Lucene入门1：索引的增、查

因工作需要接触Lucene，今天简单对Lucene索引的增删改查有了一个初步了解。先用起来，再逐步根据需要深入了解其原理。

1. 新增索引

步骤：

1)创建Directory，指定索引存放目录
2)创建索引写入配置对象，指定分词器
3)创建索引写入对象IndexWriter
4)创建Document对象，存储索引
5)为Document添加Field
6)添加Document到Index

public class luceneIndexUtil {

    private int[] ids = {1, 2, 3};
    private String[] authors = {"jason", "neo", "rzexin"};
    private String[] articles = {"vim can do everything!", "go go", "blockchain"};
    private String[] file_paths = {"/tmp/lucene/a.txt", "/tmp/lucene/b.txt", "/tmp/lucene/c.txt"};

    private static String INDEX_PATH = "/tmp/lucene/index.1";

    public void createIndex() throws IOException {

        //1)创建Directory，指定索引存放目录
        FSDirectory dir = FSDirectory.open(Paths.get(INDEX_PATH));

        //2)创建索引写入配置对象，指定分词器
        // 因存在中文文档，使用中文分词器，默认是标准分词器（StandardAnalyzer）
        // pom.xml需添加库依赖：lucene-analyzers-smartcn
        IndexWriterConfig indexWriterConfig = new IndexWriterConfig(new SmartChineseAnalyzer());

        //3)创建索引写入对象IndexWriter
        IndexWriter indexWriter = new IndexWriter(dir, indexWriterConfig);

        Document doc = null;
        for (int i=0; i<ids.length; ++i) {
            //4)创建Document对象，存储索引
            doc = new Document();

            //5)为Document添加Field
            //IntPoint - 对int字段，只索引不存储
            doc.add(new IntPoint("id", ids[i]));
            //存储Field的值
            doc.add(new StoredField("id", ids[i]));

            //StringField - 只索引不分词，即字符串作为一个整体进行索引
            //Field.Store.YES - 将这个域中内容完全存储到文件，方便进行文本还原
            doc.add(new StringField("author", authors[i], Field.Store.YES));

            //TextField - 索引并分词
            //Field.Store.NO - 不存储到文件，可以被索引，但内容不可还原
            doc.add(new TextField("article", articles[i], Field.Store.NO));

            //读取文件，默认：Field.Store.NO
            doc.add(new TextField("content", new FileReader(new File(file_paths[i]))));

            //6)添加Document到Index
            indexWriter.addDocument(doc);
        }

        indexWriter.close();
    }
}

测试：

    @Test
    public void testCreateIndex() throws IOException {
        luceneIndexUtil luceneIndexUtil = new luceneIndexUtil();
        luceneIndexUtil.createIndex();
    }

执行后，生成文件：

2. 搜索

步骤：

1)创建Directory
2)创建IndexReader
3)创建IndexSearcher
4)创建搜索的Query
5)返回TopDocs
6)获取ScoreDoc对象
7)获取具体Document对象
8)从Document对象中获取需要值

    public void searchIndex() throws IOException, ParseException {
        //1)创建Directory
        FSDirectory dir = FSDirectory.open(Paths.get(INDEX_PATH));

        //2)创建IndexReader
        IndexReader indexReader = DirectoryReader.open(dir);

        //3)创建IndexSearcher
        IndexSearcher indexSearcher = new IndexSearcher(indexReader);

        //4)创建搜索的Query
        QueryParser parser = new QueryParser("content", new SmartChineseAnalyzer());
        //搜索content包含关键字【天空】的文档
        Query query = parser.parse("天空");
        //QueryParser parser = new QueryParser("article", new StandardAnalyzer());
        //Query query = parser.parse("Go");

        //5)返回TopDocs
        TopDocs topDocs = indexSearcher.search(query, 10);

        //6)获取ScoreDoc对象
        ScoreDoc[] scoreDocs = topDocs.scoreDocs;
        for(ScoreDoc scoreDoc : scoreDocs) {
            //7)获取具体Document对象
            Document doc = indexSearcher.doc(scoreDoc.doc);

            //8)从Document对象中获取需要值
            System.out.printf("id:%s, author:%s, article:%s, content:%s\n",
                    doc.get("id"), doc.get("author"),
                    doc.get("article"), doc.get("content"));
        }

        indexReader.close();
    }

测试：

    @Test
    public void testSearchIndex() throws IOException, ParseException {
        luceneIndexUtil luceneIndexUtil = new luceneIndexUtil();
        luceneIndexUtil.searchIndex();
    }