@gojob/ts-elasticsearch

NPM publish

To publish a new version of this package, you have to build the project before run npm publish.

Description

The purpose of this library is to provide a decorated class approch to use the elasticsearch module.

Installation

yarn add @gojob/ts-elasticsearch

Usage example

import { Field, Elasticsearch, Primary } from '@gojob/ts-elasticsearch';

@Index()
class User {
  @Primary()
  @Field({ enabled: false})
  id: string;

  @Field('text') name: string;
  @Field('integer') age: number;
}

// ...

const elasticsearch = new Elasticsearch({ host: 'http://192.168.99.100:9200' });

await elasticsearch.indices.create(User);
await elasticsearch.indices.putMapping(User);

const user = new User();
user.id = 'agent-1122';
user.name = 'Smith';
user.age = 33;

await elasticsearch.index(user);

await elasticsearch.indices.refresh(User);

const { documents } = elasticsearch.search(User, { body: { query: { match_all: {} } } });

Documentation

Decorators

decorator @Index

Keep in mind that index types are scheduled to be removed in Elasticsearch 8.
This class decorator factory declares a new Elasticsearch index.
By default, it uses the class name as index (and type), but it can be overwritten.
Index settings may be added in the settings optional parameter.

@Index({index: 'twt', settings: { number_of_shards: 3 } })
class Tweeter {}

decorator @Primary

This property decorator factory declares the class primary key which will be used as _id in Elasticsearch index. When not provided, Elasticsearch generates ids by it own.

@Index()
class User {
  @Primary()
  @Field({ enabled: false})
  id: string;

  @Field('text') name: string;
  @Field('integer') age: number;
}

decorator @Field

This property decorator factory declares a new property in the current class.
It takes either a simple type as string or an elasticsearch mapping setting.

@Index()
class User {
  @Field('keyword') id: string;
  @Field({ type: 'text', boost: 2 })name: string;
  @Field('integer') age: number;
}

Special case of object or nested (array of object)

When dealing with object or nested, you have to declare a class with some fields declared. Notice in this case, the class in not decorated with @Index.
Object example:

class Country {
  @Field('text') name: string;
}

@Index()
class City {
  @Field('text') name: string;
  @Field({ object: Country })
  country: Country;
}

Nested example:

class Follower {
  @Field({ enabled: false}) name: string;
  @Field({ enabled: false}) popularity: number;
}

@Index()
class User {
  @Primary()
  @Field({ enabled: false})
  id: string;

  @Field('text') name: string;

  @Field({ nested: Follower })
  followers: Follower[];
}

Elasticsearch class

This class provides main elasticsearch features directly usable with indexed classes.
The purpose of this library is not to override all official plugin features, but only those that are relevant for managing documents using classes.
When dealing with documents, you can either uses indexed class instances or literal object associated to their indexed class.
The examples presents in this chapter are based on this setup.

import { Field, Elasticsearch, Primary } from '@gojob/ts-elasticsearch';

const elasticsearch = new Elasticsearch({ host: 'http://192.168.99.100:9200' });

@Index()
class User {
  @Primary()
  @Field({ enabled: false})
  id: string;

  @Field('text') name: string;
  @Field('integer') age: number;
}

Elasticsearch core

Instanciate this class with an IConfigOptions or an official elasticsearch.Client instance.

IConfigOptions

IConfigOptions extends the official client configuration option.
Extended options provided by IConfigOptions:
| Name | Type | Optional | Description | |:-----------:|:---------------------------------------------------------------------------------------------------------:|:--------:|---------------------------------| | client | elasticsearch.Client | X | Official client instance to use | | indexPrefix | string | X | Prefix to set to all indices |

Core functions

The main class provides the main part of the document API functions.

• bulkIndex: Index an arary or a stream of documents (bulk api
• count: Count documents in the index (count api)
• create: Add a new document to the index create api)
• delete: Delete a document from the index delete api)
• get: Retrieve a document by its id get api)
• getIndices: Return all Indexed classes
• index: Add or update a document in the index index api)
• search: Search documents in the index search api)
• update: Update a document update api)

Elasticsearch.indices

The indices provides the main part of the document API functions.

• create: Create an Index (indice create api)
• delete: Delete an Index (indice delete api)
• exists: Check if an Index exists (indice exists api)
• flush: Flush an Index (indice flush api)
• putMapping: Put mapping of an Index (indice putmapping api)
• refresh: Refresh an Index (indice refresh api)