elasticsearch_.net_client_nest2.x_到_5.x常用方法属性差异

       Elasticsearch.Net、NEST 交流群:523061899

变更1 字段类型string

  • 2.x中仅有String类型,可设置是否分词、是否索引及分词使用的分词器。
  • 5.x中拆分为Keyword类型与Text类型,Keyword 不分词,可设置是否索引;Text分词,可设置是否索引及分词使用的分词器。
            [String(Name = "Name",Index = FieldIndexOption.NotAnalyzed)]
            public string Name { get; set; }
     
            [String(Name = "Dic", Index = FieldIndexOption.Analyzed,Analyzer = "ik_max_word")]
     
     
     
            /// <summary>
            /// keyword 不分词
            /// </summary>
            [Keyword(Name = "Name", Index = true)]
            public string Name { get; set; }
     
            /// <summary>
            /// text 分词,Analyzer = "ik_max_word"
            /// </summary>
            [Text(Name = "Dic", Index = true, Analyzer = "ik_max_word")]
            public string Dic { get; set; }

     

    注:5.x已将string标记为过时特性,下个版本将移除;

    变更2 滚动扫描获取全量数据

    滚动扫描支持并行执行提高获取数据效率;

    2.x

    public void SearchScanScroll(string indexName)
            {
                string scrollid = "";
                var result = _client.Search<TestModel2>(s => s.Index(indexName).Query(q => q.MatchAll())
                    .Size(10)
                    //todo:2x中的SearchType.Scan,5.x移除
                    .SearchType(SearchType.Scan)
                    .Scroll("1m"));
                scrollid = result.ScrollId;
                int count = 0;
                while (true)
                {
                    //得到滚动扫描的id
                    //执行滚动扫描得到数据 返回数据量是 result.Shards.Successful*size(查询成功的分片数*size)
                    var result1 = _client.Scroll<TestModel2>("1m", scrollid);
                    if(result1.Documents==null || !result1.Documents.Any())
                        break;
                    foreach (var info in result1.Documents)
                    {
                        count++;
                     Console.WriteLine(info.Id+" - "+ count);   
                    }
                    //得到新的id
                    scrollid = result1.ScrollId;
                }
            }

    5.x

    Action<int> sc1 = (id) =>
                {
                    string scrollid = "";
                    //todo:5.x 多了Slice设置 移除SearchType.Scan
                    var result = _client.Search<TestModel5>(s => s.Index(indexName).Query(q => q.MatchAll())
                        .Size(15)
                        .Sort(st=>st.Descending(ds=>ds.Id))
                        .Scroll("1m")
                        //id从0开始 0,1,2...
                        //length=max
                        //例:max=3 id=0,id=1,id=2
                        .Slice(sl => sl.Id(id).Max(3))
                        );
     
                    //得到滚动扫描的id
                    scrollid = result.ScrollId;
                    foreach (var info in result.Documents)
                    {
                        Console.WriteLine(info.Id + " - " + " -批次count " + result.Documents.Count + " - 线程"+Thread.CurrentThread.ManagedThreadId);
                    }
                    while (true)
                    {
                        //执行滚动扫描得到数据 返回数据量是 result.Shards.Successful*size(查询成功的分片数*size)
                        var result1 = _client.Scroll<TestModel5>("1m", scrollid);
                        if (result1.Documents == null || !result1.Documents.Any())
                            break;
                        foreach (var info in result1.Documents)
                        {
                            Console.WriteLine(info.Id + " - " +" -批次count "+ result1.Documents.Count+ " - 线程" + Thread.CurrentThread.ManagedThreadId);
                        }
                        //得到新的id
                        scrollid = result1.ScrollId;
                    }
                };
               var t1= Task.Factory.StartNew(() => { sc1(0); });
                var t2= Task.Factory.StartNew(() => { sc1(1); });
                var t3= Task.Factory.StartNew(() => { sc1(2); });
                t1.Wait();
                t2.Wait();
                t3.Wait();

    注:5.x 增加了Slice设置可并行拉取全量数据,移除SearchType.Scan,首次查询Search时也返回结果,2.x中设置SearchType.Scan首次Search只返回ScrollId。

    注:5.x中新增深度分页方案深度分页

    分组结果对象 Nest.KeyedBucket

    • 2.x KeyedBucket
    • 5.x KeyedBucket<object>

    聚合结果解析

    2.x

               var vendorIdGroup = (BucketAggregate)result.Aggregations["Group_Group"];
                foreach (var bucket1 in vendorIdGroup.Items)
                {
                    //todo:2.x KeyedBucket 
                    var bucket = (KeyedBucket)bucket1;
                    var maxPrice = ((ValueAggregate)bucket.Aggregations["Dvalue_Max"]).Value;
                }

    5.x

                var vendorIdGroup = (BucketAggregate)result.Aggregations["Group_Group"];
                foreach (var bucket1 in vendorIdGroup.Items)
                {
                    //todo:5.x KeyedBucket<T> 
                    var bucket = (KeyedBucket<object>) bucket1;
                    var maxPrice = ((ValueAggregate)bucket.Aggregations["Dvalue_Max"]).Value;
                }

    其它

    一些新的特性或接口会持续补充。

posted @ 2017-09-13 18:07  Mr. Hu  阅读(1437)  评论(0编辑  收藏  举报
Map