {
commonAmongAlternates.set(0, UScript.CODE_LIMIT);
for (Iterator<BitSet> it = scriptSetSet.iterator(); it.hasNext(); ) {
final BitSet next = it.next();
// [Kana], [Kana Hira] => [Kana]
if (requiredScripts.intersects(next)) {
it.remove();
} else {
// [[Arab Syrc Thaa]; [Arab Syrc]] => [[Arab Syrc]]
commonAmongAlternates.and(next); // get the intersection.
for (BitSet other : scriptSetSet) {
if (next != other && contains(next, other)) {
it.remove();
break;
}
}
}
}
}
if (scriptSetSet.size() == 0) {
commonAmongAlternates.clear();
}
return this;
}