package com.limegroup.gnutella.util;
/**
* Abstract clas to use for different normalization implementations
* Helper functions like blockSplit are also defined here.
*/
abstract class AbstractI18NConverter {
/**
* This method should return the converted form of the string s
* this method should also split s into the different
* unicode blocks
* @param s String to be converted
* @return the converted string
*/
public abstract String getNorm(String s);
/**
* Simple composition of a string.
*/
public abstract String compose(String s);
/**
* Returns a string split according to the unicode blocks. A
* space '\u0020' will be splaced between the blocks.
* The index to the blockStarts array will be used to compare
* when splitting the string.
* @param String s
* @return string split into blocks with '\u0020' as the delim
*/
String blockSplit(String s) {
if(s.length() == 0) return s;
else {
int blockb4 = of(s.charAt(0));
int curBlock;
StringBuffer buf = new StringBuffer();
buf.append(s.charAt(0));
for(int i = 1, n = s.length(); i < n; i++) {
curBlock = of(s.charAt(i));
//compare the blocks of the current char and the char
//right before. Also, make sure we don't add too many
//'\u0020' chars
if(curBlock != blockb4 &&
(s.charAt(i) != '\u0020' && s.charAt(i - 1) != '\u0020'))
buf.append("\u0020");
buf.append(s.charAt(i));
blockb4 = curBlock;
}
//get rid of trailing space (if any)
return buf.toString().trim();
}
}
/**
* Returns which unicode block the parameter c
* belongs to. The returned int is the index to the blockStarts
* array.
* @param char c
* @return index to array
*/
int of(char c) {
int top, bottom, current;
bottom = 0;
top = blockStarts.length;
current = top/2;
while (top - bottom > 1) {
if (c >= blockStarts[current]) {
bottom = current;
} else {
top = current;
}
current = (top + bottom) / 2;
}
return current;
}
/**
* copy from Character.java
* the boundaries for each of the unicode blocks
*/
static final char blockStarts[] = {
'\u0000',
'\u0080',
'\u0100',
'\u0180',
'\u0250',
'\u02B0',
'\u0300',
'\u0370',
'\u0400',
'\u0500', // unassigned
'\u0530',
'\u0590',
'\u0600',
'\u0700', // unassigned
'\u0900',
'\u0980',
'\u0A00',
'\u0A80',
'\u0B00',
'\u0B80',
'\u0C00',
'\u0C80',
'\u0D00',
'\u0D80', // unassigned
'\u0E00',
'\u0E80',
'\u0F00',
'\u0FC0', // unassigned
'\u10A0',
'\u1100',
'\u1200', // unassigned
'\u13A0',
'\u1400',
'\u1680',
'\u16A0',
'\u1700',
'\u1720',
'\u1740',
'\u1760',
'\u1780',
'\u1800',
'\u1900',
'\u1950',
'\u19E0',
'\u1D00',
'\u1E00',
'\u1F00',
'\u2000',
'\u2070',
'\u20A0',
'\u20D0',
'\u2100',
'\u2150',
'\u2190',
'\u2200',
'\u2300',
'\u2400',
'\u2440',
'\u2460',
'\u2500',
'\u2580',
'\u25A0',
'\u2600',
'\u2700',
'\u27C0', // unassigned
'\u3000',
'\u3040',
'\u30A0',
'\u3100',
'\u3130',
'\u3190',
'\u3200',
'\u3300',
'\u3400', // unassigned
'\u4E00',
'\uA000', // unassigned
'\uAC00',
'\uD7A4', // unassigned
'\uD800',
'\uE000',
'\uF900',
'\uFB00',
'\uFB50',
'\uFE00', // unassigned
'\uFE20',
'\uFE30',
'\uFE50',
'\uFE70',
'\uFEFF', // special
'\uFF00',
'\uFFF0'
};
}