@arbel/firesql

FireSQL - Query Firestore using SQL syntax

What is FireSQL?

FireSQL is a library built on top of the official Firebase SDK that allows you to query Cloud Firestore using SQL syntax. It's smart enough to issue the minimum amount of queries necessary to the Firestore servers in order to get the data that you request.

On top of that, it offers some of the handy utilities that you're used to when using SQL, so that it can provide a better querying experience beyond what's offered by the native querying methods.

Installation

Just add firesql and firebase to your project:

npm install firesql firebase
# or
yarn add firesql firebase

If you want to receive realtime updates when querying, then you will also need to install rxjs and rxfire:

npm install firesql firebase rxjs rxfire
# or
yarn add firesql firebase rxjs rxfire

Usage

// You can either query the collections at the root of the database...
const dbRef = firebase.firestore();

// ... or the subcollections of some document
const docRef = firebase.firestore().doc('someDoc');

// And then just pass that reference to FireSQL
const fireSQL = new FireSQL(dbRef);

// Use `.query()` to get a one-time result
fireSQL.query('SELECT * FROM myCollection').then(documents => {
  documents.forEach(doc => {
    /* Do something with the document */
  });
});

// Use `.rxQuery()` to get an observable for realtime results.
// Don't forget to import "firesql/rx" first (see example below).
fireSQL.rxQuery('SELECT * FROM myCollection').subscribe(documents => {
  /* Got an update with the documents! */
});

Examples

One-time result (Promise)

import { FireSQL } from 'firesql';
import firebase from 'firebase/app';
import 'firebase/firestore';

firebase.initializeApp({ /* ... */ });

const fireSQL = new FireSQL(firebase.firestore());

const citiesPromise = fireSQL.query(`
  SELECT name AS city, country, population AS people
  FROM cities
  WHERE country = 'USA' AND population > 700000
  ORDER BY country, population DESC
  LIMIT 10
`);

citiesPromise.then(cities => {
  for (const city of cities) {
    console.log(
      `${city.city} in ${city.country} has ${city.people} people`
    );
  }
});

Realtime updates (Observable)

import { FireSQL } from 'firesql';
import firebase from 'firebase/app';
import 'firesql/rx'; // <-- Important! Don't forget
import 'firebase/firestore';

firebase.initializeApp({ /* ... */ });

const fireSQL = new FireSQL(firebase.firestore());

const cities$ = fireSQL.rxQuery(`
  SELECT city, category, AVG(price) AS avgPrice
  FROM restaurants
  WHERE category IN ("Mexican", "Indian", "Brunch")
  GROUP BY city, category
`);

cities$.subscribe(results => {
  /* REALTIME AGGREGATED DATA! */
});

Limitations

• Only SELECT queries for now. Support for INSERT, UPDATE, and DELETE might come in the future.
• No support for JOINs.
• LIMIT doesn't accept an OFFSET, only a single number.
• No support for aggregate function COUNT.
• • If using GROUP BY, it cannot be combined with ORDER BY nor LIMIT.
• No support for negating conditions with NOT.
• Limited LIKE. Allows for searches in the form of WHERE field LIKE 'value%', to look for fields that begin with the given value; and WHERE field LIKE 'value', which is functionally equivalent to WHERE field = 'value'.

Nested objects

You can access nested objects by using backticks around the field path. For example, if you have a collection "products" with documents like this:

{
  productName: "Firebase Hot Sauce",
  details: {
    available: true,
    stock: 42
  }
}

You could do the following queries:

SELECT *
FROM products
WHERE `details.stock` > 10

SELECT productName, `details.stock` AS productStock
FROM products
WHERE `details.available` = true

Getting the document IDs

You can use the special field __name__ to refer to the document ID (its key inside a collection). For convenience, you might want to alias it:

SELECT __name__ AS docId, country, population
FROM cities

If you always want to include the document ID, you can specify that as a global option to the FireSQL class:

const fireSQL = new FireSQL(ref, { includeId: true}); // To include it as "__name__"
const fireSQL = new FireSQL(ref, { includeId: 'fieldName'}); // To include it as "fieldName"

You can also specify that option when querying. This will always take preference over the global option:

fireSQL.query(sql, { includeId: 'id'}); // To include it as "id"
fireSQL.query(sql, { includeId: false}); // To not include it

When querying it's also possible to use the document as a search field by using __name__ directly. For example, you could search for all the documents whose IDs start with Hello:

SELECT *
FROM cities
WHERE __name__ LIKE 'Hello%'

Note: You will need to specify the includeId option if you want to obtain the document IDs when doing a SELECT * query.

How does FireSQL work?

FireSQL transforms your SQL query into one or more queries to Firestore. Once all the necessary data has been retrieved, it does some internal processing in order to give you exactly what you asked for.

For example, take the following SQL:

SELECT *
FROM cities
WHERE country = 'USA' AND population > 50000

This would get transformed into this single Firestore query:

db.collection('cities')
  .where('country', '==', 'USA')
  .where('population', '>', 50000);

That's pretty straightforward. But what about this one?

SELECT *
FROM cities
WHERE country = 'USA' OR population > 50000

There's no direct way to perform an OR query on Firestore so FireSQL splits that into 2 separate queries:

db.collection('cities').where('country', '==', 'USA');
db.collection('cities').where('population', '>', 50000);

The results are then merged and any possible duplicates are eliminated.

The same principle applies to any other query. Sometimes your SQL will result in a single Firestore query and some other times it might result in several.

For example, take a seemingly simple SQL statement like the following:

SELECT *
FROM cities
WHERE country != 'Japan' AND region IN ('north', 'east', 'west') AND (capital = true OR population > 100000)

This will need to launch a total of 12 concurrent queries to Firestore!

const cities = db.collection('cities');
cities.where('country', '<', 'Japan').where('region', '==', 'north').where('capital', '==', true);
cities.where('country', '<', 'Japan').where('region', '==', 'north').where('population', '>', 100000);
cities.where('country', '<', 'Japan').where('region', '==', 'east').where('capital', '==', true);
cities.where('country', '<', 'Japan').where('region', '==', 'east').where('population', '>', 100000);
cities.where('country', '<', 'Japan').where('region', '==', 'west').where('capital', '==', true);
cities.where('country', '<', 'Japan').where('region', '==', 'west').where('population', '>', 100000);
cities.where('country', '>', 'Japan').where('region', '==', 'north').where('capital', '==', true);
cities.where('country', '>', 'Japan').where('region', '==', 'north').where('population', '>', 100000);
cities.where('country', '>', 'Japan').where('region', '==', 'east').where('capital', '==', true);
cities.where('country', '>', 'Japan').where('region', '==', 'east').where('population', '>', 100000);
cities.where('country', '>', 'Japan').where('region', '==', 'west').where('capital', '==', true);
cities.where('country', '>', 'Japan').where('region', '==', 'west').where('population', '>', 100000);

As you can see, SQL offers a very concise and powerful way to express your query. But as they say, with great power comes great responsibility. Always be mindful of the underlying data model when using FireSQL.

Examples of supported queries:

SELECT *
FROM restaurants

SELECT name, price
FROM restaurants
WHERE city = 'Chicago'

SELECT *
FROM restaurants
WHERE category = 'Indian' AND price < 50

SELECT *
FROM restaurants
WHERE name LIKE 'Best%'

SELECT *
FROM restaurants
WHERE name LIKE 'Best%' OR city = 'Los Angeles'

SELECT *
FROM restaurants
WHERE city IN ( 'Raleigh', 'Nashvile', 'Denver' )

SELECT *
FROM restaurants
WHERE city != 'Oklahoma'

SELECT *
FROM restaurants
WHERE favorite = true

SELECT *
FROM restaurants
WHERE favorite -- Equivalent to the previous one

SELECT *
FROM restaurants
WHERE favorite IS NULL

SELECT AVG(price) AS averagePriceInChicago
FROM restaurants
WHERE city = 'Chicago'

SELECT city, MIN(price), AVG(price), MAX(price)
FROM restaurants
WHERE category = 'Indian'
GROUP BY city

SELECT *
FROM restaurants
WHERE city = 'Memphis' AND ( price < 40 OR avgRating > 8 )
ORDER BY price DESC, avgRating

SELECT *
FROM restaurants
WHERE price BETWEEN 25 AND 150
ORDER BY city, price
LIMIT 10

SELECT *
FROM restaurants
WHERE city = 'Chicago'
UNION
SELECT *
FROM restaurants
WHERE price > 200