Module org.apache.lucene.analysis.smartcn

Package org.apache.lucene.analysis.cn.smart.hhmm

Class BigramDictionary

java.lang.Object

org.apache.lucene.analysis.cn.smart.hhmm.AbstractDictionary

org.apache.lucene.analysis.cn.smart.hhmm.BigramDictionary

class BigramDictionary
extends AbstractDictionary

SmartChineseAnalyzer Bigram dictionary.

Field Summary

Fields

Modifier and Type	Field	Description
private long[]	bigramHashTable
private int[]	frequencyTable
private int	max
static final int	PRIME_BIGRAM_LENGTH
private static BigramDictionary	singleInstance
static final char	WORD_SEGMENT_CHAR

Fields inherited from class org.apache.lucene.analysis.cn.smart.hhmm.AbstractDictionary

CHAR_NUM_IN_FILE, GB2312_CHAR_NUM, GB2312_FIRST_CHAR

Constructor Summary

Constructors

Modifier	Constructor	Description
private	BigramDictionary()

Method Summary

Modifier and Type	Method	Description
private int	getAvaliableIndex(long hashId, char[] carray)
private int	getBigramItemIndex(char[] carray)
int	getFrequency(char[] carray)
static BigramDictionary	getInstance()
private void	load()
private void	load(String dictRoot)
void	loadFromFile(String dctFilePath)	Load the datafile into this BigramDictionary
private void	loadFromInputStream(InputStream serialObjectInputStream)
private boolean	loadFromObj(Path serialObj)
private void	saveToObj(Path serialObj)

Methods inherited from class org.apache.lucene.analysis.cn.smart.hhmm.AbstractDictionary

getCCByGB2312Id, getGB2312Id, hash1, hash1, hash2, hash2

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

WORD_SEGMENT_CHAR

public static final char WORD_SEGMENT_CHAR

See Also:

• Constant Field Values

singleInstance

private static BigramDictionary singleInstance

PRIME_BIGRAM_LENGTH

public static final int PRIME_BIGRAM_LENGTH

See Also:

• Constant Field Values

bigramHashTable

private long[] bigramHashTable

frequencyTable

private int[] frequencyTable

max

private int max

Constructor Details

BigramDictionary

private BigramDictionary()

Method Details

getInstance

public static BigramDictionary getInstance()

loadFromObj

private boolean loadFromObj(Path serialObj)

loadFromInputStream

private void loadFromInputStream(InputStream serialObjectInputStream)
                          throws IOException,
ClassNotFoundException

Throws:

IOException

ClassNotFoundException

saveToObj

private void saveToObj(Path serialObj)
                throws IOException

Throws:

IOException

load

private void load()
           throws IOException,
ClassNotFoundException

Throws:

IOException

ClassNotFoundException

load

private void load(String dictRoot)
           throws IOException

Throws:

IOException

loadFromFile

public void loadFromFile(String dctFilePath)
                  throws IOException

Load the datafile into this BigramDictionary

Parameters:

dctFilePath - path to the Bigramdictionary (bigramdict.dct)

Throws:

IOException - If there is a low-level I/O error

getAvaliableIndex

private int getAvaliableIndex(long hashId,
 char[] carray)

getBigramItemIndex

private int getBigramItemIndex(char[] carray)

getFrequency

public int getFrequency(char[] carray)