lexiconimpl.java

这是java 开发的的免费语音播放插件,很值得学习参考!!!!!!!!!!!!111

    	System.arraycopy(phones, 0, copy, 0, phones.length);
            return copy;
    	}
    	else return null;
    	
    }
    /**
     * Gets a phone list for a word from a given lexicon.  If a phone
     * list cannot be found, returns <code>null</code>.  The format is 
     * lexicon dependent.  If the part of speech does not matter, pass
     * in <code>null</code>.
     *
     * @param lexicon the lexicon
     * @param word the word to find
     * @param partOfSpeech the part of speech
     *
     * @return the list of phones for word or <code>null</code>
     */
    protected String[] getPhones(Map lexicon,
                                 String word,
                                 String partOfSpeech) {
        String[] phones;
        partOfSpeech = fixPartOfSpeech(partOfSpeech);
        phones = getPhones(lexicon, word+partOfSpeech);
        for (int i = 0;
             (i < partsOfSpeech.size()) && (phones == null);
             i++) {
            if (!partOfSpeech.equals((String) partsOfSpeech.get(i))) {
                phones = getPhones(lexicon,
                                   word + (String) partsOfSpeech.get(i));
            }
        }
        return phones;
    }

    /**
     * Gets a phone list for a word from a given lexicon.  If a phone
     * list cannot be found, returns <code>null</code>.
     *
     * @param lexicon the lexicon
     * @param wordAndPartOfSpeech word and part of speech concatenated
     *   together
     *
     * @return the list of phones for word or <code>null</code>
     */
    protected String[] getPhones(Map lexicon,
                                 String wordAndPartOfSpeech) {
        Object value = lexicon.get(wordAndPartOfSpeech);
        if (value instanceof String[]) {
            return (String[]) value;
        } else if (value instanceof String) {
            String[] phoneArray;
            phoneArray = getPhones((String) value);
            if (tokenizeOnLookup) {
                lexicon.put(wordAndPartOfSpeech, phoneArray);
            }
            return phoneArray;
        } else {
            return null;
        }
    }
    
    /** 
     * Turns the phone <code>String</code> into a <code>String[]</code>,
     * using " " as the delimiter.
     *
     * @param phones the phones
     *
     * @return the phones split into an array
     */
    protected String[] getPhones(String phones) {
        ArrayList phoneList = new ArrayList();
        StringTokenizer tokenizer = new StringTokenizer(phones, " ");
        while (tokenizer.hasMoreTokens()) {
            phoneList.add(tokenizer.nextToken());
        }
        return (String[]) phoneList.toArray(new String[0]);
    } 
    
    /**
     * Adds a word to the addenda.
     *
     * @param word the word to find
     * @param partOfSpeech the part of speech
     * @param phones the phones for the word
     * 
     */
    public void addAddendum(String word,
                            String partOfSpeech,
                            String[] phones) {
        String pos = fixPartOfSpeech(partOfSpeech);
        if (!partsOfSpeech.contains(pos)) {
            partsOfSpeech.add(pos);
        }
        addenda.put(word + pos, phones);
    }   

    /**
     * Removes a word from the addenda.
     *
     * @param word the word to remove
     * @param partOfSpeech the part of speech
     */
    public void removeAddendum(String word, String partOfSpeech) {
        addenda.remove(word + fixPartOfSpeech(partOfSpeech));        
    }

    /**
     * Outputs a string to a data output stream.
     *
     * @param dos the data output stream
     * @param s the string to output
     *
     * @throws IOException if errors occur during writing
     */
    private void outString(DataOutputStream dos, String s) 
			throws IOException {
	dos.writeByte((byte) s.length());
	for (int i = 0; i < s.length(); i++) {
	    dos.writeChar(s.charAt(i));
	}
    }

    /**
     * Inputs a string from a DataInputStream.  This method is not re-entrant.
     *
     * @param dis the data input stream
     *
     * @return the string 
     *
     * @throws IOException if errors occur during reading
     */
    private String getString(DataInputStream dis) throws IOException {
	int size = dis.readByte();
	for (int i = 0; i < size; i++) {
	    charBuffer[i] = dis.readChar();
	}
	return new String(charBuffer, 0, size);
    }

    /**
     * Inputs a string from a DataInputStream.  This method is not re-entrant.
     *
     * @param bb the input byte buffer
     *
     * @return the string 
     *
     * @throws IOException if errors occur during reading
     */
    private String getString(ByteBuffer bb) throws IOException {
	int size = bb.get();
	for (int i = 0; i < size; i++) {
	    charBuffer[i] = bb.getChar();
	}
	return new String(charBuffer, 0, size);
    }


    /**
     * Dumps a binary form of the database.  This method is not thread-safe.
     * 
     * <p>Binary format is:
     * <pre>
     * MAGIC
     * VERSION
     * (int) numPhonemes
     * (String) phoneme0
     * (String) phoneme1
     * (String) phonemeN
     * (int) numEntries
     * (String) nameWithPOS 
     * (byte) numPhonemes
     * phoneme index 1
     * phoneme index 2
     * phoneme index n
     * </pre>
     *
     * <p>Strings are formatted as: <code>(byte) len char0 char1 charN</code>
     *
     * <p>Limits: Strings: 128 chars
     * <p>Limits: Strings: 128 phonemes per word
     *
     * @param lexicon the lexicon to dump 
     * @param path the path to dump the file to
     */
    private void dumpBinaryLexicon(Map lexicon, String path) {
	try {
	    FileOutputStream fos = new FileOutputStream(path);
	    DataOutputStream dos = new DataOutputStream(new
		    BufferedOutputStream(fos));
	    List phonemeList = findPhonemes(lexicon);

	    dos.writeInt(MAGIC);
	    dos.writeInt(VERSION);
	    dos.writeInt(phonemeList.size());

	    for (int i = 0; i < phonemeList.size(); i++) {
		outString(dos, (String) phonemeList.get(i));
	    }

	    dos.writeInt(lexicon.keySet().size());
	    for (Iterator i = lexicon.keySet().iterator(); i.hasNext(); ) {
		String key = (String) i.next();
		outString(dos, key);
		String[] phonemes = getPhones(lexicon, key);
		dos.writeByte((byte) phonemes.length);
		for (int index = 0; index < phonemes.length; index++) {
		    int phonemeIndex = phonemeList.indexOf(phonemes[index]);
		    if (phonemeIndex == -1) {
			throw new Error("Can't find phoneme index");
		    }
		    dos.writeByte((byte) phonemeIndex);
		}
	    }
	    dos.close();
	} catch (FileNotFoundException fe) {
	    throw new Error("Can't dump binary database " +
		    fe.getMessage());
	} catch (IOException ioe) {
	    throw new Error("Can't write binary database " +
		    ioe.getMessage());
	}
    }

    /**
     * Loads the binary lexicon from the given InputStream.
     * This method is not thread safe.
     *
     * @param is the InputStream to load the database from
     * @param estimatedSize estimate of how large the database is
     *
     * @return a <code>Map</code> containing the lexicon
     *
     * @throws IOException if an IO error occurs
     */
    private Map loadMappedBinaryLexicon(FileInputStream is, int estimatedSize) 
	throws IOException {
	FileChannel fc = is.getChannel();

	MappedByteBuffer bb = fc.map(FileChannel.MapMode.READ_ONLY, 
		0, (int) fc.size());
	bb.load();
	int size = 0;
	int numEntries = 0;
	List phonemeList = new ArrayList();

	// we get better performance for some reason if we
	// just ignore estimated size
        //
	// Map lexicon = new HashMap();
        Map lexicon = new LinkedHashMap(estimatedSize * 4 / 3);

	if (bb.getInt() != MAGIC) {
	    throw new Error("bad magic number in lexicon");
	}

	if (bb.getInt() != VERSION) {
	    throw new Error("bad version number in lexicon");
	}

	size = bb.getInt();
	for (int i = 0; i < size; i++) {
	    String phoneme = getString(bb);
	    phonemeList.add(phoneme);
	}
	numEntries = bb.getInt();

	for (int i = 0; i < numEntries; i++) {
	    String wordAndPos = getString(bb);
	    String pos = Character.toString(
		    wordAndPos.charAt(wordAndPos.length() - 1));
	    if (!partsOfSpeech.contains(pos)) {
		partsOfSpeech.add(pos);
	    }

	    int numPhonemes = bb.get();
	    String[] phonemes = new String[numPhonemes];

	    for (int j = 0; j < numPhonemes; j++) {
		phonemes[j] = (String) phonemeList.get(bb.get());
	    }
	    lexicon.put(wordAndPos, phonemes);
	}
	fc.close();
	return lexicon;
    }

    /**
     * Loads the binary lexicon from the given InputStream.
     * This method is not thread safe.
     *
     * @param is the InputStream to load the database from
     * @param estimatedSize estimate of how large the database is
     *
     * @return a <code>Map</code> containing the lexicon
     *
     * @throws IOException if an IO error occurs
     */
    private Map loadBinaryLexicon(InputStream is, int estimatedSize) 
	throws IOException {
	DataInputStream dis = new DataInputStream(new
		BufferedInputStream(is));
	int size = 0;
	int numEntries = 0;
	List phonemeList = new ArrayList();

	// we get better performance for some reason if we
	// just ignore estimated size
        //
        Map lexicon = new LinkedHashMap();

	if (dis.readInt() != MAGIC) {
	    throw new Error("bad magic number in lexicon");
	}

	if (dis.readInt() != VERSION) {
	    throw new Error("bad version number in lexicon");
	}

	size = dis.readInt();
	for (int i = 0; i < size; i++) {
	    String phoneme = getString(dis);
	    phonemeList.add(phoneme);
	}
	numEntries = dis.readInt();

	for (int i = 0; i < numEntries; i++) {
	    String wordAndPos = getString(dis);
	    String pos = Character.toString(
		    wordAndPos.charAt(wordAndPos.length() - 1));
	    if (!partsOfSpeech.contains(pos)) {
		partsOfSpeech.add(pos);
	    }

	    int numPhonemes = dis.readByte();
	    String[] phonemes = new String[numPhonemes];

	    for (int j = 0; j < numPhonemes; j++) {
		phonemes[j] = (String) phonemeList.get(dis.readByte());
	    }
	    lexicon.put(wordAndPos, phonemes);
	}
	dis.close();
	return lexicon;
    }

    /**
     * Dumps this lexicon (just the compiled form). Lexicon will be
     * dumped to two binary files PATH_compiled.bin and
     * PATH_addenda.bin
     *
     * @param path the root path to dump it to
     */
    public void dumpBinary(String path) {
        String compiledPath = path + "_compiled.bin";
        String addendaPath = path + "_addenda.bin";
        
        dumpBinaryLexicon(compiled, compiledPath);
        dumpBinaryLexicon(addenda, addendaPath);
    }

    /**
     * Returns a list of the unique phonemes in the lexicon.
     *
     * @param lexicon the lexicon of interest
     *
     * @return list the unique set of phonemes
     */
    private List findPhonemes(Map lexicon) {
	List phonemeList = new ArrayList();
	for (Iterator i = lexicon.keySet().iterator(); i.hasNext(); ) {
	    String key = (String) i.next();
	    String[] phonemes = getPhones(lexicon, key);
	    for (int index = 0; index < phonemes.length; index++) {
		if (!phonemeList.contains(phonemes[index])) {
		    phonemeList.add(phonemes[index]);
		}
	    }
	}
	return phonemeList;
    }


    /**
     * Tests to see if this lexicon is identical to the other for
     * debugging purposes.
     *
     * @param other the other lexicon to compare to
     *
     * @return true if lexicons are identical
     */
    public boolean compare(LexiconImpl other) {
	return compare(addenda, other.addenda) && 
	      compare(compiled, other.compiled);
    }

    /**
     * Determines if the two lexicons are identical for debugging purposes.
     *
     * @param lex this lex
     * @param other the other lexicon to chd
     *
     * @return true if they are identical
     */
    private boolean compare(Map lex, Map other) {
	for (Iterator i = lex.keySet().iterator(); i.hasNext(); ) {
	    String key = (String) i.next();
	    String[] thisPhonemes = getPhones(lex, key);
	    String[] otherPhonemes = getPhones(other, key);
	    if (thisPhonemes == null) {
		System.out.println(key + " not found in this.");
		return false;
	    } else if (otherPhonemes == null) {
		System.out.println(key + " not found in other.");
		return false;
	    } else if (thisPhonemes.length == otherPhonemes.length) {
		for (int j = 0; j < thisPhonemes.length; j++) {
		    if (!thisPhonemes[j].equals(otherPhonemes[j])) {
			return false;
		    }
		}
	    } else {
		return false;
	    }
	}
	return true;
    }
 
    /**
     * Fixes the part of speech if it is <code>null</code>.  The
     * default representation of a <code>null</code> part of speech
     * is the number "0".
     */
    static protected String fixPartOfSpeech(String partOfSpeech) {
        return (partOfSpeech == null) ? "0" : partOfSpeech;
    }
}