Elasticsearch learning portal (B): a simple learning portal

A, ES concepts

Index
_inde, document storage areas, similar to the relational data in the database.
In fact, data is stored in the slice and the index, the index is just a fragment of one or more logical space grouped together. Index names must be all lowercase, not allowed to begin with an underscore, can not contain commas.
Document
document id is a string, when combined with _index, you can uniquely identify a document in the ElasticSearch. When you create a document, you can customize _id, but also allows automatic generation of ES help.

Second, the use postman learning portal

• Create index
  creates an index database format: Use the put request: ip: 9200 / {index}
  
  red box which is designated type before ES7, now is not necessary to use the default _doc for the type, see some It says the document type will be completely removed in 8.x inside.

{
     "settings":{
        "number_of_shards":3,
        "number_of_replicas":1
    },
    "mappings": {
        
        "properties": {
            "id": {
                "type": "long",
                "store":true
            },
            "title": {
                "type": "text",
                "store":true,
                "index":true,
                "analyzer": "standard"
            },
            "content": {
                "type": "text",
                "store":true,
                "index":true,
                "analyzer": "standard"
            }
        }
        
}
}

number_of_shards is used to set the number of fragments, number_of_replicas is used to set the number of copies. Add your own fields in the properties inside and specify the type of property

• Remove the index
  format: DELETE IP: 9200 / {index }
• Creating a document and modify documents
  * Insert Document Format: PUT (POST) IP: 9200 / {index} / _ doc / id specified document ID
  the POST the IP: 9200 / {index} / _ random ID * DOC
  
  
• Modify the document
  * Format: POST IP: 9200 / {index } / _ doc / Document ID *
  in fact and almost new, is to delete the original in the new.
• Deleting a document
  * Format: DELETE IP: 9200 / {index } / _ doc / document ID *
  

• Archie
  * Format: GET IP: 9200 / {index } / _ doc / ID document
  format: POSTIP: 9200 / {index} / _ doc / _search *
  

  Third, the tokenizer

• Standard tokenizer
  
  can not see a good standard word processing Chinese pair

• IK word breaker
  1. Download https://github.com/medcl/elasticsearch-analysis-ik/releases
  2. In the / plugins below to create a folder named ik, then downloaded via ftp pass word ik folder, and extract
  3 restart ES
  the IK word has two algorithms, ik_smart (split coarse-grained) and ik_max_word (granular split)
  
  
  custom dictionary
  needs to be configured inside the extended IKAnalyzer.cfg.xml dictionary is OK
  but the dictionary file format format utf-8 (do not select utf-8 BOM)