Jskad/source/org/thdl/lucene/TshegBarTokenizer.java

131 lines
4.4 KiB
Java

package org.thdl.lucene;
import org.apache.lucene.analysis.*;
import org.apache.lucene.analysis.Token;
import org.thdl.tib.scanner.*;
import java.io.*;
import java.util.*;
/**
* Takes stream of Unicode Tibetan text and tokenizes it
* into "syllables" or tsheg bars. Note that this is not
* equivalent to tokenizing into "words" since words frequently
* consist of more than one tsheg bar.
* <p>
* Non-Tibetan text and Tibetan punctuation is ignored by this
* class.
*
* @author Edward Garrett
*/
public class TshegBarTokenizer extends Tokenizer {
public int offset = 0;
public static void main(String[] args) {
try {
Writer out = new BufferedWriter(new OutputStreamWriter(new FileOutputStream("out.txt"), "UTF-8"));
TshegBarTokenizer tok = new TshegBarTokenizer(new StringReader(args[0]));
Token next = tok.next();
while (next != null) {
out.write(next.termText() + "\n");
next = tok.next();
}
out.flush();
out.close();
} catch (IOException ioe) {
ioe.printStackTrace();
}
}
/**
* Constructs a <code>TshegBarTokenizer</code> from a <code>Reader</code>
*/
public TshegBarTokenizer(Reader in) {
super(in);
}
/**
* Processes a stream of Tibetan text into a sequence of tsheg bars.
* Note that tsheg bars are returned in citation (dictionary) form,
* meaning in the normal case that final tshegs are added if not already
* present.
*
* @return the next tsheg bar in the stream
*/
public Token next() throws IOException {
int c;
do {
c = input.read();
offset++;
} while (c!=-1 && !isPartOfTshegBar((char)c));
if (c==-1) return null; //reached end of stream without finding token
//otherwise, this is the start of the tsheg bar
int start = offset-1;
StringBuffer buffy = new StringBuffer();
buffy.append((char)c);
c = input.read();
offset++;
while (c!=-1 && isPartOfTshegBar((char)c)) {
buffy.append((char)c);
c = input.read();
offset++;
}
buffy.append('\u0F0B'); //add tsheg to end of token
String token = buffy.toString();
if (c == '\u0F0B') {
return new Token(token.toString(), start, offset, "?"); //include tsheg for purposes of highlighting
} else {
return new Token(token.toString(), start, offset-1, "?"); //type "?" means not yet tagged
}
}
/**
* Determines whether or not passed character belongs to the "inner" (contentful)
* part of a Tibetan tsheg bar.
*
* @return <code>true</code> if <code>c</code> is both Tibetan and
* not a <code>Character.NON_SPACING_MARK</code>; <code>false</code>
* otherwise
*/
public static boolean isPartOfTshegBar(char c) {
return (Character.UnicodeBlock.of(c) == Character.UnicodeBlock.TIBETAN &&
(Character.isLetterOrDigit(c) || Character.getType(c) == Character.NON_SPACING_MARK))
? true : false;
}
//returns tsheg bar if string represents single tsheg bar; otherwise returns null
/**
* Reduces a series of possibly multiple tsheg bars to a single tsheg bar,
* ignoring all tsheg bars after the first.
*
* @return the first tsheg bar within the passed
* <code>String</code>
*/
public static String getSingleTshegBar(String val) throws IOException {
TshegBarTokenizer tokenizer = new TshegBarTokenizer(new StringReader(val));
Token next = tokenizer.next();
if (next!=null) {
if (tokenizer.next()==null) {
return next.termText();
}
}
return null;
}
/**
*
* @return array containing tsheg bars occurring in passed <code>String</String>
*/
public static String[] getTshegBars(String text_bo) throws IOException {
TshegBarTokenizer tok = new TshegBarTokenizer(new StringReader(text_bo));
List<String> tokens = new ArrayList<String>();
Token next = tok.next();
while (next != null) {
tokens.add(next.termText());
next = tok.next();
}
return (String[])tokens.toArray(new String[0]);
}
}