Elasticsearch

How search works 

=>

Source: Grokking Algorithms - An illustrated guide for programmers and other curious people

 Inverted Index

https://www.geeksforgeeks.org/inverted-index/

Data structure mainly used by search engines

https://lucene.apache.org/

Terminology

Analysis is the process of converting full text to terms.

ES stores documents represented as a JSON document

Each document is stored in an index and has a type and an id.

An index is like a table in a relational database. It has a mapping which contains a type, which contains the fields in the index.

A mapping is like a schema definition in a relational database.

A type used to represent the type of document, e.g. an email, a user, or a tweet. Types are deprecated and are in the process of being removed

Let's see :)

ES API

https://www.elastic.co/guide/en/elasticsearch/reference/current/docs.html

ES API

Index (save) API  -> http://localhost:9200/transactions/ron/1

REST like client :)

Mappig -> https://www.elastic.co/guide/en/elasticsearch/reference/current/mapping.html

ES API

http://localhost:9200/transactions/_search?q=message:ana

http://localhost:9200/transactions/_search?q=message:an

Search API

WE FOUND THE JOKE

Elasticsearch is not for search :)

ES API

http://localhost:9200/transactions/_termvectors/1

Query

Match query: match queries accept text/numerics/dates, analyzes them, and constructs a query.

Term query: The term query finds documents that contain the exact term specified in the inverted index

What is full search?

The operator dilemma @ ING

{
   "query":{
      "match":{
         "message":{
            "query":"are pere",
            "operator":"or"
         }
      }
   }
}

Query

Analizer: examines the text of fields and generates a token stream.

Tokenizer: break field data into lexical units, or tokens.

Query

Query String

{
  "query" : {
    "query_string": {
      "query": "mer per",
      "fields": [
        "message"
      ]
    }
  }
}

Takeaways

Elasticsearch is for search and is not a joke :)

Working with ES means:

         - define mappings toghetter with analizers and tokenizers

         - store/index documents

         - write a query to get documents

Personal takeaways

A lot of breaking changes in ES releases: not only APIs but also concepts

https://www.elastic.co/guide/en/elasticsearch/reference/6.0/breaking-changes-6.0.html

A lot of cool new thinks added :) -> SQL queries

Not a really good documentation

Deep concepts ->

https://www.elastic.co/guide/en/elasticsearch/guide/currentQuery DSL

https://www.elastic.co/guide/en/elasticsearch//reference/current/query-dsl.html

ES @ING