source: trunk/gli/src/org/greenstone/gatherer/util/DecodeHTMLReader.java@ 4364

package org.greenstone.gatherer.util;

import java.io.*;
import org.greenstone.gatherer.Gatherer;

public class DecodeHTMLReader
    extends PushbackReader {

    public DecodeHTMLReader(Reader source) {
    super(source, 4);
    }

    /** Read a single character. */
    public int read() {
    return decode();
    }
 
    /** Read characters into a portion of an array. */
    public int read(char[] cbuf, int off, int len) {
    int count = 0;
    for(int i = off; i < len && ready(); i++) {
        cbuf[i] = (char)decode();
        count++;
    }
    return count;
    }

    public boolean ready() {
    try {
        return super.ready();
    }
    catch (Exception error) {
    }
    return false;
    }

    /** Retrieve the next character off the stream. Unfortunately I have to do this a character at a time (which is slow). I also have to keep in mind that if a suspect encoded character turns out not to be I have to replace the extra characters. */
    private int decode() {
    int character;
    try {
        character = super.read();
    }
    catch (Exception error) {
        character = ' ';
    }
    try {
        if(character == '&') {
        int amp = super.read();
        switch(amp) {
        case 'a':
        case 'A':
            int ampap = super.read();
            int ampapo = super.read();
            int ampapos = super.read();
            int ampapossemi = super.read();
            if((ampap == 'p' || ampap == 'P') && (ampapo == 'o' || ampapo == 'O') && (ampapos == 's' || ampapos == 'S') && ampapossemi == ';') {
            // Read an &apos; so return an apostrophy
            return '\'';
            }
            // Not a apos. Return the characters removed in the correct order.
            super.unread(ampapossemi);
            super.unread(ampapos);
            super.unread(ampapo);
            super.unread(ampap);
            break;
        case 'g':
        case 'G':
            int ampgt = super.read();
            int ampgtsemi = super.read();
            if((ampgt == 't' || ampgt == 'T') && ampgtsemi == ';') {
            return '>';
            }
            super.unread(ampgtsemi);
            super.unread(ampgt);
            break;
        case 'l':
        case 'L':
            int amplt = super.read();
            int ampltsemi = super.read();
            if((amplt == 't' || amplt == 'T') && ampltsemi == ';') {
            return '<';
            }
            super.unread(ampltsemi);
            super.unread(amplt);
            break;
        case 'q':
        case 'Q':
            int ampqu = super.read();
            int ampquo = super.read();
            int ampquot = super.read();
            int ampquotsemi = super.read();
            if((ampqu == 'u' || ampqu == 'U') && (ampquo == 'o' || ampquo == 'O') && (ampquot == 't' || ampquot == 'T') && ampquotsemi == ';') {
            return '\"';
            }
            super.unread(ampquotsemi);
            super.unread(ampquot);
            super.unread(ampquo);
            super.unread(ampqu);
            break;
        case '#':
            int amphash = super.read();
            int amphash3 = super.read();
            int amphash39 = super.read();
            int amphash39semi = super.read();
            if(amphash == '#' && amphash3 == '3' && amphash39 == '9' && amphash39semi == ';') {
            return '\'';
            }
            super.unread(amphash39semi);
            super.unread(amphash39);
            super.unread(amphash3);
            super.unread(amphash);
            break;
        }
        // Not a suspect. Return the character removed.
        super.unread(amp);
        }
    }
    catch (Exception error) {
        Gatherer.printStackTrace(error);
    }
    // Nothing special. Simply return the character extracted.
    return character;
    }
}