rtContent available in: eng ita

Process Overview

This page targets developers who want to integrate AccelaSearch on platforms that are not natively supported. We refer to such platforms as custom and the integration process on those platforms as custom integration.

Custom integration process broad overview

To integrate AccelAsearch, the following steps must be taken:

• register at https://accelasearch.com/
• obtain an API key
• get access to the collector database
• insert data about your shops
• insert data about your products
• configure your search engine
• index

Once your first synchronization is complete, you may update your products’ information anytime by writing changes into your collector database. Remember that AccelaSearch requires a soft-deletion mechanism, so products and attributes must not be directly deleted. Instead, they should be marked as deleted. This helps our engine to keep track of changes in the collector. AccelaSearch will delete them eventually when they are no longer needed.

The following BPMN gives a formal representation of the synchronization process.

BPMN describing the synchronization phase

Collector Database

The collector database is a SQL database that holds data about shops and products. You have exclusive access to one single database to store information about all your shops and products. While the database schema is mostly self-explanatory, some clarifications are worthy:

• for every product, every attribute but category, image, price, and stock is handled through an EAV approach using the products_attr_* tables
• products and attributes must not be directly deleted; instead, they must be soft-deleted by setting deleted = 1 and lastupdate = NOW() both for the entity and the corresponding product; AccelaSearch will periodically hard delete items marked for deletion