/**
* Parse a Unicode identifier from the given string at the given
* position. Return the identifier, or an empty string if there
* is no identifier.
* @param str the string to parse
* @param pos INPUT-OUPUT parameter. On INPUT, pos is the
* first character to examine. It must be less than str.length(),
* and it must not point to a whitespace character. That is, must
* have pos < str.length(). On
* OUTPUT, the position after the last parsed character.
* @return the Unicode identifier, or an empty string if there is
* no valid identifier at pos.
*/
UnicodeString ICU_Utility::parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos) {
// assert(pos < str.length());
UnicodeString buf;
int p = pos;
while (p < str.length()) {
UChar32 ch = str.char32At(p);
if (buf.length() == 0) {
if (u_isIDStart(ch)) {
buf.append(ch);
} else {
buf.truncate(0);
return buf;
}
} else {
if (u_isIDPart(ch)) {
buf.append(ch);
} else {
break;
}
}
p += U16_LENGTH(ch);
}
pos = p;
return buf;
}
//
// RBBISymbolTable::parseReference This function from the abstract symbol table interface
// looks for a $variable name in the source text.
// It does not look it up, only scans for it.
// It is used by the UnicodeSet parser.
//
// This implementation is lifted pretty much verbatim
// from the rules based transliterator implementation.
// I didn't see an obvious way of sharing it.
//
UnicodeString RBBISymbolTable::parseReference(const UnicodeString& text,
ParsePosition& pos, int32_t limit) const
{
int32_t start = pos.getIndex();
int32_t i = start;
UnicodeString result;
while (i < limit) {
UChar c = text.charAt(i);
if ((i==start && !u_isIDStart(c)) || !u_isIDPart(c)) {
break;
}
++i;
}
if (i == start) { // No valid name chars
return result; // Indicate failure with empty string
}
pos.setIndex(i);
text.extractBetween(start, i, result);
return result;
}
bool IsIdStart(codepoint_t ch)
{
#ifdef HAS_REAL_ICU
if (u_isIDStart(ch))
{
return true;
}
#endif
// Following codepoints are treated as part of ID_Start
// for backwards compatibility as per section 2.5 of the Unicode 8 spec
// See http://www.unicode.org/reports/tr31/tr31-23.html#Backward_Compatibility
// The exact list is in PropList.txt in the Unicode database
switch (ch)
{
case 0x2118: return true; // SCRIPT CAPITAL P
case 0x212E: return true; // ESTIMATED SYMBOL
case 0x309B: return true; // KATAKANA-HIRAGANA VOICED SOUND MARK
case 0x309C: return true; // KATAKANA-HIRAGANA SEMI-VOICED SOUND MARK
default: return false;
}
}
static jboolean Character_isUnicodeIdentifierStartImpl(JNIEnv*, jclass, jint codePoint) {
return u_isIDStart(codePoint);
}
//static jboolean Character_isUnicodeIdentifierStartImpl(JNIEnv*, jclass, jint codePoint) {
JNIEXPORT jboolean JNICALL
Java_java_lang_Character_isUnicodeIdentifierStartImpl(JNIEnv*, jclass, jint codePoint) {
return u_isIDStart(codePoint);
}
jboolean fastiva_vm_Character_C$__isUnicodeIdentifierStartImpl(jint codePoint) {
return u_isIDStart(codePoint);
}
