首页 > 编程学习 > Lucene 3.6.1:中文分词、创建索引库、排序、多字段分页查询以及高亮显示源码
2012八月29

Lucene 3.6.1:中文分词、创建索引库、排序、多字段分页查询以及高亮显示源码

1、准备工作

下载lucene 3.6.1 : http://lucene.apache.org/

下载中文分词IK Analyzer: http://code.google.com/p/ik-analyzer/downloads/list (注意下载的是IK Analyzer 2012_u5_source.zip,其他版本有bug)

下载solr 3.6.1:  http://lucene.apache.org/solr/(编译IK Analyzer时需引用包)

OK,将lucene 、solr 相关包(lucene-core-3.6.1.jar、lucene-highlighter-3.6.1.jar、lucene-analyzers-3.6.1.jar、apache-solr-core-3.6.1.jar、apache-solr-solrj-3.6.1.jar)拷贝到项目lib下,IK源码置于项目src下。

2、从Oracle数据库中取数据创建索引(使用IK分词)

package lucene.util;

import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.index.CorruptIndexException;
import org.apache.lucene.store.FSDirectory;
import org.apache.lucene.store.Directory;
import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.util.Version;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.wltea.analyzer.lucene.IKAnalyzer;

import java.sql.Connection;
import java.io.File;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Date;

import modules.gk.Gk_info;
import modules.gk.Gk_infoSub;
import web.sys.Globals;
import web.db.DBConnector;
import web.db.ObjectCtl;
import web.util.StringUtil;
//Wizzer.cn
public class LuceneIndex {
    IndexWriter writer = null;
    FSDirectory dir = null;
    boolean create = true;//是否初始化&覆盖索引库

    public void init() {
        long a1 = System.currentTimeMillis();
        System.out.println("[Lucene 开始执行:" + new Date() + "]");
        Connection con = DBConnector.getconecttion(); //取得一个数据库连接
        try {
            final File docDir = new File(Globals.SYS_COM_CONFIG.get("sys.index.path").toString());//E:\lucene
            if (!docDir.exists()) {
                docDir.mkdirs();
            }
            String cr = Globals.SYS_COM_CONFIG.get("sys.index.create").toString();//true or false
            if ("false".equals(cr.toLowerCase())) {
                create = false;
            }
            dir = FSDirectory.open(docDir);
//            Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_36);
            Analyzer analyzer = new IKAnalyzer(true);
            IndexWriterConfig iwc = new IndexWriterConfig(Version.LUCENE_36, analyzer);
            if (create) {
                // Create a new index in the directory, removing any
                // previously indexed documents:
                iwc.setOpenMode(IndexWriterConfig.OpenMode.CREATE);
            } else {
                // Add new documents to an existing index:
                iwc.setOpenMode(IndexWriterConfig.OpenMode.CREATE_OR_APPEND);
            }
            writer = new IndexWriter(dir, iwc);
            String sql = "SELECT indexno,title,describes,pdate,keywords FROM TABLEA WHERE STATE=1 AND SSTAG<>1 ";
            int rowCount = ObjectCtl.getRowCount(con, sql);
            int pageSize = StringUtil.StringToInt(Globals.SYS_COM_CONFIG.get("sys.index.size").toString());   //每页记录数
            int pages = (rowCount - 1) / pageSize + 1; //计算总页数
            ArrayList list = null;
            Gk_infoSub gk = null;
            for (int i = 1; i < pages+1; i++) {
                long a = System.currentTimeMillis();
                list = ObjectCtl.listPage(con, sql, i, pageSize, new Gk_infoSub());
                for (int j = 0; j < list.size(); j++) {
                    gk = (Gk_infoSub) list.get(j);
                    Document doc = new Document();
                    doc.add(new Field("indexno", StringUtil.null2String(gk.getIndexno()), Field.Store.YES, Field.Index.NOT_ANALYZED_NO_NORMS));//主键不分词
                    doc.add(new Field("title", StringUtil.null2String(gk.getTitle()), Field.Store.YES, Field.Index.ANALYZED));
                    doc.add(new Field("describes", StringUtil.null2String(gk.getDescribes()), Field.Store.YES, Field.Index.ANALYZED));
                    doc.add(new Field("pdate", StringUtil.null2String(gk.getPdate()), Field.Store.YES, Field.Index.NOT_ANALYZED_NO_NORMS));//日期不分词
                    doc.add(new Field("keywords", StringUtil.null2String(gk.getKeywords()), Field.Store.YES, Field.Index.ANALYZED));
                    writer.addDocument(doc);
                    ObjectCtl.executeUpdateBySql(con,"UPDATE TABLEA SET SSTAG=1 WHERE indexno='"+gk.getIndexno()+"'");//更新已索引状态
                }

                long b = System.currentTimeMillis();
                long c = b - a;
                System.out.println("[Lucene " + rowCount + "条," + pages + "页,第" + i + "页花费时间:" + c + "毫秒]");
            }
            writer.commit();

        } catch (Exception e) {
            e.printStackTrace();
        } finally {
            DBConnector.freecon(con); //释放数据库连接
            try {
                if (writer != null) {
                    writer.close();
                }
            } catch (CorruptIndexException e) {
                e.printStackTrace();
            } catch (IOException e) {
                e.printStackTrace();
            } finally {
                try {
                    if (dir != null && IndexWriter.isLocked(dir)) {
                        IndexWriter.unlock(dir);//注意解锁
                    }
                } catch (IOException e) {
                    e.printStackTrace();
                }
            }
        }
        long b1 = System.currentTimeMillis();
        long c1 = b1 - a1;
        System.out.println("[Lucene 执行完毕,花费时间:" + c1 + "毫秒,完成时间:" + new Date() + "]");
    }
}

 

3、单字段查询以及多字段分页查询高亮显示

 

package lucene.util;

import org.apache.lucene.store.FSDirectory;
import org.apache.lucene.store.Directory;
import org.apache.lucene.search.*;
import org.apache.lucene.search.highlight.SimpleHTMLFormatter;
import org.apache.lucene.search.highlight.Highlighter;
import org.apache.lucene.search.highlight.SimpleFragmenter;
import org.apache.lucene.search.highlight.QueryScorer;
import org.apache.lucene.queryParser.QueryParser;
import org.apache.lucene.queryParser.MultiFieldQueryParser;
import org.apache.lucene.analysis.TokenStream;
import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.KeywordAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.Term;
import org.apache.lucene.util.Version;
import modules.gk.Gk_infoSub;

import java.util.ArrayList;
import java.io.File;
import java.io.StringReader;
import java.lang.reflect.Constructor;

import web.util.StringUtil;
import web.sys.Globals;
import org.wltea.analyzer.lucene.IKAnalyzer;
//Wizzer.cn
public class LuceneQuery {
    private static String indexPath;// 索引生成的目录
    private int rowCount;// 记录数
    private int pages;// 总页数
    private int currentPage;// 当前页数
    private int pageSize;   //每页记录数

    public LuceneQuery() {
        this.indexPath = Globals.SYS_COM_CONFIG.get("sys.index.path").toString();
    }

    public int getRowCount() {
        return rowCount;
    }

    public int getPages() {
        return pages;
    }

    public int getPageSize() {
        return pageSize;
    }

    public int getCurrentPage() {
        return currentPage;
    }

    /**
     * 函数功能:根据字段查询索引
     */
    public ArrayList queryIndexTitle(String keyWord, int curpage, int pageSize) {
        ArrayList list = new ArrayList();
        try {
            if (curpage <= 0) {
                curpage = 1;
            }
            if (pageSize <= 0) {
                pageSize = 20;
            }
            this.pageSize = pageSize;   //每页记录数
            this.currentPage = curpage;   //当前页
            int start = (curpage - 1) * pageSize;
            Directory dir = FSDirectory.open(new File(indexPath));
            IndexReader reader = IndexReader.open(dir);
            IndexSearcher searcher = new IndexSearcher(reader);
            Analyzer analyzer = new IKAnalyzer(true);
            QueryParser queryParser = new QueryParser(Version.LUCENE_36, "title", analyzer);
            queryParser.setDefaultOperator(QueryParser.AND_OPERATOR);
            Query query = queryParser.parse(keyWord);
            int hm = start + pageSize;
            TopScoreDocCollector res = TopScoreDocCollector.create(hm, false);
            searcher.search(query, res);

            SimpleHTMLFormatter simpleHTMLFormatter = new SimpleHTMLFormatter("<span style='color:red'>", "</span>");
            Highlighter highlighter = new Highlighter(simpleHTMLFormatter, new QueryScorer(query));
            this.rowCount = res.getTotalHits();
            this.pages = (rowCount - 1) / pageSize + 1; //计算总页数
            TopDocs tds = res.topDocs(start, pageSize);
            ScoreDoc[] sd = tds.scoreDocs;
            for (int i = 0; i < sd.length; i++) {
                Document hitDoc = reader.document(sd[i].doc);
                list.add(createObj(hitDoc, analyzer, highlighter));
            }

        } catch (Exception e) {
            e.printStackTrace();
        }

        return list;

    }
    /**
     * 函数功能:根据字段查询索引
     */
    public ArrayList queryIndexFields(String allkeyword, String onekeyword, String nokeyword, int curpage, int pageSize) {
        ArrayList list = new ArrayList();
        try {
            if (curpage <= 0) {
                curpage = 1;
            }
            if (pageSize <= 0) {
                pageSize = 20;
            }
            this.pageSize = pageSize;   //每页记录数
            this.currentPage = curpage;   //当前页
            int start = (curpage - 1) * pageSize;
            Directory dir = FSDirectory.open(new File(indexPath));
            IndexReader reader = IndexReader.open(dir);
            IndexSearcher searcher = new IndexSearcher(reader);
            BooleanQuery bQuery = new BooleanQuery();  //组合查询
            if (!"".equals(allkeyword)) {//包含全部关键词
                KeywordAnalyzer analyzer = new KeywordAnalyzer();
                BooleanClause.Occur[] flags = {BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD};//AND
                Query query = MultiFieldQueryParser.parse(Version.LUCENE_36, allkeyword, new String[]{"title", "describes", "keywords"}, flags, analyzer);
                bQuery.add(query, BooleanClause.Occur.MUST);  //AND
            }
            if (!"".equals(onekeyword)) { //包含任意关键词
                Analyzer analyzer = new IKAnalyzer(true);
                BooleanClause.Occur[] flags = {BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD};//OR
                Query query = MultiFieldQueryParser.parse(Version.LUCENE_36, onekeyword, new String[]{"title", "describes", "keywords"}, flags, analyzer);
                bQuery.add(query, BooleanClause.Occur.MUST);  //AND
            }
            if (!"".equals(nokeyword)) { //排除关键词
                Analyzer analyzer = new IKAnalyzer(true);
                BooleanClause.Occur[] flags = {BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD};//NOT
                Query query = MultiFieldQueryParser.parse(Version.LUCENE_36, nokeyword, new String[]{"title", "describes", "keywords"}, flags, analyzer);
                bQuery.add(query, BooleanClause.Occur.MUST_NOT);  //AND

            }
            int hm = start + pageSize;
            TopScoreDocCollector res = TopScoreDocCollector.create(hm, false);
            searcher.search(bQuery, res);
            SimpleHTMLFormatter simpleHTMLFormatter = new SimpleHTMLFormatter("<span style='color:red'>", "</span>");
            Highlighter highlighter = new Highlighter(simpleHTMLFormatter, new QueryScorer(bQuery));
            this.rowCount = res.getTotalHits();
            this.pages = (rowCount - 1) / pageSize + 1; //计算总页数
            System.out.println("rowCount:" + rowCount);
            TopDocs tds = res.topDocs(start, pageSize);
            ScoreDoc[] sd = tds.scoreDocs;
            Analyzer analyzer = new IKAnalyzer();
            for (int i = 0; i < sd.length; i++) {
                Document hitDoc = reader.document(sd[i].doc);
                list.add(createObj(hitDoc, analyzer, highlighter));
            }

        } catch (Exception e) {
            e.printStackTrace();
        }

        return list;

    }

    /**
     * 创建返回对象(高亮)
     */

    private synchronized static Object createObj(Document doc, Analyzer analyzer, Highlighter highlighter) {

        Gk_infoSub gk = new Gk_infoSub();
        try {

            if (doc != null) {
                gk.setIndexno(StringUtil.null2String(doc.get("indexno")));
                gk.setPdate(StringUtil.null2String(doc.get("pdate")));
                String title = StringUtil.null2String(doc.get("title"));
                gk.setTitle(title);
                if (!"".equals(title)) {
                    highlighter.setTextFragmenter(new SimpleFragmenter(title.length()));
                    TokenStream tk = analyzer.tokenStream("title", new StringReader(title));
                    String htext = StringUtil.null2String(highlighter.getBestFragment(tk, title));
                    if (!"".equals(htext)) {
                        gk.setTitle(htext);
                    }
                }
                String keywords = StringUtil.null2String(doc.get("keywords"));
                gk.setKeywords(keywords);
                if (!"".equals(keywords)) {
                    highlighter.setTextFragmenter(new SimpleFragmenter(keywords.length()));
                    TokenStream tk = analyzer.tokenStream("keywords", new StringReader(keywords));
                    String htext = StringUtil.null2String(highlighter.getBestFragment(tk, keywords));
                    if (!"".equals(htext)) {
                        gk.setKeywords(htext);
                    }
                }
                String describes = StringUtil.null2String(doc.get("describes"));
                gk.setDescribes(describes);
                if (!"".equals(describes)) {
                    highlighter.setTextFragmenter(new SimpleFragmenter(describes.length()));
                    TokenStream tk = analyzer.tokenStream("keywords", new StringReader(describes));
                    String htext = StringUtil.null2String(highlighter.getBestFragment(tk, describes));
                    if (!"".equals(htext)) {
                        gk.setDescribes(htext);
                    }
                }

            }
            return gk;
        }
        catch (Exception e) {

            e.printStackTrace();
            return null;
        }
        finally {
            gk = null;
        }

    }

    private synchronized static Object createObj(Document doc) {

        Gk_infoSub gk = new Gk_infoSub();
        try {

            if (doc != null) {
                gk.setIndexno(StringUtil.null2String(doc.get("indexno")));
                gk.setPdate(StringUtil.null2String(doc.get("pdate")));
                gk.setTitle(StringUtil.null2String(doc.get("title")));
                gk.setKeywords(StringUtil.null2String(doc.get("keywords")));
                gk.setDescribes(StringUtil.null2String(doc.get("describes")));
            }
            return gk;
        }
        catch (Exception e) {

            e.printStackTrace();
            return null;
        }
        finally {
            gk = null;
        }

    }
}

 

单字段查询:

        long a = System.currentTimeMillis();
        try {
            int curpage = StringUtil.StringToInt(StringUtil.null2String(form.get("curpage")));
            int pagesize = StringUtil.StringToInt(StringUtil.null2String(form.get("pagesize")));
            String title = StringUtil.replaceLuceneStr(StringUtil.null2String(form.get("title")));
            LuceneQuery lu = new LuceneQuery();
            form.addResult("list", lu.queryIndexTitle(title, curpage, pagesize));
            form.addResult("curPage", lu.getCurrentPage());
            form.addResult("pageSize", lu.getPageSize());
            form.addResult("rowCount", lu.getRowCount());
            form.addResult("pageCount", lu.getPages());
        } catch (Exception e) {
            e.printStackTrace();
        }
        long b = System.currentTimeMillis();
        long c = b - a;
        System.out.println("[搜索信息花费时间:" + c + "毫秒]");

多字段查询:

        long a = System.currentTimeMillis();
        try {
            int curpage = StringUtil.StringToInt(StringUtil.null2String(form.get("curpage")));
            int pagesize = StringUtil.StringToInt(StringUtil.null2String(form.get("pagesize")));
            String allkeyword = StringUtil.replaceLuceneStr(StringUtil.null2String(form.get("allkeyword")));
            String onekeyword = StringUtil.replaceLuceneStr(StringUtil.null2String(form.get("onekeyword")));
            String nokeyword = StringUtil.replaceLuceneStr(StringUtil.null2String(form.get("nokeyword")));
            LuceneQuery lu = new LuceneQuery();
            form.addResult("list", lu.queryIndexFields(allkeyword,onekeyword,nokeyword, curpage, pagesize));
            form.addResult("curPage", lu.getCurrentPage());
            form.addResult("pageSize", lu.getPageSize());
            form.addResult("rowCount", lu.getRowCount());
            form.addResult("pageCount", lu.getPages());
        } catch (Exception e) {
            e.printStackTrace();
        }
        long b = System.currentTimeMillis();
        long c = b - a;
        System.out.println("[高级检索花费时间:" + c + "毫秒]");

4、Lucene通配符查询

            BooleanQuery bQuery = new BooleanQuery();  //组合查询
            if (!"".equals(title)) {
                WildcardQuery w1 = new WildcardQuery(new Term("title", title+ "*"));

                bQuery.add(w1, BooleanClause.Occur.MUST);  //AND
            }
            int hm = start + pageSize;
            TopScoreDocCollector res = TopScoreDocCollector.create(hm, false);
            searcher.search(bQuery, res);

 

5、Lucene嵌套查询

实现SQL:(unitid like ‘unitid%’  and idml like ‘id2%’) or (tounitid like ‘unitid%’ and tomlid like ‘id2%’ and tostate=1)

                    BooleanQuery bQuery = new BooleanQuery();
                    BooleanQuery b1 = new BooleanQuery();
                    WildcardQuery w1 = new WildcardQuery(new Term("unitid", unitid + "*"));
                    WildcardQuery w2 = new WildcardQuery(new Term("idml", id2 + "*"));
                    b1.add(w1, BooleanClause.Occur.MUST);//AND
                    b1.add(w2, BooleanClause.Occur.MUST);//AND
                    bQuery.add(b1, BooleanClause.Occur.SHOULD);//OR
                    BooleanQuery b2 = new BooleanQuery();
                    WildcardQuery w3 = new WildcardQuery(new Term("tounitid", unitid + "*"));
                    WildcardQuery w4 = new WildcardQuery(new Term("tomlid", id2 + "*"));
                    WildcardQuery w5 = new WildcardQuery(new Term("tostate", "1"));
                    b2.add(w3, BooleanClause.Occur.MUST);//AND
                    b2.add(w4, BooleanClause.Occur.MUST);//AND
                    b2.add(w5, BooleanClause.Occur.MUST);//AND
                    bQuery.add(b2, BooleanClause.Occur.SHOULD);//OR

6、Lucene先根据时间排序后分页

下面这种方式不太合理,建议在创建索引库的时候排序,这样查询的时候只用分页即可,若有多个排序条件可单独创建索引库。
int hm = start + pageSize;
Sort sort = new Sort(new SortField(“pdate”, SortField.STRING, true));
TopScoreDocCollector res = TopScoreDocCollector.create(pageSize, false);
searcher.search(bQuery, res);
this.rowCount = res.getTotalHits();
this.pages = (rowCount – 1) / pageSize + 1; //计算总页数
TopDocs tds =searcher.search(bQuery,rowCount,sort);// res.topDocs(start, pageSize);
ScoreDoc[] sd = tds.scoreDocs;
System.out.println(“rowCount:” + rowCount);
int i=0;
for (ScoreDoc scoreDoc : sd) {
i++;
if(i<start){
continue;
}
if(i>hm){
break;
}
Document doc = searcher.doc(scoreDoc.doc);
list.add(createObj(doc));
}


最新的排序写法:

            int hm = start + pageSize;
            Sort sort = new Sort();
            SortField sortField = new SortField("pdate", SortField.STRING, true);
            sort.setSort(sortField);
            TopDocs hits = searcher.search(bQuery, null, hm, sort);
            this.rowCount = hits.totalHits;
            this.pages = (rowCount - 1) / pageSize + 1; //计算总页数
            for (int i = start; i < hits.scoreDocs.length; i++) {
                ScoreDoc sdoc = hits.scoreDocs[i];
                Document doc = searcher.doc(sdoc.doc);
                list.add(createObj(doc));
            }

ps:
周一完成创建索引库定时任务,周二实现模糊查询中文分词高亮显示及分页,今天实现了通配符查询、嵌套查询、先排序后分页,从零接触到实现Lucene主要功能花了三天时间,当然,性能如何还待测试和优化。

Loading

本文地址:https://wizzer.cn/archives/2299 , 转载请保留.

3 Responses to “Lucene 3.6.1:中文分词、创建索引库、排序、多字段分页查询以及高亮显示源码”

  1. #1 285872852@qq.com 回复 | 引用 Post:2012-11-22 14:13

    TopDocs hits = searcher.search(bQuery, null, hm, sort);
    先排序后分页.这个明显不对哟,这里的数据,根本就不对。
    int hm = start + pageSize;
    这里并不是全部的数据,有一部份少掉了。

发表评论