Elastic Search 学习入门之Search全文检索(七)

蔚落 2024-03-24 15:56 57阅读 0赞

ES是基于[Lucene][]的开源搜索引擎，其查询语法关键字部分和Lucene大致一样：

分页： from/size、字段：fields、排序：sort、查询：query

过滤：filter、高亮：highlight、统计：facet

### ES的搜索类型有4种(以下说明是基于elasticsearch2.3): ###

query and fetch (速度最快)(返回N倍数据量) **受保护，5.3之前可用**

query then fetch (默认的搜索方式)

DFS query and fetch **没有了**

DFS query then fetch (可以更精确控制搜索打分和排名)

DFS：这个D可能是Distributed，F可能是frequency的缩写，至于S可能是[Scatter][]的缩写，整个单词可能是分布式词频率和文档频率散发的缩写。

初始化散发：从ES的官方网站可以发现，初始化散发其实就是在进行真正的查询之前，先把各个分片的词频率和文档频率收集一下，然后进行词搜索的时候，各分片依据全局的词频率和文档频率进行搜索和排名。显然如果使用DFS\_QUERY\_THEN\_FETCH这种查询方式，效率是最低的，因为一个搜索，可能要请求3次分片。但使用DFS方法，搜索精度应该是最高的。

综上，从性能考虑：

QUERY\_AND\_FETCH是最快的，DFS\_QUERY\_THEN\_FETCH是最慢的。

从搜索的精确度：

DFS要比非DFS的准确度更高。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzI4ODQ0NzY3_size_16_color_FFFFFF_t_70][]

[ElasticSearch][]查询：

对应每个查询项，我们可以通过must、should、mustNot方法对QueryBuilder进行组合，形成多条件查询。(must => and, should=>or)

Luncene 支持基于词条的TermQuery、RangeQuery、PrefixQuery、BooleanQuery、PhraseQuery、WildcardQuery、FuzzQuery

*  TermQuery与QueryParser

单个单词作为查询表达式时，它相当于一个单独的项，如果表达式是由单个单词构成，QueryParser的parse()方法会返回一个TermQuery对象。

如查询表达式为content:hello, QueryParser会返回一个域为content，值为hello的TermQuery。

Query query = new TermQuery("content", "hello")

*  RangeQuery与QUeryParser

QueryParser可以使用\[ 起始 To 终止 \] 或 \{ 起始 To 终止 \}表达式来构造RangeQuery。

如查询表达式：time：\[20181010 To 20181210\], QueryParser会返回一个域为time，下限为20181010，上限为20181210的RangeQuery。

Term t1 = new Term("time", "20181010");

Term t2 = new Term("time", "20181210");

Query query = new RangeQuery(t1, t2, true);

*  PrefixQuery与QueryParser
    
     *  当查询表达式中短语以星号(\*)结尾时，QueryParser会创建一个PrefixQuery对象。
     *  如查询表达式为content：luc\*, 则QueryParser会返回一个域为content，值为luc的PrefixQuery
        
        Query query = new PrefixQuery(luc);
 *  BooleanQuery与QueryParser
    
     *  当查询表达式中包含多个项时，QueryParser可以方便的构建BooleanQuery。QueryParser使用圆括号分组，通过-，+，AND，OR及NOT来指定所生成的BooleanQuery。

*  PhraseQuery与QueryParser

在QueryParser的分析表达式中双引号的若干项会被转换为一个PhraseQuery对象，默认情况下，Slop因子为0，可以在表达式中通过~n来指定slop因子的值。

如查询表达式为content:"hello world" ~3, 则QueryParser会返回一个域为content，内容为"hello world", slope为3的短语查询。

Query query = new PhraseQuery();

query.setSlop(3);

query.add(new Term("content", "hello"));

query.add(new Term("content", "world"));

*  Wildcard与QueryParser

Luncene使用两个标准的通配符号，\*代表0或多个字母，？代表0或1个字母。但查询表达式中包含\*或者？时，则QueryParser会返回一个WildcardQuery对象。但要注意的是，当\*出现在查询表达式的末尾时，会被优化为PrefixQuery；并且查询表达式的首个字符不能是通配符，防止用户输入以通配符\*为前缀的搜索表达式，导致Lucene枚举所有的项而耗费巨大的资源。

*  FuzzyQuery和QueryParser

QueryParser通过在某个项之后添加"~"来支持FuzzyQuery类的模糊查询。

### 代码实现： ###

*  #### 简单查询及显示所有内容： ####
 *      @Test
            public void testQuery1(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        /**
                         * 设置检索方式：
                         *  QUERY_AND_FETCH:  在5.3之前，之后受保护
                         *  QUERY_THEN_FETCH:  默认
                         *  DFS_QUERY_AND_FETCH:  直接移除，新版本没有
                         *  DFS_QUERY_THEN_FETCH:
                         */
                        .setSearchType(SearchType.DEFAULT)
                        /**
                         * 设置要检索的内容
                         * 基于不同的检索方式，是否能够检索到想要的数据，就逐渐衍生出来了一个职位SEO，搜索引擎优化
                         */
        //                .setQuery(QueryBuilders.matchPhrasePrefixQuery("firstname", "V*")) // 在firstname字段上检索以V开头的数据
        //                .setQuery(QueryBuilders.matchQuery("state", "NM"))
                        .setQuery(QueryBuilders.termQuery("age", 40))
                        //分页，每页显示M条，显示第N页的数据setFrom((N - 1) * M ).setSize()
                        .setFrom(1)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
         
                /**
                 * "hits": [
                 * {
                 * "_index": "product",
                 * "_type": "bigdata",
                 * "_id": "5",
                 * "_score": 1,
                 * "_source": {
                 * "name": "redis",
                 * "author": "redis",
                 * "version": "5.0.0"
                 * }
                 * }
                 */
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){
                    System.out.println("--------------------------------------------");
                    String index = hit.getIndex();
                    String type = hit.getType();
                    String id = hit.getId();
                    float score = hit.getScore();
                    System.out.println("index: " + index);
                    System.out.println("type: " + type);
                    System.out.println("id: " + id);
                    System.out.println("score: " + score);
                    Map<String, Object> source = hit.getSourceAsMap();
                    source.forEach((field, value) ->{
                        System.out.println(field + "--->" + value);
                    });
         
                }
            }
 *  #### 查询字段部分高亮显示： ####
 *      @Test
            public void testHightLight(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .highlighter(//设置高亮显示
                                SearchSourceBuilder.highlight()
                                        .field("address")
                                        .preTags("<font color='red' size='16px'>")
                                        .postTags("</font>")
                        )
                        .setFrom(0)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){//获取高亮显示的内容
                    System.out.println("-------------------------------------------");
                    //高亮字段内容
                    Map<String, HighlightField> highlightFields = hit.getHighlightFields();
                    highlightFields.forEach((key,highlightField) -> {
                        System.out.println("key: " + key);
                        String address = "";
                        Text[] fragments = highlightField.fragments();
                        for (Text fragment : fragments){
                            address += fragment.toString();
                        }
                        System.out.println("address: " + address);
         
                    });
         
         
                }
         
            }
 *  #### 按照某个字段进行排序显示： ####
 *      @Test
            public void testSort(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .highlighter(//设置高亮显示
                                SearchSourceBuilder.highlight()
                                        .field("address")
                                        .preTags("<font color='red' size='16px'>")
                                        .postTags("</font>")
                        )
                        .addSort("age", SortOrder.ASC)
        //                .addSort("age", SortOrder.DESC)
                        .setFrom(0)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){//获取高亮显示的内容
                    System.out.println("-------------------------------------------");
                    Map<String, Object> source = hit.getSourceAsMap();
                    Object firstname = source.get("firstname");
                    Object age = source.get("age");
                    System.out.println("firstname: " + firstname);
                    System.out.println("age: " + age);
                    //高亮字段内容
                    Map<String, HighlightField> highlightFields = hit.getHighlightFields();
                    highlightFields.forEach((key,highlightField) -> {
                        System.out.println("key: " + key);
                        String address = "";
                        Text[] fragments = highlightField.fragments();
                        for (Text fragment : fragments){
                            address += fragment.toString();
                        }
                        System.out.println("address: " + address);
                    });
                }
         
            }
 *  #### 聚合操作测试： ####
 *      @Test
            public void testAggr(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .addAggregation(
                                AggregationBuilders
                                        .avg("avg_age")//select avg(age) avg_age --> 这里面的name就是最好显示的别名
                                        .field("age")//select max(age), min(age), avg(age) avg_age --> 这里面的field就是这里的age对应列，或者索引库中的field
                        )
                        .get();
                Aggregations aggrs = response.getAggregations();//是个集合
        //        System.out.println(aggrs);
                for (Aggregation aggr : aggrs){
        //            System.out.println(aggr);
        //            System.out.println(aggr.getName());
        //            System.out.println(aggr.getType());
                    InternalAvg avg = (InternalAvg) aggr;
                    double value = avg.getValue();
                    System.out.println(avg.getName() + "-->" + value);
                }
            }
 *  #### 过滤部分字段范围测试： ####
 *      @Test
            public void testFilter(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .highlighter(//设置高亮显示
                                SearchSourceBuilder.highlight()
                                        .field("address")
                                        .preTags("<font color='red' size='16px'>")
                                        .postTags("</font>")
                        )
                        //过滤年龄在30~35之间的数据
                        .setPostFilter(
                                QueryBuilders.rangeQuery("age").gte(30).lte(35)
                        )
                        .addSort("age", SortOrder.ASC)
                        .setFrom(0)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){//获取高亮显示的内容
                    System.out.println("-------------------------------------------");
                    Map<String, Object> source = hit.getSourceAsMap();
                    Object firstname = source.get("firstname");
                    Object age = source.get("age");
                    System.out.println("firstname: " + firstname);
                    System.out.println("age: " + age);
                    //高亮字段内容
                    Map<String, HighlightField> highlightFields = hit.getHighlightFields();
                    highlightFields.forEach((key,highlightField) -> {
                        System.out.println("key: " + key);
                        String address = "";
                        Text[] fragments = highlightField.fragments();
                        for (Text fragment : fragments){
                            address += fragment.toString();
                        }
                        System.out.println("address: " + address);
                    });
                }
            }
 *  #### 全代码： ####
    
    elasticsearch.conf:
    
        cluster.name=rk-ES
        cluster.host.port=hadoop01:9300,hadoop02:9300,hadoop03:9300
    
        package rk.constants;
         
        /**
         * @Author rk
         * @Date 2018/12/10 15:14
         * @Description:
         **/
        public interface Constants {
            String CLUSTER_NAME = "cluster.name";
            String CLUSTER_HOST_PORT = "cluster.host.port";
         
        }
 *      package rk.elastic;
         
        import org.elasticsearch.action.search.SearchResponse;
        import org.elasticsearch.action.search.SearchType;
        import org.elasticsearch.client.transport.TransportClient;
        import org.elasticsearch.common.settings.Settings;
        import org.elasticsearch.common.text.Text;
        import org.elasticsearch.common.transport.TransportAddress;
        import org.elasticsearch.index.query.QueryBuilders;
        import org.elasticsearch.search.SearchHit;
        import org.elasticsearch.search.SearchHits;
        import org.elasticsearch.search.aggregations.Aggregation;
        import org.elasticsearch.search.aggregations.AggregationBuilders;
        import org.elasticsearch.search.aggregations.Aggregations;
        import org.elasticsearch.search.aggregations.metrics.avg.InternalAvg;
        import org.elasticsearch.search.builder.SearchSourceBuilder;
        import org.elasticsearch.search.fetch.subphase.highlight.HighlightField;
        import org.elasticsearch.search.sort.SortOrder;
        import org.elasticsearch.transport.client.PreBuiltTransportClient;
        import org.junit.After;
        import org.junit.Before;
        import org.junit.Test;
        import rk.constants.Constants;
         
        import java.io.IOException;
        import java.io.InputStream;
        import java.net.InetSocketAddress;
        import java.util.Map;
        import java.util.Properties;
         
        /**
         * @Author rk
         * @Date 2018/12/10 15:06
         * @Description:
         **/
        public class ElasticSearchTest2 {
         
            private TransportClient client;
            @Before
            public void setUp() throws IOException {
                Properties properties = new Properties();
                InputStream in = ElasticSearchTest2.class.getClassLoader().getResourceAsStream("elasticsearch.conf");
                properties.load(in);
                Settings setting = Settings.builder()
                        .put(Constants.CLUSTER_NAME,properties.getProperty(Constants.CLUSTER_NAME))
                        .build();
                client = new PreBuiltTransportClient(setting);
                String hostAndPorts = properties.getProperty(Constants.CLUSTER_HOST_PORT);
                for (String hostAndPort : hostAndPorts.split(",")){
                    String[] fields = hostAndPort.split(":");
                    String host = fields[0];
                    int port = Integer.valueOf(fields[1]);
                    TransportAddress ts = new TransportAddress(new InetSocketAddress(host, port));
                    client.addTransportAddresses(ts);
                }
                System.out.println("cluster.name = " + client.settings().get("cluster.name"));
            }
         
            String[] indices = {"product","test"};
         
            @Test
            public void testQuery1(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        /**
                         * 设置检索方式：
                         *  QUERY_AND_FETCH:  在5.3之前，之后受保护
                         *  QUERY_THEN_FETCH:  默认
                         *  DFS_QUERY_AND_FETCH:  直接移除，新版本没有
                         *  DFS_QUERY_THEN_FETCH:
                         */
                        .setSearchType(SearchType.DEFAULT)
                        /**
                         * 设置要检索的内容
                         * 基于不同的检索方式，是否能够检索到想要的数据，就逐渐衍生出来了一个职位SEO，搜索引擎优化
                         */
        //                .setQuery(QueryBuilders.matchPhrasePrefixQuery("firstname", "V*")) // 在firstname字段上检索以V开头的数据
        //                .setQuery(QueryBuilders.matchQuery("state", "NM"))
                        .setQuery(QueryBuilders.termQuery("age", 40))
                        //分页，每页显示M条，显示第N页的数据setFrom((N - 1) * M ).setSize()
                        .setFrom(1)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
         
                /**
                 * "hits": [
                 * {
                 * "_index": "product",
                 * "_type": "bigdata",
                 * "_id": "5",
                 * "_score": 1,
                 * "_source": {
                 * "name": "redis",
                 * "author": "redis",
                 * "version": "5.0.0"
                 * }
                 * }
                 */
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){
                    System.out.println("--------------------------------------------");
                    String index = hit.getIndex();
                    String type = hit.getType();
                    String id = hit.getId();
                    float score = hit.getScore();
                    System.out.println("index: " + index);
                    System.out.println("type: " + type);
                    System.out.println("id: " + id);
                    System.out.println("score: " + score);
                    Map<String, Object> source = hit.getSourceAsMap();
                    source.forEach((field, value) ->{
                        System.out.println(field + "--->" + value);
                    });
         
                }
            }
         
            @Test
            public void testHightLight(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .highlighter(//设置高亮显示
                                SearchSourceBuilder.highlight()
                                        .field("address")
                                        .preTags("<font color='red' size='16px'>")
                                        .postTags("</font>")
                        )
                        .setFrom(0)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){//获取高亮显示的内容
                    System.out.println("-------------------------------------------");
                    //高亮字段内容
                    Map<String, HighlightField> highlightFields = hit.getHighlightFields();
                    highlightFields.forEach((key,highlightField) -> {
                        System.out.println("key: " + key);
                        String address = "";
                        Text[] fragments = highlightField.fragments();
                        for (Text fragment : fragments){
                            address += fragment.toString();
                        }
                        System.out.println("address: " + address);
         
                    });
         
         
                }
         
            }
         
            @Test
            public void testSort(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .highlighter(//设置高亮显示
                                SearchSourceBuilder.highlight()
                                        .field("address")
                                        .preTags("<font color='red' size='16px'>")
                                        .postTags("</font>")
                        )
                        .addSort("age", SortOrder.ASC)
        //                .addSort("age", SortOrder.DESC)
                        .setFrom(0)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){//获取高亮显示的内容
                    System.out.println("-------------------------------------------");
                    Map<String, Object> source = hit.getSourceAsMap();
                    Object firstname = source.get("firstname");
                    Object age = source.get("age");
                    System.out.println("firstname: " + firstname);
                    System.out.println("age: " + age);
                    //高亮字段内容
                    Map<String, HighlightField> highlightFields = hit.getHighlightFields();
                    highlightFields.forEach((key,highlightField) -> {
                        System.out.println("key: " + key);
                        String address = "";
                        Text[] fragments = highlightField.fragments();
                        for (Text fragment : fragments){
                            address += fragment.toString();
                        }
                        System.out.println("address: " + address);
                    });
                }
         
            }
         
         
            @Test
            public void testAggr(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .addAggregation(
                                AggregationBuilders
                                        .avg("avg_age")//select avg(age) avg_age --> 这里面的name就是最好显示的别名
                                        .field("age")//select max(age), min(age), avg(age) avg_age --> 这里面的field就是这里的age对应列，或者索引库中的field
                        )
                        .get();
                Aggregations aggrs = response.getAggregations();//是个集合
        //        System.out.println(aggrs);
                for (Aggregation aggr : aggrs){
        //            System.out.println(aggr);
        //            System.out.println(aggr.getName());
        //            System.out.println(aggr.getType());
                    InternalAvg avg = (InternalAvg) aggr;
                    double value = avg.getValue();
                    System.out.println(avg.getName() + "-->" + value);
                }
            }
         
            @Test
            public void testFilter(){
                SearchResponse response = client
                        .prepareSearch(indices) // 指定要检索的索引库
                        .setSearchType(SearchType.DEFAULT)
                        .setQuery(QueryBuilders.matchQuery("address", "Avenue"))
                        .highlighter(//设置高亮显示
                                SearchSourceBuilder.highlight()
                                        .field("address")
                                        .preTags("<font color='red' size='16px'>")
                                        .postTags("</font>")
                        )
                        //过滤年龄在30~35之间的数据
                        .setPostFilter(
                                QueryBuilders.rangeQuery("age").gte(30).lte(35)
                        )
                        .addSort("age", SortOrder.ASC)
                        .setFrom(0)//从哪一条开始显示
                        .setSize(5)//每页显示的内容
                        .get();
                // 返回检索结果数据，被封装SearchHits对象中
                SearchHits searchHits = response.getHits();
                long totalHits = searchHits.totalHits;
                System.out.println("搜索到"+totalHits+"个结果");
                SearchHit[] hits = searchHits.getHits();
                for (SearchHit hit : hits){//获取高亮显示的内容
                    System.out.println("-------------------------------------------");
                    Map<String, Object> source = hit.getSourceAsMap();
                    Object firstname = source.get("firstname");
                    Object age = source.get("age");
                    System.out.println("firstname: " + firstname);
                    System.out.println("age: " + age);
                    //高亮字段内容
                    Map<String, HighlightField> highlightFields = hit.getHighlightFields();
                    highlightFields.forEach((key,highlightField) -> {
                        System.out.println("key: " + key);
                        String address = "";
                        Text[] fragments = highlightField.fragments();
                        for (Text fragment : fragments){
                            address += fragment.toString();
                        }
                        System.out.println("address: " + address);
                    });
                } 
            }
           
            
            @After
            public void cleanUp(){
                client.close();
            }
         
        }
    
    #### elasticsearch2.3版本时的代码实现： ####
    
        package rk.elastic;
         
         
        import com.fasterxml.jackson.databind.ObjectMapper;
        import org.dom4j.Document;
        import org.dom4j.Element;
        import org.dom4j.io.SAXReader;
        import org.elasticsearch.action.bulk.BulkRequestBuilder;
        import org.elasticsearch.action.bulk.BulkResponse;
        import org.elasticsearch.action.index.IndexRequest;
        import org.elasticsearch.action.search.SearchRequestBuilder;
        import org.elasticsearch.action.search.SearchResponse;
        import org.elasticsearch.action.search.SearchType;
        import org.elasticsearch.client.transport.TransportClient;
        import org.elasticsearch.common.settings.Settings;
        import org.elasticsearch.common.text.Text;
        import org.elasticsearch.common.transport.TransportAddress;
        import org.elasticsearch.index.query.QueryBuilders;
        import org.elasticsearch.search.SearchHit;
        import org.elasticsearch.search.SearchHits;
        import org.elasticsearch.search.fetch.subphase.highlight.HighlightField;
        import org.elasticsearch.transport.client.PreBuiltTransportClient;
        import org.junit.Before;
        import org.junit.Test;
        import rk.constants.Constants;
         
        import java.io.File;
        import java.io.IOException;
        import java.io.InputStream;
        import java.net.InetSocketAddress;
        import java.util.*;
         
        /**
         * @Author rk
         * @Date 2018/12/10 15:06
         * @Description:
         *
         *      准备数据:
         *      <doc>
         *         <url>http://gongyi.sohu.com/20120730/n349358066.shtml</url>
         *         <docno>fdaa73d52fd2f0ea-34913306c0bb3300</docno>
         *          <contenttitle>失独父母中年遇独子夭折　称不怕死亡怕养老生病</contenttitle>
         *          <content></content>
         *      </doc>
         *
         **/
         
        class Article{
            private String url;
            private String docno;
            private String content;
            private String contenttitle;
         
            public Article() {
            }
         
            public Article(String url, String docno, String content, String contenttitle) {
                this.url = url;
                this.docno = docno;
                this.content = content;
                this.contenttitle = contenttitle;
            }
         
            public String getUrl() {
                return url;
            }
         
            public void setUrl(String url) {
                this.url = url;
            }
         
            public String getDocno() {
                return docno;
            }
         
            public void setDocno(String docno) {
                this.docno = docno;
            }
         
            public String getContent() {
                return content;
            }
         
            public void setContent(String content) {
                this.content = content;
            }
         
            public String getContenttitle() {
                return contenttitle;
            }
         
            public void setContenttitle(String contenttitle) {
                this.contenttitle = contenttitle;
            }
         
            @Override
            public String toString() {
                return "Article{" +
                        "url='" + url + '\'' +
                        ", docno='" + docno + '\'' +
                        ", content='" + content + '\'' +
                        ", contenttitle='" + contenttitle + '\'' +
                        '}';
            }
        }
         
        //  解析代码，取其中前20条
        class XmlParser {
            public static List<Article> getArticle() {
                List<Article> list = new ArrayList<Article>();
                SAXReader reader = new SAXReader();
                Document document;
                try {
                    document = reader.read(new File("news_sohusite_xml"));
                    Element root = document.getRootElement();
                    Iterator<Element> iterator = root.elementIterator("doc");
                    Article article = null;
                    int count = 0;
                    while(iterator.hasNext()) {
                        Element doc = iterator.next();
                        String url = doc.elementTextTrim("url");
                        String docno = doc.elementTextTrim("docno");
                        String content = doc.elementTextTrim("content");
                        String contenttitle = doc.elementTextTrim("contenttitle");
                        article = new Article();
                        article.setContent(content);
                        article.setDocno(docno);
                        article.setContenttitle(contenttitle);
                        article.setUrl(url);
                        if(++count > 20) {
                            break;
                        }
                        list.add(article);
                    }
                } catch (Exception e) {
                    e.printStackTrace();
                }
                return list;
            }
         
        }
         
         
         
         
         
        public class ElasticSearchTest2_3 {
         
            private TransportClient client;
            @Before
            public void setUp() throws IOException {
                Properties properties = new Properties();
                InputStream in = ElasticSearchTest2.class.getClassLoader().getResourceAsStream("elasticsearch.conf");
                properties.load(in);
                Settings setting = Settings.builder()
                        .put(Constants.CLUSTER_NAME,properties.getProperty(Constants.CLUSTER_NAME))
                        .build();
                client = new PreBuiltTransportClient(setting);
                String hostAndPorts = properties.getProperty(Constants.CLUSTER_HOST_PORT);
                System.out.println(hostAndPorts);
                for (String hostAndPort : hostAndPorts.split(",")){
                    String[] fields = hostAndPort.split(":");
                    String host = fields[0];
                    int port = Integer.valueOf(fields[1]);
                    TransportAddress ts = new TransportAddress(new InetSocketAddress(host, port));
                    client.addTransportAddresses(ts);
                }
         
                TransportAddress[] transportAddrs = new TransportAddress[3];
                client.addTransportAddresses(transportAddrs);
                System.out.println("cluster.name = " + client.settings().get("cluseter.name"));
         
            }
         
            String index = "search";
            //  批量导入ES库
            @Test
            public void bulkInsert() throws Exception {
                List<Article> list = XmlParser.getArticle();
                ObjectMapper oMapper = new ObjectMapper();
                BulkRequestBuilder bulkRequestBuilder = client.prepareBulk();
                for (int i = 0; i < list.size(); i++) {
                    Article article = list.get(i);
                    String val = oMapper.writeValueAsString(article);
                    bulkRequestBuilder.add(new IndexRequest(index, "news",
                            article.getDocno()).source(val));
                }
                BulkResponse response = bulkRequestBuilder.get();
            }
         
            //查询
            @Test
            public void testSearch() {
                String indices = "bigdata";//指的是要搜索的哪一个索引库
                SearchRequestBuilder builder = client.prepareSearch(indices)
                        .setSearchType(SearchType.DEFAULT)
                        .setFrom(0)
                        .setSize(5)//设置分页
                        /**
                         * 这是最新的
                         * .highlighter(//设置高亮显示
                         *          SearchSourceBuilder.highlight()
                         *                      .field("address")
                         *                      .preTags("<font color='red' size='16px'>")
                         *                      .postTags("</font>")
                         *      )
                         */
                        .addHighlightedField("name")//设置高亮字段
                        .setHighlighterPreTags("<font color='blue'>")
                        .setHighlighterPostTags("</font>");//高亮风格
                builder.setQuery( QueryBuilders.fuzzyQuery("name", "hadoop"));
                SearchResponse searchResponse = builder.get();
                SearchHits searchHits = searchResponse.getHits();
                SearchHit[] hits = searchHits.getHits();
                long total = searchHits.getTotalHits();
                System.out.println("总共条数：" + total);//总共查询到多少条数据
                for (SearchHit searchHit : hits) {
         
                    Map<String, Object> source = searchHit.getSource();//这是最新的：searchHit.getSourceAsMap()
                    Map<String, HighlightField> highlightFields = searchHit.getHighlightFields();
                    System.out.println("---------------------------");
                    String name = source.get("name").toString();
                    String author = source.get("author").toString();
                    System.out.println("name=" + name);
                    System.out.println("author=" + author);
                    HighlightField highlightField = highlightFields.get("name");
                    if(highlightField != null) {
                        Text[] fragments = highlightField.fragments();
                        name = "";
                        for (Text text : fragments) {
                            name += text.toString();
                        }
                    }
                    System.out.println("name: " + name);
                    System.out.println("author: " + author);
                }
            }
         
        }
    
    与SQL使用LIMIT来控制单“页”数量类似，Elasticsearch使用的是from以及size两个参数：
    
    from：从哪条结果开始，默认值为0
    
    size：每次返回多少个结果，默认值为10
    
    假设每页显示5条结果，那么1至3页的请求就是：
    
    GET /\_search?size=5
    
    GET /\_search?size=5&from=5
    
    GET /\_search?size=5&from=10
    
    注意：不要一次请求过多或者页码过大的结果，这么会对服务器造成很大的压力。因为它们会在返回前排序。一个请求会经过多个分片。每个分片都会生成自己的排序结果。然后再进行集中整理，以确保最终结果的正确性。  
    ————————————————  
    版权声明：本文为CSDN博主「R\_记忆犹新」的原创文章，遵循CC 4.0 BY-SA版权协议，转载请附上原文出处链接及本声明。  
    原文链接：https://blog.csdn.net/qq\_28844767/article/details/84946433

[Lucene]: https://so.csdn.net/so/search?q=Lucene&spm=1001.2101.3001.7020
[Scatter]: https://so.csdn.net/so/search?q=Scatter&spm=1001.2101.3001.7020
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzI4ODQ0NzY3_size_16_color_FFFFFF_t_70]: https://image.dandelioncloud.cn/pgy_files/images/2024/03/24/ca97606ece7247cd8aa650be3fa1d941.png
[ElasticSearch]: https://so.csdn.net/so/search?q=ElasticSearch&spm=1001.2101.3001.7020