dictionarytest.java

基于词典和最大匹配算法的的中文分词组件

/**
 * 
 */
package org.solol.mmseg.test;

import org.solol.mmseg.core.IDictionary;
import org.solol.mmseg.core.IWord;
import org.solol.mmseg.internal.Dictionary;
import org.solol.mmseg.internal.Word;

import junit.framework.TestCase;

/**
 * @author solo L
 *
 */
public class DictionaryTest extends TestCase {
	
	IDictionary dictionary = new Dictionary();
	
	String word1Value = "我们";
	IWord word1 = new Word(word1Value,Word.CJK_WORD);
	
	String word2Value = "是";
	IWord word2 = new Word(word2Value,Word.CJK_WORD);
	
	String word3Value = "中学生";
	IWord word3 = new Word(word3Value,Word.CJK_WORD);	
	
	String word4Value = "学生";
	IWord word4 = new Word(word4Value,Word.CJK_WORD);
	
	/* (non-Javadoc)
	 * @see junit.framework.TestCase#setUp()
	 */
	protected void setUp() throws Exception {
		dictionary.addWord(word1Value,Word.CJK_WORD);
		dictionary.addWord(word2Value,Word.CJK_WORD);
		dictionary.addWord(word3Value,Word.CJK_WORD);
	}

	/* (non-Javadoc)
	 * @see junit.framework.TestCase#tearDown()
	 */
	protected void tearDown() throws Exception {
	}

	public final void testIsMatched() {
		assertTrue(dictionary.isMatched(word1Value));
		assertTrue(dictionary.isMatched(word2Value));
		assertTrue(dictionary.isMatched(word3Value));
		assertFalse(dictionary.isMatched(word4Value));
	}

	public final void testAddWord() {
		dictionary.addWord(word4Value,Word.CJK_WORD);	
		assertTrue(dictionary.isMatched(word4Value));
		dictionary.removeWord(word4Value);
		assertFalse(dictionary.isMatched(word4Value));
	}
	
	public final void testGetWord() {
		assertEquals(word1, dictionary.getWord(word1Value));
		assertEquals(word2, dictionary.getWord(word2Value));
		assertEquals(word3, dictionary.getWord(word3Value));
	}

	public final void testRemoveWord() {
		dictionary.addWord(word4Value,Word.CJK_WORD);	
		assertTrue(dictionary.isMatched(word4Value));
		dictionary.removeWord(word4Value);	
		assertFalse(dictionary.isMatched(word4Value));
	}

}