@codemask-labs/nestjs-elasticsearch

NestJS Elasticsearch Module

Introduction

Welcome to Nestjs Elasticsearch module based on @nestjs/elasticsearch package.

Motive

This package originates from our experience with using Elasticsearch in production environments, which leaded to maintenance issues when extensively used aggregations, searches and filters (especially aggregations).

The main issues we encountered and which our package fixes are:

1. Current Elasticsearch NestJS Module does not provide autocompletion for queries.
2. Elasticsearch response forgets about types of aggregations.
3. Since Elasticsearch indexes can be schema-less we got no proper feedback about what fields we should expect on the index.
4. Writing utility methods for all filters and aggregations queries caused a lot of boilerplate code in each project.

Features

• Quick Setup: Get up and running in minutes using our easy-to-understand API.

• Developer Experience: Designed with developers in mind, package prioritizes ease of use and efficiency throughout the development process.

• Full TypeScript Support: Enjoy the benefits of code autocompletion and types for both request and response objects. ⁤⁤Unlike the original Elasticsearch library, this package provides full type definitions in order to provide better development experience and minimize runtime errors. ⁤

• Utility Methods: Say goodbye to repetitive boilerplate code. The package offers set of utility methods for most common Elasticsearch filtering, sorting, pagination and aggregations use cases.

• Schema definitions: Elasticsearch index itself may be schema-less, but we integrated schema definitions into our package. Each schema maps to an Elasticsearch index. This gives us clear data model, which can be used when building request object, ensuring only fields available for given index are used. These definitions are also used to register indexes in module scope and inject them into a service, similar to how it is approached in TypeORM NestJS module.

Instalation

You can install package using yarn or npm

$ yarn add @codemask-labs/nestjs-elasticsearch
// or
$ npm i @codemask-labs/nestjs-elasticsearch

Getting started

Once the package is installed, you can start with importing the ElasticsearchModule into the AppModule.

import { ElasticsearchModule } from '@codemask-labs/nestjs-elasticsearch'

@Module({
    imports: [
        ElasticsearchModule.register({
            node: 'http://localhost:9200'
        })
    ]
})
class AppModule {}

The register() method supports all the configuration properties available in ClientOptions from the @elastic/elasticsearch package.

Registering the index

You can define index schema with @RegisterIndex() decorator.

import { RegisterIndex } from '@codemask-labs/nestjs-elasticsearch'

@RegisterIndex('examples')
export class ExampleDocument {
    readonly id: string
    readonly exampleField: string
    readonly exampleNumericField: number
}

Add index to a module

The ElasticsearchModule provides the forFeature() method to configure the module and define which indexes should be registered in the current scope.

import { ElasticsearchModule } from '@codemask-labs/nestjs-elasticsearch'
import { ExampleDocument } from './example.document'

@Module({
    imports: [ElasticsearchModule.forFeature([ExampleDocument])],
    providers: [ExampleService]
})
export class ExampleModule {}

Inject index in a service

With module configuration in place, we can inject the ExampleDocument into the ExampleService using the @InjectIndex() decorator:

import { Injectable } from '@nestjs/common'
import { Index } from '@codemask-labs/nestjs-elasticsearch'
import { ExampleDocument } from './example.document'

@Injectable()
export class ExampleService {
    @InjectIndex(ExampleDocument)
    private readonly exampleIndex: Index<ExampleDocument>

    getExampleDocuments() {
        return this.exampleIndex.search()
    }
}

Now you can start creating request to Elasticsearch.

Api Reference

Once you finish the Getting Started guide, you can start building Elasticsearch request objects.

You can put request object directly in the search() method

import { getBoolQuery, getTermQuery, Order } from '@codemask-labs/nestjs-elasticsearch'

getExampleDocuments() {
    return this.exampleIndex.search({
        size: 10,
        query: getBoolQuery({
            must: [
                getTermQuery('exampleField.keyword', 'Some value'),
                getRangeQuery('exampleNumericField', {
                    gte: 1,
                    lte: 10
                })
            ]
        }),
        sort: {
            'exampleField.keyword': {
                order: Order.ASC
            }
        }
    })
}

or use getSearchRequest() method if you want to move request creation to some other place, but still laverage full type support and autocompletion.

import {
    getBoolQuery,
    getTermQuery,
    getSearchRequest,
    Order
} from '@codemask-labs/nestjs-elasticsearch'
import { ExampleDocument } from './example.document'

const searchRequestBody = getSearchRequest(ExampleDocument, {
    size: 10,
    query: getBoolQuery({
        must: [
            getTermQuery('exampleField.keyword', 'Some value'),
            getRangeQuery('exampleNumericField', {
                gte: 1,
                lte: 10
            })
        ]
    }),
    sort: {
        'exampleField.keyword': {
            order: Order.ASC
        }
    }
})

Filtering

As for now the package provides utils for the following filter queries:

Boolean query

Use getBoolQuery() for Boolean query and its most common occurrence types:

• getMustNotQuery() for Must not query
• getShouldQuery() for Should query
• getMustQuery() for Must query

Together with that you can also use getMinimumShouldMatchParameter() for minimum_should_match parameter

Full text queries

• getMatchPhrasePrefixQuery() for Match phrase prefix query
• getMatchQuery() for Match query

Term-level queries

• getTermQuery() for Term query
• getTermsQuery() for Terms query
• getRangeQuery() for Range query
• getExistsQuery for Exists query

Aggregations

As for now the package provides utils for the following aggregation queries:

Bucket aggregations

• getCompositeAggregation() for Composite aggregation
• getDateHistogramAggregation() for Date histogram aggregation
• getMissingValueAggregation() for Missing aggregation
• getHistogramAggregation() Histogram aggregation
• getRangeAggregation() for Range aggregation
• getTermsAggregation() for Terms aggregation

Metrics aggregations:

• getAvgAggregation() for Avg aggregation
• getCardinalityAggregation() for Cardinality aggregation
• getMaxAggregation() for Max aggregation
• getMinAggregation() for Min aggregation
• getPercentileAggregation() for Percentiles aggregation
• getSumAggregation() for Sum aggregation
• getTopHitsAggregation() for Top hits aggregation
• getValueCountAggregation() for Value count aggregation

Pipiline aggregations

• getStatsBucketAggregation() for Stats bucket aggregation

License

MIT