Skip to content

kontent-ai/delivery-sdk-js

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

npm version Build npm Known Vulnerabilities GitHub license Gzip browser bundle Discord

JavaScript Delivery SDK Documentation

This library is build for fetching Kontent.ai data with Delivery API. Works in browser & node.

Tip

It's highly recommended to use this library in combination with Model Generator. This will enable you to access your data in a strongly typed manner and will greatly simplify working with your data.

Kontent.ai Delivery SDK

Installation

npm i @kontent-ai/delivery-sdk --save

UMD

When using UMD bundle and including this library in script tag on your html page, you can find it under the kontentDelivery global variable.

Bundles are distributed in dist/bundles folder:

  • dist/bundles/kontent-delivery.umd.js
  • dist/bundles/kontent-delivery.umd.min.js

Getting started

import { type IContentItem, type Elements, createDeliveryClient } from '@kontent-ai/delivery-sdk';

/**
 * Generate models with @kontent-ai/model-generator
 */
export type Movie = IContentItem<{
    readonly title: Elements.TextElement;
    readonly plot: Elements.RichTextElement;
    readonly released: Elements.DateTimeElement;
    readonly length: Elements.NumberElement;
    readonly poster: Elements.AssetsElement;
    readonly category: Elements.MultipleChoiceElement;
    readonly stars: Elements.LinkedItemsElement<Actor>;
    readonly seoname: Elements.UrlSlugElement;
    readonly releasecategory: Elements.TaxonomyElement;
}>;

// initialize delivery client
const deliveryClient = createDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>'
});

// fetch items
const response = await deliveryClient.items<Movie>().type('<CONTENT_TYPE_CODENAME>').toPromise();

// read data of first item
const movieText = response.data.items[0].elements.title.value;

Configuration

Property type description
environmentId string environmentId of your Kontent.ai project
elementResolver? ElementResolver Element resolver used to map custom elements
previewApiKey? string Preview API key used to get unpublished content items
defaultLanguage? string Sets default language that will be used for all queries unless overridden with query parameters
proxy? IDeliveryClientProxyConfig Can be used to configure custom URLs. Useful when you use reverse proxy or have a need to transform URL - e.g. to remove 'environmentId'
secureApiKey? string Secured API key: Use secured API only when running on Node.JS server, otherwise you can expose your key
defaultQueryConfig? IQueryConfig Default configuration for all queries. Can be overridden by individual queries
httpService ? IHttpService Can be used to inject custom http service for performing requests
globalHeaders? (queryConfig: IQueryConfig) => IHeader[] Adds ability to add extra headers to each http request
retryStrategy? IRetryStrategyOptions Retry strategy configuration
linkedItemsReferenceHandler? LinkedItemsReferenceHandler Indicates if content items are automatically mapped. Available values: 'map' or 'ignore'
assetsDomain? string Custom domain for assets. Changes url of assets in both asset & rich text elements
defaultRenditionPreset? string Codename of rendition preset to be applied by default to the base asset URL path when present. When set, the SDK will provide the URL of customized images by default. Right now the only supported preset codename is default.
excludeArchivedItems? boolean Can be used to exclude archived items from all queries by default. Only applicable when preview API is used.

Use custom models for Custom elements

You can register an ElementResolver to map custom elements into dedicated element models and work with data more effectively.

For example, if you have a custom color element in your Kontent.ai project with json data like:

 "color": {
  "type": "custom",
  "name": "Color",
  "value": "{\"red\":167,\"green\":96,\"blue\":197}"
  }

You can create a ColorElement type with strongly typed properties for red, green & blue values that you will parse from the json.

import { ElementModels, Elements, createDeliveryClient } from '@kontent-ai/delivery-sdk';

type ColorElement = Elements.CustomElement<{
    red: number;
    green: number;
    blue: number;
}>;

To resolve your custom element into ColorElement, use the ElementResolver in your delivery client config:

const client = createDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>',
    elementResolver: (elementWrapper) => {
        if (elementWrapper.rawElement.type === 'color') {
            const parsed = JSON.parse(elementWrapper.rawElement.value);

            return {
                red: parsed.red,
                green: parsed.green,
                blue: parsed.blue
            };
        }

        return undefined;
    }
});

You can then use this custom element type in your models:

type Movie = IContentItem<{
    color: ColorElement;
    title: Elements.TextElement;
}>;

Querying

The SDK supports the following query parameters: depthParameter, elementsParameter, excludeElementsParameter, limitParameter, orderParameter, skipParameter and languageParameter. For more information about the parameters, see the SDK query methods below. You can also head over to Delivery API reference.

// Gets 5 items based on the Movie type
deliveryClient.items<Movie>().type('movie').limitParameter(5).skipParameter(2).toPromise();
Filter
depthParameter
elementsParameter
excludeElementsParameter
limitParameter
orderParameter
skipParameter
languageParameter

Filtering

Filters are also considered query parameters and can be combined. See Content filtering in API reference for more examples.

// Gets items based on the Movie type with 'Warrior' in their 'Title' element
deliveryClient.items<Movie>().type('movie').equalsFilter('elements.title', 'Warrior').toPromise();
Filter
type
types
allFilter
anyFilter
containsFilter
equalsFilter
greaterThanFilter
greaterThanOrEqualFilter
inFilter
lessThanFilter
lessThanOrEqualFilter
rangeFilter
emptyFilter
NotEmptyFilter
notEqualsFilter
notInFilter

Soring

You can sort data by using any of the following methods:

deliveryClient.items<Movie>().type('movie').orderByDescending('elements.title').toPromise();
deliveryClient.items<Movie>().type('movie').orderByAscending('elements.title').toPromise();
deliveryClient.items<Movie>().type('movie').orderParameter('elements.title', 'desc').toPromise();

Execute queries with a custom URL

When you have an URL (i.e. for next page in paging, for testing purposes or just if you prefer to build it on your own) and still want to leverage SDK functionality such as type mapping, property resolving etc., use withUrl parameter on any query such as:

deliveryClient
    .items<Movie>()
    .withUrl('https://deliver.kontent.ai/da5abe9f-fdad-4168-97cd-b3464be2ccb9/items?system.type=movie')
    .toPromise();

Custom parameters

In case you need to use custom parameters to build up an URL, use withParameter method:

deliveryClient.items<Movie>().withParameter('name', 'value').toPromise();

Get localized items

You can specify language of items with languageParameter of a particular query. You can also define default language that will be used if languageParameter is not used during the initialization of delivery client.

import { createDeliveryClient } from '@kontent-ai/delivery-sdk';

var deliveryClient = new createDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>',
    defaultLanguage: 'es'
});

// Gets items in 'es' language because it is marked as default
deliveryClient.item('warrior').toPromise();

// Gets items in 'en' language because language parameter has priority over the default one
deliveryClient.item('warrior').languageParameter(`en`).toPromise();

Preview mode

You can enable the preview mode either globally (when initializing the DeliveryClient) or per query. In each case, you need to set previewApiKey in the delivery client global configuration.

Enable preview mode globally

import { createDeliveryClient } from '@kontent-ai/delivery-sdk';

const deliveryClient = createDeliveryClient({
  environmentId: '<YOUR_ENVIRONMENT_ID>';
  previewApiKey: '<YOUR_PREVIEW_API_KEY>',
  defaultQueryConfig: {
    usePreviewMode: true
  }
});

Enable preview mode per query

deliveryClient
    .items()
    .queryConfig({
        usePreviewMode: true
    })
    .toPromise();

Secure Delivery API

Using the Delivery API with secure access enabled is recommend only when the request is not being executed on the client (browser) because otherwise you will expose the API key publicly.

import { createDeliveryClient } from '@kontent-ai/delivery-sdk';

const deliveryClient = createDeliveryClient({
  environmentId: '<YOUR_ENVIRONMENT_ID>';
  secureApiKey: '<YOUR_SECURE_ACCESS_KEY>',
    defaultQueryConfig: {
      // Enabled secure access for all queries
      useSecuredMode: true
  }
});

As with preview mode, you can also override global settings on query level.

deliveryClient
    .items()
    .queryConfig({
        // Enables secure access only for this query
        useSecuredMode: true
    })
    .toPromise();

Image transformation

Use ImageUrlBuilder for your image transformations on asset URLs.

import { transformImageUrl } from '@kontent-ai/delivery-sdk';

// Sample asset URL; You'll find URLs like these in asset and rich text elements
const assetUrl = `https://assets-eu-01.kc-usercontent.com/ede994d8-bb05-01b5-9c33-8b65e7372306/4f45e0a8-4fc3-4143-a81f-55b7e4ce7daa/warrior.jpg`;

const transformedUrl = transformImageUrl(assetUrl)
    .withDpr(2)
    .withCompression('lossless')
    .withQuality(4)
    .withHeight(200)
    .withWidth(100)
    .getUrl();

Paging

All listing queries support automatic paging. To use automatic paging, use toAllPromise extension method:

// this executed multiple HTTP requests until it gets all items
const response = await deliveryClient.items().limitParameter(5).toAllPromise();

// only gets 3 pages at maximum
const response = await deliveryClient.items().limitParameter(5).toAllPromise({
    pages: 3
});

Resolving rich text elements

Rich text elements in Kontent.ai may contain linked items and components. For example, if you write a blog post, you might want to insert a video or testimonial to a specific place in your article.

To learn how to work with Rich text element have a look at @kontent-ai/rich-text-resolver library.

Get content types

deliveryClient.type('movie').toPromise();
deliveryClient.types().toPromise();
deliveryClient.types().toAllPromise();

Get taxonomies

deliveryClient.taxonomy('taxonomyGroupName').toPromise();
deliveryClient.taxonomies().toPromise();
deliveryClient.taxonomies().toAllPromise();

Proxy configuration

If you want to use a proxy server, you need to use a different domain or otherwise transform the URL. By using proxy configuration option you can define your own base domains as well as URL format. This is useful if you need to for example hide the environmentId from the URL.

IDeliveryClientProxyConfig offers 3 ways of configuring proxy URL:

  1. baseUrl - Base URL used for all requests. Defaults to 'deliver.kontent.ai'
  2. basePreviewUrl - Base url used for preview requests. Defaults to 'preview-deliver.kontent.ai'
  3. advancedProxyUrlResolver - Resolver function where you get IProxyUrlData context data (includes domain, action, query parameters..) and can fully customize the final URL.

Examples:

const client = createDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>',
    // Will be used instead of 'https://deliver.kontent.ai' for all requests.
    // Parameters, filters, project Id and other parts of the URL will use default values.
    proxy: {
        baseUrl: 'http://my-proxy.io'
    }
});
const client = createDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>',
    proxy: {
        advancedProxyUrlResolver: (data) => {
            const action = data.action; // /items
            const domain = data.domain; // https://deliver.kontent.ai
            const environmentId = data.environmentId; // xxx
            const queryString = data.queryString; // e.g. ?depth=1&elements=xElement
            const queryParameters = data.queryParameters; // array with query parameters parameters
            const queryConfig = data.queryConfig; // query configuration
            return `http://my-proxy.io${action}${queryString}`; // proxy URL with omitted project Id
        }
    }
});

Error handling

If the error originates in Kontent.ai (see error responses), you will get a DeliveryError object instance with more specific information. Otherwise, you will get the original error.

import { DeliveryError } from '@kontent-ai/delivery-sdk';

try {
    await deliveryClient.item('invalid_codename').toPromise();
} catch (error) {
    if (error instanceof DeliveryError) {
        // delivery specific error (e.g. item with codename not found...)
        console.log(err.message, err.errorCode);
    } else {
        // some other error
        console.log(error);
    }
}

Remapping json responses

In some scenarios, you might want to store json response for later use and use SDK to map the response for you. There are 2 ways you can map previously stored json:

const result = await deliveryClient.item<Movie>('codename').toPromise();
const json = result.response.data;

// approach #1
const remappedData = deliveryClient.mappingService.viewContentItemResponse<Movie>(json);

// approach #2
const remappedData = deliveryClient.item<Movie>(movieCodename).map(json);

Handling circular references

By default, the SDK automatically maps content items present in linked items & rich text elements. Linked items can reference other linked items in their tree (e.g. child item referencing parent) which may cause infinite nesting (circular reference). This behavior is not an issue for most scenarios, in fact it is beneficial as you can easily access all linked items. However, you cannot easily serialize such model. Using e.g. JSON.stringify would fail if there are circular references.

For this reason, you may disable mapping of linked items with linkedItemsReferenceHandler configuration option.

const client = getTestDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>',
    linkedItemsReferenceHandler: 'ignore' // or 'map'
});

Using custom HTTP service

The SDK allows you to inject your own instance of a class implementing the IHttpService interface. This way you can easily mock responses, implement your own http service, or modify the requests in some other way.

Sample of a test HttpService implementation can be found at https://github.com/kontent-ai/core-sdk-js

Once you have your HttpService, use it in delivery client initialization:

const deliveryClient = createDeliveryClient({
    environmentId: '<YOUR_ENVIRONMENT_ID>',
    httpService: YourHttpServiceImplementation
});

Debugging

Accessing request data

Every response from this SDK contains additional debug data you can use to inspect when something is not right or if you need to access response headers or other network related properties.

const deliveryResponse = await createDeliveryClient({ environmentId: 'environmentId' })
    .item('itemCodename')
    .toPromise();
const rawResponseData = deliveryResponse.response; // contains raw response data, headers, status etc..
const responseHeaders = deliveryResponse.response.headers;

Getting URL of a query

In case you need to get the raw URL of a request before calling it, use the getUrl() method on any query.

const queryText = deliveryClient
    .items()
    .type('movie')
    .limitParameter(10)
    .orderParameter('system.codename', 'desc')
    .getUrl();

console.log(queryText);
// outputs:
// https://deliver.kontent.ai/b52fa0db-84ec-4310-8f7c-3b94ed06644d/items?limit=10&order=system.codename[desc]&system.type=movie

Testing

Note: You need to have Chrome installed in order to run tests via Karma.

  • npm run test:browser Runs tests in Chrome
  • npm run test:node Runs tests in node.js
  • npm run test:all Runs all test

If you want to mock http responses, it is possible to use custom Http Service as a part of the delivery client configuration.

Feedback & Contribution

Feedback & Contributions are welcomed. Feel free to take/start an issue & submit PR.