Security News
GitHub Removes Malicious Pull Requests Targeting Open Source Repositories
GitHub removed 27 malicious pull requests attempting to inject harmful code across multiple open source repositories, in another round of low-effort attacks.
Stent is combining the ideas of Redux with the concept of state machines.
State machine is a mathematical model of computation. It's an abstract concept where the machine may have different states but at a given time fulfills only one of them. It accepts input and based on that (plus its current state) transitions to another state. Isn't it sounds familiar? Yes, it sounds like a front-end application. That's why this model/concept applies nicely to UI development.
Disclaimer: there are different types of state machines. I think the one that makes sense for front-end development is Mealy state machine.
The library is available as a npm module so npm install stent
or yarn add stent
will do the job. There's also a standalone version here (only core functionalities) which you can directly add to your page.
To create a new machine we simply import the Machine
object and call its create
method.
import { Machine } from 'stent';
const machine = Machine.create('name-of-the-machine', {
state: { name: 'idle' },
transitions: {
'idle': {
'run': 'running'
},
'running': {
'stop': 'idle'
}
}
});
{ name: 'idle'}
is the initial state of the machine and inside transitions
we define another one running
. run
and stop
are actions(inputs) that transition the machine to a new state. Notice that stop
is not available when we are at idle
state and run
when we are at running
state.
Stent library is enforcing declarative approach of programming. Which means that by defining the possible states and actions for them we clearly define what's happening in our application. The user and data flows become a lot more predictable simply because we restrict ourselves of dispatching actions are the wrong time/state. For example, does not make sense that we say stop
when we are not running
.
After the definition of the machine it knows what to expect and automatically creates a couple of things for us so we can trigger the logic. Based on the transitions
property Stent generates:
idle
state produces isIdle()
method, for running
we have isRunning()
.run()
and stop()
.We may use spaces or dashes in the state or action names but the rule of thumb is that Stent transforms the string to a camel case. For example if we have fetching data
state the machine will have isFetchingData()
method, get fresh todos
action will result in getFetchTodos()
method.
So, here's an example of how to use the machine above:
if (machine.isIdle()) {
machine.run();
}
if (machine.isRunning()) {
machine.stop();
}
console.log(machine.isIdle()); // true
The created machine accepts more then a string as a handler of the action. We may pass a function which accepts two arguments. The first one is the current state and the second one is some meta data traveling with the action (if any). For example:
const machine = Machine.create('todo-app', {
state: { name: 'idle', todos: [] },
transitions: {
'idle': {
'add todo': function (state, todo) {
return {
name: 'idle',
todos: [...state.todos, todo]
};
}
}
}
});
machine.addTodo({ title: 'Fix that damn bug' })
The state in the context of Stent is a vanilla JavaScript object literal. The only one reserved property is name
which represents the state's name. Everything else depends on our business logic. In the example above that's the todos
array.
The handler function accepts the previous state and should return a new state in a immutable fashion. Same as the Redux's reducer, whatever we return becomes the new state.
The actual todo item is passed to the addTodo
method and it comes as a second argument of the handler.
Stent also accepts a generator function as a handler. That's inspired by the redux-saga project. The generators have couple of interesting characteristics and this library uses two of them - the ability to generate multiple results (from a single function) and the ability to pause the execution. What if we need to fetch data from the server and want to handle that process with multiple states - idle
, fetching
, done
or error
. Here's how to do it with a generator as a handler:
const machine = Machine.create('todo-app', {
state: { name: 'idle', todos: [] },
transitions: {
'idle': {
'fetch todos': function * () {
yield { name: 'fetching' };
try {
const todos = yield call(getTodos, '/api/todos');
} catch (error) {
return { name: 'error', error };
}
return { name: 'done', todos };
}
}
}
});
Assuming that getTodos
is a function that accepts an endpoint as a string and returns a promise. Inside the generator we are allowed to yield
two type of things:
call
. (more about those helpers below)Generator as an action handler is suitable for the cases where we do more then one thing and/or have async operations.
<state object>
The state object is just a normal object literal. The only one required property is name
and it is used to indicate the state of the machine:
{
name: 'idle',
user: {
firstName: '...',
lastName: '...'
},
someOtherProperty: '...'
}
If you try transitioning to a state which is not defined into the transitions
section or it has no actions in it State will throw an exception. It's because once you get into that new state you are basically stuck.
Machine.<create|get|flush|connect>
The Machine
object is used for creating/managing and fetching machines.
import { Machine } from 'stent';
const appMachine = Machine.create(
'app', // name of the machine
{
state: <state object>, // initial state
transitions: {
<state name>: {
<action name>: <action handler>,
<action name>: <action handler>,
...
},
<state name>: {
<action name>: <action handler>,
<action name>: <action handler>,
...
},
...
}
}
);
// later in the code
const appMachine = Machine.get('app');
The created machine has dynamically created methods associated with the provided configuration:
is<state name>
method so we can check if the machine is in that state. For example, to check if the machine is in a fetching remote data
state we may call machine.isFetchingRemoteData()
method. The alternative is machine.state.name === 'fetching remote data'
.add new todos
is available as machine.addNewTodo(<todo data here>)
.Machine.flush()
can be used to delete the currently created machines and middlewares.
Machine.connect()
is the same as the connect helper.
<action handler>
The action handler may be just a string. In the following example fetching
is the same as { name: 'fetching' }
state object.
Machine.create('app', {
'idle': {
'fetch data': 'fetching'
}
});
Could be also a state object:
Machine.create('app', {
'idle': {
'fetch data': { name: 'fetching', data: [], pending: false }
}
});
Another variant is to use a function that returns a string. Which again results in { name: 'fetching' }
.
Machine.create('app', {
'idle': {
'fetch data': function (state, payload) {
return 'fetching';
}
}
});
Notice that the function receives the current state and some payload passed when the action is fired.
And of course we may return the actual state object. That's actually a common case because very often we want to keep some data alongside:
Machine.create('app', {
'idle': {
'fetch data': function (state, payload) {
return { name: 'fetching', answer: 42 };
}
}
});
The context of the action handler function (or generator) is the machine itself. This means that this
inside the function points to the created machine and we may call its methods. For example:
Machine.create('app', {
'idle': {
'fetch data': function (state, payload) {
if (this.isIdle()) {
this.request('/api/todos');
}
},
'request': function (state, endpoint) {
console.log(endpoint); // endpoint = /api/todos
}
}
});
In some cases you don't want to change the state but only handle the action. So feel free to skip the return
statement. If the handler returns undefined
the machine keeps its state.
We may also use a generator if we have more complex operations or/and async tasks.
Machine.create('app', {
'idle': {
'fetch data': function * (state, payload) {
yield 'fetching'; // transition to a `fetching` state
yield { name: 'fetching' } // the same but using a state object
}
}
});
More for generators and what could be yielded in the Helpers used inside generators section below.
connect
and disconnect
connect
is the short way to do Machine.get
and retrieving one or more created machines. It also provides a mechanism for subscribing for state changes.
import { connect } from 'stent/lib/helpers';
Machine.create('MachineA', ...);
Machine.create('MachineB', ...);
connect()
.with('MachineA', 'MachineB')
.map((MachineA, MachineB) => {
// called multiple times
});
The mapping function by default is called once initially and then every time when the state of the connected machines changes. So, if you need only that first call use mapOnce
instead.
connect()
.with('MachineA', 'MachineB')
.mapOnce((MachineA, MachineB) => {
// this gets called only once
});
You may also need to disconnect
which makes sense if you use the map
function. If you are connecting with mapOnce
your mapping function is getting called only once anyway.
const disconnect = connect()
.with('MachineA', 'MachineB')
.map((MachineA, MachineB) => {
// called multiple times
});
// at some point later
disconnect();
There's also a helper for integrating with React. It creates a HoC:
import React from 'react';
import { connect } from 'stent/lib/react';
class TodoList extends React.Component {
render() {
const { isIdle, todos } = this.props;
...
}
}
// `todos` and `authorization` are machines defined
// using `Machine.create` function
export default connect(TodoList)
.with('MachineA', 'MachineB')
.map((MachineA, MachineB) => {
isIdle: MachineA.isIdle,
todos: MachineB.state.todos
});
The result of the map
function goes as props to our component. Similarly to Redux's connect function. And of course the mapping function is disconnect
ed when the component is unmounted.
yield call(<function>, ...args)
It's blocking the generator function and calls <function>
with the given ...args
. <function>
could be:
import { call } from 'stent/lib/helpers';
Machine.create('app', {
'idle': {
'fetch data': function * () {
const data = yield call(requestToBackend, '/api/todos/', 'POST');
}
}
});
requestToBackend
is getting called with /api/todos/
and POST
as arguments.
yield wait(<action name/s>)
It's blocking the generator and waits for action/s. The function accepts a single argument string or array of strings.
import { wait } from 'stent/lib/helpers';
Machine.create('app', {
'idle': {
'fetch data': function * () {
const initActionPayload = yield wait('init');
const [ userProfilePayload, dataPayload ] = yield wait([
'user profile fetched',
'data processed'
]);
...
}
}
});
If you want to extend the library with some additional functionalities you may add a middleware. It's an object with a set of functions that hook to the lifecycle methods of Stent.
import { Machine } from 'stent';
Machine.addMiddleware({
onActionDispatched(next, actionName, ...args) {
console.log(`Action dispatched: ${ actionName }`);
next();
console.log(`After ${ actionName } action our state is ${ this.state.name }`);
},
onStateChanged(next) {
console.log(`The new state will be: ${ this.state.name }`);
next();
console.log(`The state now is: ${ this.state.name }`);
}
});
The hooks above are getting called just before running the internal Stent's logic. At this moment nothing in the machine is changing/executing. Calling next
will pass the control flow to Stent. Also have in mind that these methods are fired with the machine as a context. Which means that you have an access to the current state and methods.
import { Machine } from 'stent';
const machine = Machine.create('app', {
state: { name: 'idle', todos: [] },
transitions: {
'idle': {
'add new todo': function ({ todos }, todo) {
return { name: 'idle', todos: [...todos, todo] };
},
'delete todo': function ({ todos }, index) {
return { name: 'idle', todos: todos.splice(index, 1) };
},
'fetch todos': function * () {
yield 'fetching';
try {
const todos = yield call(getTodos, '/api/todos');
} catch (error) {
return { name: 'fetching failed', error };
}
return { name: 'idle', todos };
}
},
'fetching failed': {
'fetch todos': function * () {
yield { name: 'idle', error: null };
this.fetchTodos();
}
}
}
});
machine.fetchTodos();
import React from 'react';
import { connect } from 'stent/lib/react';
class TodoList extends React.Component {
render() {
const { todos, error, isFetching, fetchTodos, deleteTodo, isAuthorized } = this.props;
if (isFetching()) return <p>Loading</p>;
if (error) return (
<div>
Error fetching todos: { error }<br />
<button onClick={ fetchTodos }>try again</button>
</div>
);
return (
<ul>
{ todos.map(({ text}) => <li onClick={ deleteTodo }>{ text }</li>) }
</ul>
);
}
}
// `todos` and `authorization` are machines defined
// using `Machine.create` function
export default connect(TodoList)
.with('todos', 'authorization')
.map(({ state, isFetching, fetchTodos, deleteTodo }, { isAuthorized }) => {
todos: state.todos,
error: state.error,
isFetching,
fetchTodos,
deleteTodo,
isAuthorized
});
FAQs
Stent is combining the ideas of redux with the concept of state machines
The npm package stent receives a total of 1,395 weekly downloads. As such, stent popularity was classified as popular.
We found that stent demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
GitHub removed 27 malicious pull requests attempting to inject harmful code across multiple open source repositories, in another round of low-effort attacks.
Security News
RubyGems.org has added a new "maintainer" role that allows for publishing new versions of gems. This new permission type is aimed at improving security for gem owners and the service overall.
Security News
Node.js will be enforcing stricter semver-major PR policies a month before major releases to enhance stability and ensure reliable release candidates.