es api------高级搜索、高亮查询、聚合查询、加权排序

es的搜索类型（SearchType）有4种

query and fetch(速度最快)(返回N倍数据量，多个副本连同自己分片中的数据都会呈现出来，数据重复！)

 query then fetch（默认的搜索方式） 

 DFS query and fetch  ~>用来给检索到的每条记录计算分数

 DFS query then fetch(可以更精确控制搜索打分和排名。)

从性能考虑QUERY_AND_FETCH是最快的，DFS_QUERY_THEN_FETCH是最慢的。从搜索的准确度来说，DFS要比非DFS的准确度更高。

分页检索api代码

public class SearchType {
    /**
     * TransportClient实例，用于访问远程的es集群
     */
    private TransportClient client;

    /**
     * 初始化
     */
    @Before
    public void init() throws UnknownHostException {
        Settings settings = Settings.builder().put("cluster.name", "cluster").build();
        client = new PreBuiltTransportClient(settings);
        client.addTransportAddress(new TransportAddress(InetAddress.getByName("mini1"), 9300));
        client.addTransportAddress(new TransportAddress(InetAddress.getByName("mini2"), 9300));
        client.addTransportAddress(new TransportAddress(InetAddress.getByName("mini3"), 9300));
    }

    
    //测试分页查询
    @Test
    public void testSearch() {
        //设置每页的大小
        int pageSize=3;
        //设置查询第几页的内容
        int currentPageNo=1;
        SearchResponse res = client
                .prepareSearch("bigdata")
                .setTypes("product")
          .setSearchType(org.elasticsearch.action.search.SearchType.DFS_QUERY_THEN_FETCH)
                .setFrom((currentPageNo-1)*pageSize)
                .setSize(pageSize)
                //精准查询
//                .setQuery(QueryBuilders.termQuery("name","sqoop"))
                //正则,name字段里以s开头的都会查到
//                .setQuery(QueryBuilders.regexpQuery("name","s.*"))
                //通配符，意思同上
                .setQuery(QueryBuilders.wildcardQuery("name","s*"))
                .get();
        //查询结果