core
core is a concept introduced by Nicholas C. Zakas in this video
It helps you create scalable applications written in Javascript, giving you some structure and patterns to keep everything separated.
tl;dr
- Check out this video - Introduction (portuguese)
- Check out this video - Isolation of DOM (portuguese)
The Idea
Conceptually, everything in your application is a module, and your modules should work independently from each other, so if one module breaks, the others should not.
A module should never talk directly to another module, for that you use a combination of listeners and notifications.
Getting Started
So let's think about the twitter page, and how we could re-build it using the core concept
Everything inside a red square is a module, they work in a way that they don't depend on any other modules, and they should be programmed that way as well.
Installing
npm i @eroc/core
Import
Raw import
import {Core} from "./node_modules/@eroc/core/dist/core.js";
With rollup, webpack or parcel
import {Core} from "@eroc/core";
NodeJs or Browserify
const {Core} = require("@eroc/core");
Let's start with the tweet module.
Building modules
Core.register('tweet', function(sandbox) {
return {
init: function() {
console.log('starting tweet module');
}
}
});
This way you can register a new module, and the method init
will be called once the module is started, if it exists.
Core.start('tweet');
For every module you have an enviroment called sandbox, this is the only guy your module will ever speak to.
It garantees that if your module tries to call something that doesn't exists or is broken, it won't break the module itself.
Now let's think about the tweet list, which fetches all tweets from your timeline.
Core.register('tweet-list', function(sandbox) {
return {
init: function() {
}
}
});
If you have multiple modules you can start everything using Core.start();
with no parameters, and everything will be started. But if you need some specific order for starting your modules, you can call the modules you want first, and then use Core.start()
.
Destroying modules
You might want to stop a module in some point, this can be easily done using the method Core.stop()
.
Core.register('tweet-list', function(sandbox) {
return {
destroy: function() {
console.log('Module destroyed');
}
}
});
Core.start('tweet-list');
Core.stop('tweet-list');
When you stop a module, the method destroy
will be called, if it exists.
Comunicating between modules
Now, thinking about Twitter, everytime you tweet something, it should appear on your tweet list right? but since our modules don't talk directly to each other, let's use sandbox
to do this for us:
First of all, our tweet
module should notify other modules that something has happened.
Core.register('tweet', function(sandbox) {
return {
init: function(sandbox) {
setInterval(function() {
sandbox.notify({
type: 'new-tweet',
data: {
author: 'Mauricio Soares',
text: 'core is pretty #cool'
}
})
}, 5 * 1000)
}
}
});
Every 5 seconds, this module notifies everything that is listening to new-tweet
that something has happened. If nothing is listening to it, than nothing happens.
Our tweet-list
is going to listen for this notifications.
Core.register('tweet-list', function(sandbox) {
return {
init: function() {
sandbox.listen('new-tweet', this.newTweet);
},
newTweet: function(data) {
}
}
});
Cool right? If one of those modules stop working, than it won't break the other one!
Isolation of DOM
If you have a DOM element with the same ID as the module name, it will be accessible inside the module using this.el
.
<div id="tweet"></div>
Core.register('tweet', function() {
return {
init: function() {
console.log(this.el);
}
}
});
Core.start('tweet');
If there's no DOM element, then this.el
will return null
.
Core.register('tweet', function(sandbox) {
return {
init: function() {
jQuery('#tweet').on('click', this.newTweet);
},
newTweet: function() {
}
};
});
A module should not talk to other modules directly anything else but the sandbox
.
Last thoughts
This is basically how core works, below you will find the documentation of methods and parameters.
Docs
Core.register( moduleName, constructor )
Register a new module.
moduleName
(string): The name of the moduleconstructor
(function): The implementation of the module
Usage
Core.register('module', function() {})
Core.start( moduleName )
Starts the named module. If a value is returned in the init
method, it can be grabbed in the return of the method Core.start
. If no parameters are passed, it starts all unstarted modules.
moduleName
(string): The name of the module
Usage
Core.start('module');
Core.stop( moduleName )
Stops the named module. If a value is returned in the destroy
method, it can be grabbed in the return of the method Core.stop
, If no parameters are passed, it stop all modules.
moduleName
(string): The name of the module
Usage
Core.stop('module');
sandbox.listen( notification, callback, context, force )
Listens to other modules notifications, to overwrite a notification you must use the parameter force
notification
(string | array): The name of the notification you are listening tocallback
(function): The callback when the notification is triggeredcontext
(object): The value of this
inside the callbackforce
(boolean): If you want to overwrite a notification, use true
here
sandbox.notify( config )
Notifies other modules
config
(object): The configuration object
type
(string): The notification that will be triggereddata
(function | string | number | boolean | array): The data that will be passed in the callback
Maintainer
Contributing
- Fork core
- Create a topic branch -
git checkout -b my_branch
- Push to your branch -
git push origin my_branch
- Send me a Pull Request
- That's it!
Please respect the indentation rules and code style.
Use 2 spaces, not tabs.
New features? Would you mind testing it? :)
Testing
You need NodeJS installed on your machine
npm i
npm run bundle
npm t
Release History
- 2018-10-30 v0.13.0 remove deprecated startAll and stopAll
- 2018-10-29 v0.12.0 Drop bower and publish on npm
- 2018 Various changes
- 2015-06-15 v0.7.3 Refactor UMD
- 2015-05-14 v0.7.2 Hotfix with ID's
- 2015-02-15 v0.7.0 Deprecate
Core.stopAll
- 2015-02-12 v0.6.0 Deprecate
Core.startAll
- 2015-02-05 v0.5.0 Changes
x
to use
in Sandbox
- 2015-01-17 v0.4.0 Add UMD
- 2015-01-15 v0.3.0 Ability to return values from init and destroy methods
- 2015-01-10 v0.2.1 Improve error messages
- 2014-12-30 v0.2.0 Isolation of DOM in modules
- 2014-12-21 v0.1.0 Release usable version
License
(The MIT License)
Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
'Software'), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.