pouchdb-svelte's Introduction

pouchdb-svelte

PouchDB-Svelte is a PouchDB adaptation for Svelte Rollup.

Usage

npm i pouchdb-svelte
npm i -D rollup-plugin-node-polyfills

Svelte

//  rollup.config.js

import nodePolyfills from 'rollup-plugin-node-polyfills';

export default {
    plugins: [
        nodePolyfills(),
        ...
    ]
}

import PouchDB from 'pouchdb-svelte';

const testDB = new PouchDB('test');

Sapper

//  rollup.config.js

import nodePolyfills from 'rollup-plugin-node-polyfills';

export default {
    client: {
        plugins: [
            nodePolyfills(),
            ...
        ]
    }
}

//not in the server nor in the serviceworker

import PouchDB from 'pouchdb-svelte';

(() => {
    if(typeof process !== 'undefined') return; // stops the Sapper SSR 
    const testDB = new PouchDB('test');
})();

Notes:

You can put this script in any .svelte file in your app.

It is really important to block Sapper SSR in this script. Because SSR will try to run pouchdb-svelte (pouchdb-browser) and attempt to create the LevelDB database on the server. This is not a good thing for two reasons:

The LevelDB adapter is not present in the pouchdb-browser vanilla.

If you want to run a LevelDB database on your server (polka / express), it is better to use pouchdb or pouchdb-node.

You could use pouchdb-svelte for the Front and pouchdb for the Back.

Why don't use onMount ?

Example: We have two .svelte files.

parent.svelte
Child.svelte

Child.svelte is a component of parent.svelte.
In these two files there is an onMount function.

The svelte runtime will trigger Child.svelte onMount at the first time, then parent.svelte onMount at the second time.
This is a problem if you create a pouchdb instance in parent.svelte and if you want to use it in Child.svelte.

With pouchdb-svelte you don't have to use onMount and so you can create a single instance that can be passed to components.