ElasticSearch NEST 术语查询不返回任何结果

Question

这是我的架构

[ElasticType(Name = "importFile")]
public class ImportFile : DocumentMapping
{
    [ElasticProperty(Store = false, Index = FieldIndexOption.NotAnalyzed)]
    public string FileName { get; set; }

    [ElasticProperty(Store = false, Index = FieldIndexOption.NotAnalyzed)]
    public string GroupId { get; set; }

    [ElasticProperty(Store = false, Index = FieldIndexOption.Analyzed)]
    public string FilePath { get; set; }
}

我做了一个像这样的 NEST 查询：

    var res = ElasticClient.Search<ImportFile>(s => s
        .Index(ElasticIndexName)
        .Filter(f =>
            f.Term(t => t.FileName, "Group-1.uhh"))).Documents.ToArray();

并返回零个元素！

如果我查看数据库内部（使用邮递员），我可以看到我的文档：

{
  "took": 5,
  "timed_out": false,
  "_shards": {
    "total": 2,
    "successful": 2,
    "failed": 0
  },
  "hits": {
    "total": 2,
    "max_score": 14.069489,
    "hits": [
      {
        "_index": "reviewer-bdd-test-index",
        "_type": "importFile",
        "_id": "AU9kUka2hr5Jg98UXOae",
        "_score": 14.069489,
        "_source": {
          "fileName": "Group-1.uhh",
          "groupId": "0ae1206d0644eabd82ae490e612732df5da2cd141fdee70dc64207f86c96094f",
          "filePath": ""
        }
      },
      {
        "_index": "reviewer-bdd-test-index",
        "_type": "importFile",
        "_id": "AU9kZO25hr5Jg98UXRnk",
        "_score": 14.069489,
        "_source": {
          "fileName": "group-1.uhh",
          "groupId": "0ae1206d0644eabd82ae490e612732df5da2cd141fdee70dc64207f86c96094f",
          "filePath": ""
        }
      }
    ]
  }
}

Answer 1

听起来您可能没有在索引文档之前明确地将类型的映射放入索引中，因此 Elasticsearch 已根据其文档中字段的默认映射推断出映射见过。例如，给定以下类型

[ElasticType(Name = "importFile")]
public class ImportFile
{
    [ElasticProperty(Store = false, Index = FieldIndexOption.NotAnalyzed)]
    public string FileName { get; set; }

    [ElasticProperty(Store = false, Index = FieldIndexOption.NotAnalyzed)]
    public string GroupId { get; set; }

    [ElasticProperty(Store = true, Index = FieldIndexOption.Analyzed)]
    public string FilePath { get; set; }
}

如果我们如下索引一些文档

void Main()
{
    var settings = new ConnectionSettings(new Uri("http://localhost:9200"));            
    var client = new ElasticClient(settings);

    client.Index<ImportFile>(
        new ImportFile{
            FileName = "Group-1.uhh",
            FilePath = "",
            GroupId = "0ae1206d0644eabd82ae490e612732df" + 
                      "5da2cd141fdee70dc64207f86c96094"
        },
        index => index
            .Index("reviewer-bdd-test-index")
            .Type("importFile")
            .Refresh());

    client.Index<ImportFile>(
        new ImportFile
        {
            FileName = "group-1.uhh",
            FilePath = "",
            GroupId = "0ae1206d0644eabd82ae490e612732df" + 
                      "5da2cd141fdee70dc64207f86c96094"
        },
        index => index
            .Index("reviewer-bdd-test-index")
            .Type("importFile")
            .Refresh());

    var results = client.Search<ImportFile>(s => s
                .Index("reviewer-bdd-test-index")
                .Type("importFile")
                .Query(q => q
                   .Filtered(fq => fq
                        .Filter(f => f
                            .Term(p => p.FileName, "Group-1.uhh")
                        )
                    )
                )
            );

    Console.WriteLine(string.Format("{0} {1}", results.RequestInformation.RequestMethod, results.RequestInformation.RequestUrl));
    Console.WriteLine(Encoding.UTF8.GetString(results.RequestInformation.Request)); 
    Console.WriteLine("Matching document count: {0}", results.Documents.Count());
}

以下是控制台的输出

POST http://localhost:9200/reviewer-bdd-test-index/importFile/_search
{
  "query": {
    "filtered": {
      "filter": {
        "term": {
          "fileName": "Group-1.uhh"
        }
      }
    }
  }
}
Matching document count: 0

我们没有得到匹配的文件。使用

检查 Elasticsearch 中的映射

curl -XGET "http://localhost:9200/reviewer-bdd-test-index/_mapping"

我们看到importFile 类型的映射是

{
   "reviewer-bdd-test-index": {
      "mappings": {
         "importFile": {
            "properties": {
               "fileName": {
                  "type": "string"
               },
               "groupId": {
                  "type": "string"
               }
            }
         }
      }
   }
}

这不是我们所期望的； fileName 和 groupId 也应该有 "index": "not_analyzed" 和 filePath 甚至不在映射中。这两者都是因为 Elasticsearch 已经根据它传递的文档推断出映射 - fileName 和 groupId 已被映射为字符串类型，并将使用标准分析器进行分析，我相信 em> filePath 尚未映射，因为两个看到的文档的字段都有一个空字符串值，因此应用于该字段的 standard analyzer 不会为倒排索引生成任何标记，因此字段不包含在映射中。

因此，为了确保事情按预期工作，我们需要在索引任何文档之前添加到索引的映射

void Main()
{
    var settings = new ConnectionSettings(new Uri("http://localhost:9200"));
    var client = new ElasticClient(settings);

    // Add the mapping for ImportFile to the index
    client.CreateIndex(indexSelector => indexSelector
        .Index("reviewer-bdd-test-index")
        .AddMapping<ImportFile>(mapping => mapping
            .MapFromAttributes()
        )
    );

    // ... Same as above after this point
}

结果

POST http://localhost:9200/reviewer-bdd-test-index/importFile/_search
{
  "query": {
    "filtered": {
      "filter": {
        "term": {
          "fileName": "Group-1.uhh"
        }
      }
    }
  }
}
Matching document count: 1

成功！我们有一个匹配的文档。检查 Elasticsearch 中的映射会产生我们预期的结果

{
   "reviewer-bdd-test-index": {
      "mappings": {
         "importFile": {
            "properties": {
               "fileName": {
                  "type": "string",
                  "index": "not_analyzed"
               },
               "filePath": {
                  "type": "string",
                  "store": true
               },
               "groupId": {
                  "type": "string",
                  "index": "not_analyzed"
               }
            }
         }
      }
   }
}

另外，属性映射可以替换为流利的映射

var indexResult = client.CreateIndex(indexDescriptor => indexDescriptor
    .Index("reviewer-bdd-test-index")
    .AddMapping<ImportFile>(mapping => mapping
        .Type("importFile")
        .MapFromAttributes()
        .Properties(properties => properties
            .String(s => s
                .Name(file => file.FileName)
                .Store(false)
                .Index(FieldIndexOption.NotAnalyzed))
            .String(s => s
                .Name(file => file.GroupId)
                .Store(false)
                .Index(FieldIndexOption.NotAnalyzed))
            .String(s => s
                .Name(file => file.FilePath)
                .Store(true))
        )
    )
);

此时无论是属性映射还是流畅映射都可以，但是有些事情只能通过流畅映射来实现，例如 multi_fields。