ES操作纪要

我们的ES是5.5版本，所以参考看的文档是这个：https://www.elastic.co/guide/en/elasticsearch/reference/5.5/index.html

1. 创建索引、创建mapping映射、设置分词

PUT tangxue_test_index_20201222
{
    "settings":{
        "index":{
            "number_of_shards":"3",
            "number_of_replicas":0,
            "refresh_interval":"1"
        },
        "analysis":{
            "analyzer":{
                "ngram_analyzer":{
                    "tokenizer":"my_tokenizer"
                },
                "ik_unsmart_analyzer":{
                    "filter":[
                        "lowercase"
                    ],
                    "type":"custom",
                    "tokenizer":"ik_max_word"
                }
            },
            "tokenizer":{
                "my_tokenizer":{
                    "filter":[
                        "lowercase"
                    ],
                    "type":"ngram",
                    "min_gram":1,
                    "max_gram":3,
                    "token_chars":[
                        "letter",
                        "digit"
                    ]
                }
            }
        }
    },
    "mappings":{
        "medicalrecord":{
            "properties":{
                "fullFieldName":{
                    "type":"keyword",
                    "fields":{
                        "ngramFullFieldName":{
                            "type":"text",
                            "analyzer":"ngram_analyzer"
                        },
                        "ikFullFieldName":{
                            "type":"text",
                            "analyzer":"ik_unsmart_analyzer"
                        }
                    }
                }
            }
        }
    }
}

image.png

2. 获取索引数据

GET tangxue_test_index_20201222

获取结果为：

{
  "tangxue_test_index_20201222": {
    "aliases": {},
    "mappings": {
      "medicalrecord": {
        "properties": {
          "fullFieldName": {
            "type": "keyword",
            "fields": {
              "ikFullFieldName": {
                "type": "text",
                "analyzer": "ik_unsmart_analyzer"
              },
              "ngramFullFieldName": {
                "type": "text",
                "analyzer": "ngram_analyzer"
              }
            }
          }
        }
      }
    },
    "settings": {
      "index": {
        "refresh_interval": "-1",
        "number_of_shards": "3",
        "provided_name": "tangxue_test_index_20201222",
        "creation_date": "1608606744812",
        "analysis": {
          "analyzer": {
            "ik_unsmart_analyzer": {
              "filter": [
                "lowercase"
              ],
              "type": "custom",
              "tokenizer": "ik_max_word"
            },
            "ngram_analyzer": {
              "tokenizer": "my_tokenizer"
            }
          },
          "tokenizer": {
            "my_tokenizer": {
              "filter": [
                "lowercase"
              ],
              "token_chars": [
                "letter",
                "digit"
              ],
              "min_gram": "1",
              "type": "ngram",
              "max_gram": "3"
            }
          }
        },
        "number_of_replicas": "0",
        "uuid": "-QzDmvr1RiO4Ce44RKom7A",
        "version": {
          "created": "5050399"
        }
      }
    }
  }
}

3. 创建文档

POST tangxue_test_index_20201222/medicalrecord
{
  "fullFieldName":"姓名"
}
POST tangxue_test_index_20201222/medicalrecord
{
  "fullFieldName":"姓"
}
POST tangxue_test_index_20201222/medicalrecord
{
  "fullFieldName":"名"
}

4. 搜索文档（细写细读，也就是写入和读取用的分词一样）

GET tangxue_test_index_20201222/_search
{
  "query" : {
    "bool" : {
      "should" : [
        {
          "match": {
            "fullFieldName.ngramFullFieldName": "姓名"
          }
        }
      ],
      "disable_coord" : false,
      "adjust_pure_negative" : true,
      "boost" : 1.0
    }
  }
}

返回结果：

{
  "took": 1,
  "timed_out": false,
  "_shards": {
    "total": 3,
    "successful": 3,
    "failed": 0
  },
  "hits": {
    "total": 7,
    "max_score": 4.9663877,
    "hits": [
      {
        "_index": "tangxue_test_index_20201222",
        "_type": "medicalrecord",
        "_id": "AXaIjKHEsDP0SDHXyJsa",
        "_score": 4.9663877,
        "_source": {
          "fullFieldName": "姓名"
        }
      },
      {
        "_index": "tangxue_test_index_20201222",
        "_type": "medicalrecord",
        "_id": "AXaIlF0xsDP0SDHXyJse",
        "_score": 1.5325457,
        "_source": {
          "fullFieldName": "名"
        }
      },
      {
        "_index": "tangxue_test_index_20201222",
        "_type": "medicalrecord",
        "_id": "AXaIlCH-sDP0SDHXyJsd",
        "_score": 0.44839138,
        "_source": {
          "fullFieldName": "姓"
        }
      }
      }
    ]
  }
}

5. 修改_settings

POST tangxue_test_index_20201222/_close

PUT tangxue_test_index_20201222/_settings
{
  "analysis":{
      "analyzer":{
          "ngram_analyzer":{
              "tokenizer":"my_tokenizer"
          },
          "ik_unsmart_analyzer":{
              "filter":[
                  "lowercase"
              ],
              "type":"custom",
              "tokenizer":"ik_max_word"
          }
      },
      "tokenizer":{
          "my_tokenizer":{
              "filter":[
                  "lowercase"
              ],
              "type":"ngram",
              "min_gram":2,
              "max_gram":3,
              "token_chars":[
                  "letter",
                  "digit"
              ]
          }
      }
  }
}

POST tangxue_test_index_20201222/_open

6.增加Mappings实现细写粗读

一般情况，索引分词（写）应该按照最细力度分词，搜索分词（读）可按照最粗力度分词，即所谓的细写粗读

修改Settings

PUT tangxue_test_index_20201222/_settings
{
  "analysis":{
      "analyzer":{
          "ngram_analyzer1":{
              "tokenizer":"my_tokenizer1"
          },
          "ngram_analyzer2":{
              "tokenizer":"my_tokenizer2"
          },
          "ik_unsmart_analyzer":{
              "filter":[
                  "lowercase"
              ],
              "type":"custom",
              "tokenizer":"ik_max_word"
          }
      },
      "tokenizer":{
          "my_tokenizer1":{
              "filter":[
                  "lowercase"
              ],
              "type":"ngram",
              "min_gram":1,
              "max_gram":3,
              "token_chars":[
                  "letter",
                  "digit"
              ]
          },
          "my_tokenizer2":{
              "filter":[
                  "lowercase"
              ],
              "type":"ngram",
              "min_gram":2,
              "max_gram":3,
              "token_chars":[
                  "letter",
                  "digit"
              ]
          }
      }
  }
}

增加Mappings

PUT tangxue_test_index_20201222/_mapping/medicalrecord?update_all_types
{
  "properties": {
    "fullFieldName":{
        "type":"keyword",
        "fields":{
            "ngramFullFieldName_new":{
                "type":"text",
                "analyzer":"ngram_analyzer1",
                "search_analyzer":"ngram_analyzer2"
            }
        }
    }
  }
}

写入数据

POST tangxue_test_index_20201222/medicalrecord
{
  "fullFieldName":"姓名"
}
POST tangxue_test_index_20201222/medicalrecord
{
  "fullFieldName":"姓"
}
POST tangxue_test_index_20201222/medicalrecord
{
  "fullFieldName":"姓名tang"
}

读取数据，注意要用新的mapping字段ngramFullFieldName_new

GET tangxue_test_index_20201222/_search
{
  "query" : {
    "bool" : {
      "should" : [
        {
          "match": {
            "fullFieldName.ngramFullFieldName_new": "姓名"
          }
        }
      ],
      "disable_coord" : false,
      "adjust_pure_negative" : true,
      "boost" : 1.0
    }
  }
}

读取结果，会发现返回结果是没有单独“姓”这个返回的

{
  "took": 1,
  "timed_out": false,
  "_shards": {
    "total": 3,
    "successful": 3,
    "failed": 0
  },
  "hits": {
    "total": 2,
    "max_score": 0.49191087,
    "hits": [
      {
        "_index": "tangxue_test_index_20201222",
        "_type": "medicalrecord",
        "_id": "AXaI8FkysDP0SDHXyJsp",
        "_score": 0.49191087,
        "_source": {
          "fullFieldName": "姓名"
        }
      },
      {
        "_index": "tangxue_test_index_20201222",
        "_type": "medicalrecord",
        "_id": "AXaI9ewmsDP0SDHXyJsr",
        "_score": 0.28004453,
        "_source": {
          "fullFieldName": "姓名tang"
        }
      }
    ]
  }
}

7. 查看特定分词器下的分词结果

GET tangxue_test_index_20201222/_analyze
{
  "analyzer":"ngram_analyzer1",
  "text":"姓名美美 ！*1"
}

结果为：

{
  "tokens": [
    {
      "token": "姓",
      "start_offset": 0,
      "end_offset": 1,
      "type": "word",
      "position": 0
    },
    {
      "token": "姓名",
      "start_offset": 0,
      "end_offset": 2,
      "type": "word",
      "position": 1
    },
    {
      "token": "姓名美",
      "start_offset": 0,
      "end_offset": 3,
      "type": "word",
      "position": 2
    },
    {
      "token": "名",
      "start_offset": 1,
      "end_offset": 2,
      "type": "word",
      "position": 3
    },
    {
      "token": "名美",
      "start_offset": 1,
      "end_offset": 3,
      "type": "word",
      "position": 4
    },
    {
      "token": "名美美",
      "start_offset": 1,
      "end_offset": 4,
      "type": "word",
      "position": 5
    },
    {
      "token": "美",
      "start_offset": 2,
      "end_offset": 3,
      "type": "word",
      "position": 6
    },
    {
      "token": "美美",
      "start_offset": 2,
      "end_offset": 4,
      "type": "word",
      "position": 7
    },
    {
      "token": "美",
      "start_offset": 3,
      "end_offset": 4,
      "type": "word",
      "position": 8
    },
    {
      "token": "1",
      "start_offset": 7,
      "end_offset": 8,
      "type": "word",
      "position": 9
    }
  ]
}

8. 查看所有索引详细数据

GET /_cat/indices?v

9. 查看某一个数据的信息（举例比如通过字段 “_id” 获取）

GET tangxue_test_index_20201222/_search
{
  "query": {
    "term": {
      "_id": {
        "value": "AXapehbzsDP0SDHXyJyz"
      }
    }
  }
}

结果为：

{
  "took": 1,
  "timed_out": false,
  "_shards": {
    "total": 3,
    "successful": 3,
    "failed": 0
  },
  "hits": {
    "total": 1,
    "max_score": 1,
    "hits": [
      {
        "_index": "tangxue_test_index_20201222",
        "_type": "medicalrecord",
        "_id": "AXapehbzsDP0SDHXyJyz",
        "_score": 1,
        "_source": {
          "fullFieldName": "美美姓名思"
        }
      }
    ]
  }
}

10. 清除ES某个索引的缓存

POST /tangxue_test_index_20201222/_cache/clear