// if numChars is 0, this means match as many as you want
private int __parseUnsignedInteger(int radix, int minDigits, int maxDigits)
throws MalformedPatternException {
int num, digits = 0;
StringBuffer buf;
// We don't expect huge numbers, so an initial buffer of 4 is fine.
buf = new StringBuffer(4);
while (Character.digit(__lookahead, radix) != -1 && digits < maxDigits) {
buf.append((char) __lookahead);
__match(__lookahead);
++digits;
}
if (digits < minDigits || digits > maxDigits)
throw new MalformedPatternException(
"Parse error: unexpected number of digits at position " + __bytesRead);
try {
num = Integer.parseInt(buf.toString(), radix);
} catch (NumberFormatException e) {
throw new MalformedPatternException(
"Parse error: numeric value at " + "position " + __bytesRead + " is invalid");
}
return num;
}
private SyntaxNode __backslashToken() throws MalformedPatternException {
SyntaxNode current;
char token;
int number;
__match('\\');
if (__lookahead == 'x') {
__match('x');
// Parse a hexadecimal number
current = _newTokenNode((char) __parseUnsignedInteger(16, 2, 2), __position++);
} else if (__lookahead == 'c') {
__match('c');
// Create a control character
token = Character.toUpperCase(__lookahead);
token = (char) (token > 63 ? token - 64 : token + 64);
current = new TokenNode(token, __position++);
__match(__lookahead);
} else if (__lookahead >= '0' && __lookahead <= '9') {
__match(__lookahead);
if (__lookahead >= '0' && __lookahead <= '9') {
// We have an octal character or a multi-digit backreference.
// Assume octal character for now.
__putback();
number = __parseUnsignedInteger(10, 2, 3);
number = Integer.parseInt(Integer.toString(number), 8);
current = _newTokenNode((char) number, __position++);
} else {
// We have either \0, an escaped digit, or a backreference.
__putback();
if (__lookahead == '0') {
// \0 matches the null character
__match('0');
current = new TokenNode('\0', __position++);
} else {
// Either an escaped digit or backreference.
number = Character.digit(__lookahead, 10);
current = _newTokenNode(__lookahead, __position++);
__match(__lookahead);
}
}
} else if (__lookahead == 'b') {
// Inside of a character class the \b means backspace, otherwise
// it means a word boundary
// if(__inCharacterClass)
// \b always means backspace
current = new TokenNode('\b', __position++);
/*
else
current = new TokenNode((char)LeafNode._WORD_BOUNDARY_MARKER_TOKEN,
position++);
*/
__match('b');
} /*else if(__lookahead == 'B' && !__inCharacterClass){
current = new TokenNode((char)LeafNode._NONWORD_BOUNDARY_MARKER_TOKEN,
position++);
__match('B');
} */ else {
CharacterClassNode characterSet;
token = __lookahead;
switch (__lookahead) {
case 'n':
token = '\n';
break;
case 'r':
token = '\r';
break;
case 't':
token = '\t';
break;
case 'f':
token = '\f';
break;
}
switch (token) {
case 'd':
characterSet = new CharacterClassNode(__position++);
characterSet._addTokenRange('0', '9');
current = characterSet;
break;
case 'D':
characterSet = new NegativeCharacterClassNode(__position++);
characterSet._addTokenRange('0', '9');
current = characterSet;
break;
case 'w':
characterSet = new CharacterClassNode(__position++);
characterSet._addTokenRange('0', '9');
characterSet._addTokenRange('a', 'z');
characterSet._addTokenRange('A', 'Z');
characterSet._addToken('_');
current = characterSet;
break;
case 'W':
characterSet = new NegativeCharacterClassNode(__position++);
characterSet._addTokenRange('0', '9');
characterSet._addTokenRange('a', 'z');
characterSet._addTokenRange('A', 'Z');
characterSet._addToken('_');
current = characterSet;
break;
case 's':
characterSet = new CharacterClassNode(__position++);
characterSet._addToken(' ');
characterSet._addToken('\f');
characterSet._addToken('\n');
characterSet._addToken('\r');
characterSet._addToken('\t');
current = characterSet;
break;
case 'S':
characterSet = new NegativeCharacterClassNode(__position++);
characterSet._addToken(' ');
characterSet._addToken('\f');
characterSet._addToken('\n');
characterSet._addToken('\r');
characterSet._addToken('\t');
current = characterSet;
break;
default:
current = _newTokenNode(token, __position++);
break;
}
__match(__lookahead);
}
return current;
}
