Integrating with Algolia

You can integrate Elastic Path Commerce Cloud with Algolia quickly and easily in Commerce Manager.

Whenever you publish a catalog, the full catalog is synced to Algolia using an index name composed of the catalog name and the first part of the catalog ID. This is triggered using events delivered by a webhook.

1. When a catalog is published, a catalog-release.updated event is triggered.

2. Once a catalog-release.updated event is received by the webhook, the integration begins to execute.

3. The integration performs the following:

   • Retrieves product and taxonomy data for the catalog.
   • Retrieves the mapping for Elastic Path Commerce Cloud fields to Algolia fields.
   • Builds or updates the Algolia index.

For more information, watch a video.

Before you begin configuring your Algolia integration in Commerce Manager, you need to collect the necessary setup information from Algolia and Commerce Manager.

You can find this information in the Settings, Teams and Access, API keys section on the Algolia Dashboard.

Collect the following Elastic Path Commerce Cloud API keys. You can find this information in System > Application Keys in Commerce Manager when logged in as a user with Seller Admin privileges.

Now that you've collected the setup information, let's begin by configuring the integration.

1. In Commerce Manager, go to SYSTEM > Integrations Hub.
2. Under Search, click on the Algolia Full /Delta / Large Catalog integration.
3. Review the Summary page and click Configure. The current version information is displayed.
4. Click Next.
5. Configure the following information and click Finish.

The integration starts an execution which:

• Configures a webhook to listen to the catalog-release.updated event, named EPCC Algolia Integration.
• Creates a flow for containing field mappings between Elastic Path Commerce Cloud and Algolia, called Algolia Mapping.

Now that you've configured the integration, let's learn how you use it.

To sync the catalog to Algolia, you create a map of the data to pass from Elastic Path Commerce Cloud to Algolia. The mapping consists of:

• Source field.
• Destination field.
• Any flags to trigger optional functions.

The integration uses a flow to contain the fields that you want to use. The source of the Elastic Path Commerce Cloud data is the catalog delta file. Additionally, there are some special fields that perform additional transformation work. When the integration is initially configured, a default set of mapping fields are created, but you can change these depending on your requirements. See the table below.

The following table describes some example fields in Commerce Manager and their corresponding fields in Algolia.

The following are special field names:

• main_image triggers a lookup of the URL from the file associated to the main image relationship on the product.
• taxonomy retrieves the hierarchy and builds a custom Algolia-friendly format.

1. In Commerce Manager, go to Flows > Algolia Mapping. (If you have just configured your integration, refresh Commerce Manager to pick up the newly-added flow from the integration.)
2. Click New Entry.
3. Create the field mappings you desire by supplying the source field and destination field.
4. In Algolia, create the fields described above. Refer to the documentation provided with Algolia for more information.

The following mapping flags are available.

• includeHierarchy - Use this flag if you want the name of the hierarchy to be included in the bread_crumbs.
• forceSaleFlag - By default, if a price is on sale, an on_sale flag is set to true in the price book. If you also want an on_sale flag at the root of the product, you can add this flag to the price mapping.
• forceJSON - It is a boolean value that is used to determine whether the input value should be parsed as JSON or not. This is useful if you want to treat an Elastic Path field as an embedded structure and not as a string.
• convertDateToTS - A boolean value used to determine whether the input value is parsed as a date. The date is converted into a number that represents a UNIX timestamp.

Once the integration is configured and you have the mappings you want, the catalog is synced with Algolia with every new publish of your catalogs.

If a catalog has already been published and changes are made to the catalog and the catalog is published again, this results in a partial catalog update. In other words, only the products that have been created, updated or removed from the previous catalog release are updated. These are made directly to the live index in Algolia.

A full catalog update can be triggered in the following scenarios.

1. The first time a catalog is published. The delta file is then used to create a temporary index. The settings, rules, and so on, from the live version are then copied onto the temporary index. Once the index is completely loaded, the temporary index is moved to the live index.
2. You can trigger a full catalog update by setting export_full_delta to true when publishing a catalog. Here's an sample request:

curl --location -g '{{pcmUrl}}/catalogs/{{catalogId}}/releases' \
--header 'accept: application/json' \
--header 'content-type: application/json' \
--header 'x-moltin-auth-store: {{storeID}}' \
--data '{
    "export_full_delta" : true
}'

Index names are created from the catalog name and a portion of the catalog id. This allows the integration to manage many catalogs at the same time. This dependency is important to consider when naming catalogs in Elastic Path Commerce Cloud and indexes in Algolia.