spring - Spring Elastic Search存储库将不存储数据

Question

我遇到了一个非常奇怪的问题，我基本上是在尝试使用springboot和存储库在 Elasticsearch 中存储对象，并且由于某些原因，我的对象从未持久化。

这是使用scala完成的

案例类:

 @Document(collection = "SPECTRUM")
      @org.springframework.data.elasticsearch.annotations.Document(indexName = "spectrum", `type` = "spectrum", shards = 1, replicas = 0, refreshInterval = "-1")
  case class Spectrum(
                   @(Field@field)(`type` = FieldType.Nested)
                   biologicalCompound: Compound,
                   @(Field@field)(`type` = FieldType.Nested)
                   chemicalCompound: Compound,
                   @(Field@field)(`type` = FieldType.Nested)
                   predictedCompound: Compound,
                   @(Indexed@field)
                   deleted: Boolean,
                   @(Id@field)
                   id: String,
                   lastUpdated: String,
                   @(Field@field)(`type` = FieldType.Nested)
                   metaData: Array[MetaData],
                   score: Score,
                   spectrum: String,
                   splash: Splash,
                   submitter: Submitter,
                   @(Field@field)(`type` = FieldType.Nested)
                   tags: Array[Tags],
                   @(Field@field)(`type` = FieldType.Nested)
                   authors: Array[Author]
                 )

}

存储库

@Repository("spectrumElasticRepository")
trait ISpectrumElasticRepositoryCustom  extends ElasticsearchRepository[Spectrum, String]  with     SpectrumElasticRepositoryCustom{    
  def findByBiologicalCompoundInchiKey(inchiKey: String) : java.util.List[Spectrum]
}

测试代码

  getRepository.deleteAll()
  assert(getRepository.count() == 0)

      s"we should be able to store our data" in {
        for (spectrum <- exampleRecords) {
          val result = getRepository.save(spectrum)
          assert(result.isInstanceOf[Spectrum])
        }

    assert(getRepository.count() == 58)

        val data:Iterable[Spectrum] = getRepository.findAll()
      }

一旦测试代码达到这一行

            val data:Iterable[Spectrum] = getRepository.findAll()

它导致以下异常

failed to map source [ {}] to class Spectrum
org.springframework.data.elasticsearch.ElasticsearchException: failed to map source [ {}] to class Spectrum

并直接查看 Elasticsearch 服务器

localhost:9200/spectrum/_search

似乎没有任何报告的匹配实际附加任何数据

  {
    "_index": "spectrum",
    "_type": "spectrum",
    "_id": "AVNhcpHjnm4IHnHomcXj",
    "_score": 1,
    "_source": {}
  },

因为源是空的。

任何想法是什么原因造成的？

Answer 1

这似乎与scala有关。

一旦确定了要存储在 Elasticsearch 数据库中的每个属性都将用

@BeanProperty

像这儿

  @Document(collection = "SPECTRUM")
  @org.springframework.data.elasticsearch.annotations.Document(indexName = "spectrum", `type` = "spectra", shards = 1, replicas = 0, refreshInterval = "-1")
  case class Spectrum(
                   @BeanProperty
                   @(Field@field)(`type` = FieldType.Nested)
                   biologicalCompound: Compound,
                   @BeanProperty
                   @(Field@field)(`type` = FieldType.Nested)
                   chemicalCompound: Compound,
                   @BeanProperty
                   @(Field@field)(`type` = FieldType.Nested)
                   predictedCompound: Compound,
                   @BeanProperty
                   @(Indexed@field)
                   deleted: Boolean,
                   @BeanProperty
                   @(Id@field)
                   id: String,
                   lastUpdated: String,
                   @BeanProperty
                   @(Field@field)(`type` = FieldType.Nested)
                   metaData: Array[MetaData],
                   @BeanProperty
                   score: Score,
                   @BeanProperty
                   spectrum: String,
                   @BeanProperty
                   splash: Splash,
                   @BeanProperty
                   submitter: Submitter,
                   @BeanProperty
                   @(Field@field)(`type` = FieldType.Nested)
                   tags: Array[Tags],
                   @(Field@field)(`type` = FieldType.Nested)
                   authors: Array[Author]
                     )

}

它可以正常工作并保留该对象。

显然，这太丑陋了，它使您想知道是否还有更好的选择，或者编写一个可识别scala的包装器是否有意义。

或者，您可以选择自己的一种覆盖使用的对象映射器和实体impl。这可以在您的配置类中完成。

  @Autowired
  val objectMapper:ObjectMapper = null

  @Bean
  def elasticsearchTemplate: ElasticsearchOperations = {
    new ElasticsearchTemplate(client,new EntityMapperImpl(objectMapper))
  }

  @Bean
  def client: Client = {
    val client = new TransportClient()
    val address = new InetSocketTransportAddress(hostname, port)
    client.addTransportAddress(address)

    client
  }

  class EntityMapperImpl(val mapper: ObjectMapper) extends EntityMapper {

    override def mapToString(`object`: scala.Any): String =     mapper.writeValueAsString(`object`)

    override def mapToObject[T](source: String, clazz: Class[T]): T =     mapper.readValue(source, clazz)
  }

对象映射器的确切配置如下所示

val mapper = new ObjectMapper() with ScalaObjectMapper

mapper.registerModule(DefaultScalaModule)

//required, in case we are provided with a list of value
mapper.enable(DeserializationFeature.ACCEPT_SINGLE_VALUE_AS_ARRAY)
mapper.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false)
mapper.setSerializationInclusion(Include.NON_NULL);
mapper

只需要在某处注册为bean