org.wltea.analyzer.query.SWMCQueryBuilder Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of ik-analyzers Show documentation
bestwu ik-analyzer
The newest version!
package org.wltea.analyzer.query;

import lombok.extern.slf4j.Slf4j;
import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.queryparser.classic.ParseException;
import org.apache.lucene.queryparser.classic.QueryParser;
import org.apache.lucene.search.Query;
import org.wltea.analyzer.core.IKSegmenter;
import org.wltea.analyzer.core.Lexeme;

import java.io.IOException;
import java.io.StringReader;
import java.util.ArrayList;
import java.util.List;

/**
 * Single Word Multi Char Query Builder
 * IK分词算法专用
 *
 * @author linliangyi
 */
@Slf4j
public class SWMCQueryBuilder {

	/**
	 * 生成SWMCQuery
	 *
	 * @param fieldName fieldName
	 * @param keywords  keywords
	 * @param quickMode quickMode
	 * @return Lucene Query
	 */
	public static Query create(String fieldName, String keywords, boolean quickMode) {
		if (fieldName == null || keywords == null) {
			throw new IllegalArgumentException("参数 fieldName 、 keywords 不能为null.");
		}
		//1.对keywords进行分词处理
		List lexemes = doAnalyze(keywords);
		//2.根据分词结果，生成SWMCQuery
		return getSWMCQuery(fieldName, lexemes, quickMode);
	}

	/**
	 * 分词切分，并返回结链表
	 *
	 * @param keywords keywords
	 * @return 结链表
	 */
	private static List doAnalyze(String keywords) {
		List lexemes = new ArrayList<>();
		IKSegmenter ikSeg = new IKSegmenter(new StringReader(keywords), true);
		try {
			Lexeme l;
			while ((l = ikSeg.next()) != null) {
				lexemes.add(l);
			}
		} catch (IOException e) {
			log.error(e.getMessage(), e);
		}
		return lexemes;
	}

	/**
	 * 根据分词结果生成SWMC搜索
	 *
	 * @param fieldName fieldName
	 * @param lexemes   lexemes
	 * @param quickMode quickMode
	 * @return return
	 */
	private static Query getSWMCQuery(String fieldName, List lexemes, boolean quickMode) {
		//构造SWMC的查询表达式
		StringBuilder keywordBuffer = new StringBuilder();
		//精简的SWMC的查询表达式
		StringBuilder keywordBuffer_Short = new StringBuilder();
		//记录最后词元长度
		int lastLexemeLength = 0;
		//记录最后词元结束位置
		int lastLexemeEnd = -1;

		int shortCount = 0;
		int totalCount = 0;
		for (Lexeme l : lexemes) {
			totalCount += l.getLength();
			//精简表达式
			if (l.getLength() > 1) {
				keywordBuffer_Short.append(' ').append(l.getLexemeText());
				shortCount += l.getLength();
			}

			if (lastLexemeLength == 0) {
				keywordBuffer.append(l.getLexemeText());
			} else if (lastLexemeLength == 1 && l.getLength() == 1
					&& lastLexemeEnd == l.getBeginPosition()) {//单字位置相邻，长度为一，合并)
				keywordBuffer.append(l.getLexemeText());
			} else {
				keywordBuffer.append(' ').append(l.getLexemeText());

			}
			lastLexemeLength = l.getLength();
			lastLexemeEnd = l.getEndPosition();
		}

		//借助lucene queryparser 生成SWMC Query
		QueryParser qp = new QueryParser(fieldName, new StandardAnalyzer());
		qp.setDefaultOperator(QueryParser.AND_OPERATOR);
		qp.setAutoGeneratePhraseQueries(true);

		if (quickMode && (shortCount * 1.0f / totalCount) > 0.5f) {
			try {
				return qp.parse(keywordBuffer_Short.toString());
			} catch (ParseException e) {
				log.error(e.getMessage(), e);
			}

		} else {
			if (keywordBuffer.length() > 0) {
				try {
					return qp.parse(keywordBuffer.toString());
				} catch (ParseException e) {
					log.error(e.getMessage(), e);
				}
			}
		}
		return null;
	}
}