Helix Querybuilder

AEM Querybuilder for JavaScript (Browser, Node, Deno)

Status

Installation

$ npm install @adobe/helix-querybuilder

Background

The AEM QueryBuilder is a Java and REST API for executing server-side queries using a custom Query Builder Language (QBL). QBL was designed to be:

1. implementation agnostic
2. HTML-form friendly (you should not need JavaScript to build a query)
3. simple (no joins or projections)

QBL Language and Notation

As URL Query String

The most common way of expressing queries is as a query string appended to the URL of the resource that is able to execute queries. In the context of Project Helix, this would be the Helix Data Embed Action.

An example query might look like this:

https://adobeioruntime.net/api/v1/web/helix/helix-services/data-embed@v1/https://blogs.adobe.com/psirt/?feed=atom&hlx_property=author&hlx_property.value=svishnoi

The query is encoded in the URL parameters hlx_property=author&hlx_property.value=svishnoi.

As a multi-line text

This URL query string notation is most practical in day-to-day use, but a bit hard to read. Therefore a multi-line text notation is used that uses line breaks to separate key-value-pairs, does not use prefixes, nor URL-encoding:

property=author
property.value=svishnoi

As JSON or YAML

When using QBL in configuration files or JavaScript applications, it can be convenient to represent QBL in JSON like this:

{
  "property": {
    "property": "author",
    "value": "svishnoi"
  }
}

or as YAML like:

property:
  property: author
  value: svishnoi

In the following examples the multi-line and YAML notation will be used.

Of course, repeating the name of the predicate is boring and tedious, so the short-hand _ can be used instead of the inner repetition of the predicate name:

{
  "property": {
    "_": "author",
    "value": "svishnoi"
  }
}

or as YAML like:

property:
  _: author
  value: svishnoi

Usage

Simple Usage

import { qb } from '@adobe/querybuilder';

const filter = qb.filter(window.location.search);
const filtered = filter(dataarray);

Loading from JSON to create a filter

// other loaders are availale, e.g. text and url
import { load } from '@adobe/querybuilder/src/loaders/json.js'
// other adapters will be made available soon
import { adapt } from '@adobe/querybuilder/src/adapters/filter.js'

const qb = load(JSON.parse(input));
const filter = adapt(qb);

filter([
  { foo: 'bar'}
]);

API Reference

For more, see the API documentation.

Other Adapters

• filter creates a function that filters an in-memory array
• algolia returns a pair of search string and options that can be used with algoliasearch
• odata creates an OData filter object that can be used with Azure cognitive search or Excel

Development

Build

$ npm install

Test

$ npm test

Lint

$ npm run lint

Developing Loaders

A loader must implement and export a load function that accepts a Query Builder Language representation in any form and returns a Query Builder AST object.

If you have the QBL as key-value pairs, then you can use { nest } from '@adobe/querybuilder/src/utils.js' for a quick transformation.