elasticsearch模糊搜索，搜索建议（suggest），自动补全

首先构建mapping，并制定字段解析器。解析器采用中文ik_max_word类型，es默认不支持，需要自己下载解析包扩展。中文解析包下载地址： https://github.com/medcl/elasticsearch-analysis-ik/releases

加载elasticsearch对应的解析包并配置到elaticsearch/plugins/ik,ik文件夹需要自己创建，创建完后将对应下载包copy到文件夹下即可，然后重启es才会生效，重启控制台会加载输出ik相关字眼的log。

 

配置完es以后构建es mapping字段

首先导入相关组件并构建解析器

# -*- coding: utf-8 -*- 
# @Time : 2020/4/1 8:59 AM 
# @Author : wywinstonwy
# #@desc:

from datetime import datetime
from elasticsearch_dsl import Document, Date, Integer, Keyword, Text, connections,Search,Completion
from elasticsearch import Elasticsearch
from elasticsearch_dsl import analyzer, tokenizer
from elasticsearch_dsl.analysis import CustomAnalyzer as _CustomAnalyzer
# Define a default Elasticsearch client
connections.create_connection(hosts=['localhost'])
my_analyzer = analyzer('my_analyzer',
    tokenizer=tokenizer('trigram', 'nGram', min_gram=2, max_gram=3),
    filter=['lowercase']
)
# my_analyzer = analyzer('my_analyzer',
#     tokenizer=tokenizer('ik_max_word',  min_gram=2, max_gram=3),
#     filter=['lowercase']
# )

class customAnalyzer(_CustomAnalyzer):
    def get_analysis_definition(self):
        return {}

ik_analyzer =customAnalyzer('ik_max_word',filter=['lowercase'])

 

其次是构建mapping字端，并指定对应解析器，suggest字端为搜索建议补全字段。

class Article(Document):
    title = Text(analyzer='snowball', fields={'raw': Keyword()})
    suggest =Completion(analyzer=ik_analyzer)
    content = Text(analyzer='snowball')
    category = Keyword()
    publish_time =Date()
    published_from = Date()
    lines = Integer()
    author_name = Keyword()
    url = Keyword
    objhashurl = Keyword

    class Index:
        name = 'cnblog'
        _type ='article'
        settings = {
          "number_of_shards": 2,
          "number_of_replicas": 1

        }

    def save(self, ** kwargs):
        self.lines = len(self.content.split())
        return super(Article, self).save(** kwargs)

    def is_published(self):
        return datetime.now() > self.publish_time

GET cnblog/_search
{
"query": {
"bool": {
"must": [
{
"match_all": { }
}
],
"must_not": [ ],
"should": [ ]
}
},
"from": 0,
"size": 10000,
"sort": [ ],
"aggs": { }
}
#搜索
GET cnblog/_search
{
  "query": {
    "fuzzy": {
      "title": "java"
    }
  }
  , "_source": ["title"]
}

#fuzzy模糊搜索
GET cnblog/_search
{
  "query": {
    "fuzzy": {
      "title": {
        "value": "jav",
        "fuzziness": 2,
        "prefix_length": 2
      }
    }
  }
  , "_source": "title"
 
}
#建议搜索补全
GET cnblog/_search
{
  "suggest": {
    "suggest": {
      "prefix":"jaba",
      "text": "jaba",
      "completion": {
        "field": "suggest",
        "fuzzy":{
          "fuzziness":1
        }
      }
    }
  }
  , "_source": ["title","category"]
}

 

fuzziness 为编辑距离，可以设置适合的大小范围。

 

最后通过接口实现查询：

def suggest():
    keyword ='jav'
    response1 = client.search(
        index='cnblog',
        body={
            "suggest": {
                "suggest": {
                    "text": keyword,
                    "completion": {
                        "field": "suggest",
                        "fuzzy": {
                            "fuzziness": 1
                        }
                    }
                }
            }
            , "_source": ["title", "category"]
        }
    )

    return response1