I have an application developed in Angular with typescript which has 20+ screens with list views of different entities, such as List of Movies, List of Actors, List of theatres etc. I have created a BaseListComponent<T extends Base> in angular and encapsulated most of the common functions such as view, filter, delete, search, pagination etc into this BaseListComponent. And I am creating children of it like MoviesListComponent extends BaseListComponent<Movies>. This way I need to add minimum functionality into children classes which reduces scope of errors and also efforts for unit test cases.
Now I am converting this application into React with typescript. I was studying about react hooks and learnt that many developers are preferring function based approach over classes based components. How can I create above structure with functions. As per my understanding there is no inheritance in functions. So I am doubtful if Generics will be feasible with functions. If not, how can I solve this problem using react's popular approach of decomposition?
Is there any guideline for using Generics or decomposition in react?
To do it "react style", first you have to make a component BaseListComponent. Then put the functions you listed "view, filter, delete, search, pagination etc" in this one. But whenever you want some functions to behave differently for deriving classes, you put a callback in that function.
For example:
renderView = () => { // function of BaseListComponent
const header = (
<div>--I am Header--</div>
);
const footer = () => (
<div>--------</div>
);
const letDerivedClassDecide = this.props.renderCenter();
return (
<div style={{ display: 'flex', flexDirection: 'column' }}>
{header}
{letDerivedClassDecide}
{footer}
</div>
);
}
This example means, you could control the basic layout as a vertical flex div, but you let the deriving class implement the this.props.renderCenter, so that its result can be combined with basic layout.
view = () => { // function of BaseListComponent
let someResult = 100;
someResult = this.props.view();
return someResult;
}
The same idea, this.props.view is implemented by deriving class.
Finally in the deriving class it would be like:
render() { // function of DeriveComponent; can be MoviesListComponent
return (
<BaseListComponent
renderCenter={this.renderCenter}
view={this.view}
>
);
}
Note that, the deriving class is not actually "derive" or extends the base component, it just composite it in the render() function. So both BaseListComponent and MoviesListComponent extends React.Component (or React.PureComponent when considering performance issue).
In the React documentation they say:
React also supports using a string (instead of a callback) as a ref prop on any component, although this approach is mostly legacy at this point.
https://facebook.github.io/react/docs/more-about-refs.html
Take the following example:
class Foo extends Component {
render() {
return <input onClick={() => this.action()} ref={input => (this._input = input)} />;
}
action() {
console.log(this._input.value);
}
}
Why should I prefer this, instead of:
class Foo extends Component {
render() {
return <input onClick={() => this.action()} ref='input' />;
}
action() {
console.log(this.refs.input.value);
}
}
?
It seems just much more clean and easier the second example.
Are there risks that the string method will be deprecated?
NB: I'm looking for the "official" answer to the statement in the documentation, I'm not asking about personal preferences and so on.
While perhaps more simple, the old refs API can get difficult in some edge cases, like when used in a callback. All kind of static analysis is a pain with strings, too. The callback based API can do everything the string API can do and more with just a little added verbosity.
class Repeat extends React.Component {
render() {
return <ul> {
[...Array(+this.props.times)].map((_, i) => {
return <li key={i}> { this.props.template(i) } </li>
})
} </ul>
}
}
class Hello extends React.Component {
constructor() {
super();
this.refDict = {};
}
render() {
return <Repeat times="3" template={i => <span ref= {el => this.refDict[i] = el}> Hello {i} </span>} />
{/* ^^^ Try doing this with the string API */}
}
}
Further discussion and a bit more comprehensive list of the possible issues with the string based api can be found from issue #1373, where the callback based api was introduced. I'll include here a list from the issue description:
The ref API is broken is several aspects.
You have to refer to this.refs['myname'] as strings to be Closure Compiler Advanced Mode compatible.
It doesn't allow the notion of multiple owners of a single instance.
Magical dynamic strings potentially break optimizations in VMs.
It needs to be always consistent, because it's synchronously resolved. This means that asynchronous batching of rendering introduces potential bugs.
We currently have a hook to get sibling refs so that you can have one component refer to it's sibling as a context reference. This only works one level. This breaks the ability to wrap one of those in an encapsulation.
It can't be statically typed. You have to cast it at any use in languages like TypeScript.
There's no way to attach the ref to the correct "owner" in a callback invoked by a child. <Child renderer={index => <div ref="test">{index}</div>} /> -- this ref will be attached where the callback is issued, not in the current owner.
The docs call the old string API "legacy" to make it clearer that the callback-based API is the preferred approach, as is discussed in this commit and in this PR which are the ones that actually put those statements to the documentation in the first place. Also note that a few of the comments imply that the string based refs api might be deprecated at some point.
Originally posted by danabramov on https://news.ycombinator.com/edit?id=12093234
String refs are not composable. A wrapping component can’t “snoop” on a ref to a child if it already has an existing string ref. On the other hand, callback refs don’t have a single owner, so you can always compose them.
String refs don’t work with static analysis like Flow. Flow can’t guess the magic that framework does to make the string ref “appear” on this.refs, as well as its type (which could be different). Callback refs are friendlier to static analysis.
The owner for a string ref is determined by the currently executing component. This means that with a common “render callback” pattern (e.g. <DataTable renderRow={this.renderRow} />), the wrong component will own the ref (it will end up on DataTable instead of your component defining renderRow).
String refs force React to keep track of currently executing component. This is problematic because it makes react module stateful, and thus causes weird errors when react module is duplicated in the bundle.
I'm coming from the angular world where I could extract logic to a service/factory and consume them in my controllers.
I'm trying to understand how can I achieve the same in a React application.
Let's say that I have a component that validates user's password input (it's strength). It's logic is pretty complex hence I don't want to write it in the component it self.
Where should I write this logic? In a store if I'm using flux? Or is there a better option?
The issue becomes extremely simple when you realize that an Angular service is just an object which delivers a set of context-independent methods. It's just the Angular DI mechanism which makes it look more complicated. The DI is useful as it takes care of creating and maintaining instances for you but you don't really need it.
Consider a popular AJAX library named axios (which you've probably heard of):
import axios from "axios";
axios.post(...);
Doesn't it behave as a service? It provides a set of methods responsible for some specific logic and is independent from the main code.
Your example case was about creating an isolated set of methods for validating your inputs (e.g. checking the password strength). Some suggested to put these methods inside the components which for me is clearly an anti-pattern. What if the validation involves making and processing XHR backend calls or doing complex calculations? Would you mix this logic with mouse click handlers and other UI specific stuff? Nonsense. The same with the container/HOC approach. Wrapping your component just for adding a method which will check whether the value has a digit in it? Come on.
I would just create a new file named say 'ValidationService.js' and organize it as follows:
const ValidationService = {
firstValidationMethod: function(value) {
//inspect the value
},
secondValidationMethod: function(value) {
//inspect the value
}
};
export default ValidationService;
Then in your component:
import ValidationService from "./services/ValidationService.js";
...
//inside the component
yourInputChangeHandler(event) {
if(!ValidationService.firstValidationMethod(event.target.value) {
//show a validation warning
return false;
}
//proceed
}
Use this service from anywhere you want. If the validation rules change you need to focus on the ValidationService.js file only.
You may need a more complicated service which depends on other services. In this case your service file may return a class constructor instead of a static object so you can create an instance of the object by yourself in the component. You may also consider implementing a simple singleton for making sure that there is always only one instance of the service object in use across the entire application.
The first answer doesn't reflect the current Container vs Presenter paradigm.
If you need to do something, like validate a password, you'd likely have a function that does it. You'd be passing that function to your reusable view as a prop.
Containers
So, the correct way to do it is to write a ValidatorContainer, which will have that function as a property, and wrap the form in it, passing the right props in to the child. When it comes to your view, your validator container wraps your view and the view consumes the containers logic.
Validation could be all done in the container's properties, but it you're using a 3rd party validator, or any simple validation service, you can use the service as a property of the container component and use it in the container's methods. I've done this for restful components and it works very well.
Providers
If there's a bit more configuration necessary, you can use a Provider/Consumer model. A provider is a high level component that wraps somewhere close to and underneath the top application object (the one you mount) and supplies a part of itself, or a property configured in the top layer, to the context API. I then set my container elements to consume the context.
The parent/child context relations don't have to be near each other, just the child has to be descended in some way. Redux stores and the React Router function in this way. I've used it to provide a root restful context for my rest containers (if I don't provide my own).
(note: the context API is marked experimental in the docs, but I don't think it is any more, considering what's using it).
//An example of a Provider component, takes a preconfigured restful.js
//object and makes it available anywhere in the application
export default class RestfulProvider extends React.Component {
constructor(props){
super(props);
if(!("restful" in props)){
throw Error("Restful service must be provided");
}
}
getChildContext(){
return {
api: this.props.restful
};
}
render() {
return this.props.children;
}
}
RestfulProvider.childContextTypes = {
api: React.PropTypes.object
};
Middleware
A further way I haven't tried, but seen used, is to use middleware in conjunction with Redux. You define your service object outside the application, or at least, higher than the redux store. During store creation, you inject the service into the middleware and the middleware handles any actions that affect the service.
In this way, I could inject my restful.js object into the middleware and replace my container methods with independent actions. I'd still need a container component to provide the actions to the form view layer, but connect() and mapDispatchToProps have me covered there.
The new v4 react-router-redux uses this method to impact the state of the history, for example.
//Example middleware from react-router-redux
//History is our service here and actions change it.
import { CALL_HISTORY_METHOD } from './actions'
/**
* This middleware captures CALL_HISTORY_METHOD actions to redirect to the
* provided history object. This will prevent these actions from reaching your
* reducer or any middleware that comes after this one.
*/
export default function routerMiddleware(history) {
return () => next => action => {
if (action.type !== CALL_HISTORY_METHOD) {
return next(action)
}
const { payload: { method, args } } = action
history[method](...args)
}
}
I needed some formatting logic to be shared across multiple components and as an Angular developer also naturally leaned towards a service.
I shared the logic by putting it in a separate file
function format(input) {
//convert input to output
return output;
}
module.exports = {
format: format
};
and then imported it as a module
import formatter from '../services/formatter.service';
//then in component
render() {
return formatter.format(this.props.data);
}
Keep in mind that the purpose of React is to better couple things that logically should be coupled. If you're designing a complicated "validate password" method, where should it be coupled?
Well you're going to need to use it every time the user needs to input a new password. This could be on the registration screen, a "forgot password" screen, an administrator "reset password for another user" screen, etc.
But in any of those cases, it's always going to be tied to some text input field. So that's where it should be coupled.
Make a very small React component that consists solely of an input field and the associated validation logic. Input that component within all of the forms that might want to have a password input.
It's essentially the same outcome as having a service/factory for the logic, but you're coupling it directly to the input. So you now never need to tell that function where to look for it's validation input, as it is permanently tied together.
Same situation: Having done multiple Angular projects and moving to React, not having a simple way to provide services through DI seems like a missing piece (the particulars of the service aside).
Using context and ES7 decorators we can come close:
https://jaysoo.ca/2015/06/09/react-contexts-and-dependency-injection/
Seems these guys have taken it a step further / in a different direction:
http://blog.wolksoftware.com/dependency-injection-in-react-powered-inversifyjs
Still feels like working against the grain. Will revisit this answer in 6 months time after undertaking a major React project.
EDIT: Back 6 months later with some more React experience. Consider the nature of the logic:
Is it tied (only) to UI? Move it into a component (accepted answer).
Is it tied (only) to state management? Move it into a thunk.
Tied to both? Move to separate file, consume in component through a selector and in thunks.
Some also reach for HOCs for reuse but for me the above covers almost all use cases. Also, consider scaling state management using ducks to keep concerns separate and state UI-centric.
I also came from Angular.js area and the services and factories in React.js are more simple.
You can use plain functions or classes, callback style and event Mobx like me :)
// Here we have Service class > dont forget that in JS class is Function
class HttpService {
constructor() {
this.data = "Hello data from HttpService";
this.getData = this.getData.bind(this);
}
getData() {
return this.data;
}
}
// Making Instance of class > it's object now
const http = new HttpService();
// Here is React Class extended By React
class ReactApp extends React.Component {
state = {
data: ""
};
componentDidMount() {
const data = http.getData();
this.setState({
data: data
});
}
render() {
return <div>{this.state.data}</div>;
}
}
ReactDOM.render(<ReactApp />, document.getElementById("root"));
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/15.1.0/react.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/15.1.0/react-dom.min.js"></script>
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width">
<title>JS Bin</title>
</head>
<body>
<div id="root"></div>
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/15.1.0/react.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/react/15.1.0/react-dom.min.js"></script>
</body>
</html>
Here is simple example :
I am from Angular as well and trying out React, as of now, one recommended(?) way seems to be using High-Order Components:
A higher-order component (HOC) is an advanced technique in React for
reusing component logic. HOCs are not part of the React API, per se.
They are a pattern that emerges from React’s compositional nature.
Let's say you have input and textarea and like to apply the same validation logic:
const Input = (props) => (
<input type="text"
style={props.style}
onChange={props.onChange} />
)
const TextArea = (props) => (
<textarea rows="3"
style={props.style}
onChange={props.onChange} >
</textarea>
)
Then write a HOC that does validate and style wrapped component:
function withValidator(WrappedComponent) {
return class extends React.Component {
constructor(props) {
super(props)
this.validateAndStyle = this.validateAndStyle.bind(this)
this.state = {
style: {}
}
}
validateAndStyle(e) {
const value = e.target.value
const valid = value && value.length > 3 // shared logic here
const style = valid ? {} : { border: '2px solid red' }
console.log(value, valid)
this.setState({
style: style
})
}
render() {
return <WrappedComponent
onChange={this.validateAndStyle}
style={this.state.style}
{...this.props} />
}
}
}
Now those HOCs share the same validating behavior:
const InputWithValidator = withValidator(Input)
const TextAreaWithValidator = withValidator(TextArea)
render((
<div>
<InputWithValidator />
<TextAreaWithValidator />
</div>
), document.getElementById('root'));
I created a simple demo.
Edit: Another demo is using props to pass an array of functions so that you can share logic composed by multiple validating functions across HOCs like:
<InputWithValidator validators={[validator1,validator2]} />
<TextAreaWithValidator validators={[validator1,validator2]} />
Edit2: React 16.8+ provides a new feature, Hook, another nice way to share logic.
const Input = (props) => {
const inputValidation = useInputValidation()
return (
<input type="text"
{...inputValidation} />
)
}
function useInputValidation() {
const [value, setValue] = useState('')
const [style, setStyle] = useState({})
function handleChange(e) {
const value = e.target.value
setValue(value)
const valid = value && value.length > 3 // shared logic here
const style = valid ? {} : { border: '2px solid red' }
console.log(value, valid)
setStyle(style)
}
return {
value,
style,
onChange: handleChange
}
}
https://stackblitz.com/edit/react-shared-validation-logic-using-hook?file=index.js
If you are still looking for a service like Angular, you can try the react-rxbuilder library
You can use #Injectable to register the service, and then you can use useService or CountService.ins to use the service in the component
import { RxService, Injectable, useService } from "react-rxbuilder";
#Injectable()
export class CountService {
static ins: CountService;
count = 0;
inc() {
this.count++;
}
}
export default function App() {
const [s] = useService(CountService);
return (
<div className="App">
<h1>{s.count}</h1>
<button onClick={s.inc}>inc</button>
</div>
);
}
// Finally use `RxService` in your root component
render(<RxService>{() => <App />}</RxService>, document.getElementById("root"));
Precautions
Depends on rxjs and typescript
Cannot use arrow functions in the service
Service is not limited to Angular, even in Angular2+,
Service is just collection of helper functions...
And there are many ways to create them and reuse them across the application...
1) They can be all separated function which are exported from a js file, similar as below:
export const firstFunction = () => {
return "firstFunction";
}
export const secondFunction = () => {
return "secondFunction";
}
//etc
2) We can also use factory method like, with collection of functions... with ES6 it can be a class rather than a function constructor:
class myService {
constructor() {
this._data = null;
}
setMyService(data) {
this._data = data;
}
getMyService() {
return this._data;
}
}
In this case you need make an instance with new key...
const myServiceInstance = new myService();
Also in this case, each instance has it's own life, so be careful if you want to share it across, in that case you should export only the instance you want...
3) If your function and utils not gonna be shared, you can even put them in React component, in this case, just as function in your react component...
class Greeting extends React.Component {
getName() {
return "Alireza Dezfoolian";
}
render() {
return <h1>Hello, {this.getName()}</h1>;
}
}
4) Another way you may handle things, could be using Redux, it's a temporary store for you, so if you have it in your React application, it can help you with many getter setter functions you use... It's like a big store that keep tracks of your states and can share it across your components, so can get rid of many pain for getter setter stuffs we use in the services...
It's always good to do a DRY code and not repeating what needs to be used to make the code reusable and readable, but don't try to follow Angular ways in React app, as mentioned in item 4, using Redux can reduce your need of services and you limit using them for some reuseable helper functions like item 1...
I am in the same boat like you. In the case you mention, I would implement the input validation UI component as a React component.
I agree the implementation of the validation logic itself should (must) not be coupled. Therefore I would put it into a separate JS module.
That is, for logic that should not be coupled use a JS module/class in separate file, and use require/import to de-couple the component from the "service".
This allows for dependency injection and unit testing of the two independently.
In the React world we have two types of logic: Stateful and stateless. Now this is the main concept to grasp when starting with React. That here we update state which should update UI as opposed to Angular's direct updates of dom. The two types of logics are:
That do not depend on state changes, i.e. static logic which doesn't need to re-render something based on state changes. For such cases just create regular js files and import them like a library or helper methods
If you have some code that depends on state and u need to resuse it then two options - hocs and the newer hooks. Hooks are a bit hard to wrap our heads around but basically they would force their parent to rerender if their internal state changes so any stateful logic can be defined and reused in different components, and each hook instance would have its own isolated scope.
It's a little bit of a thinking shift to understand state and declarative components but feel free to ask followup questions in comments
or you can inject the class inheritance "http" into React Component
via props object.
update :
ReactDOM.render(<ReactApp data={app} />, document.getElementById('root'));
Simply edit React Component ReactApp like this:
class ReactApp extends React.Component {
state = {
data: ''
}
render(){
return (
<div>
{this.props.data.getData()}
</div>
)
}
}
It is possible to use export keyword to use functions from file which contains necessary methods.
Let me show an example. Let's say we have a file called someService.ts:
export const foo = (formId: string) => {
// ... the code is omitted for the brevity
}
export const bar = (): Entity[] => [
// ... the code is omitted for the brevity
]
export default {
foo,
bar,
}
Then we can use this service in component like this:
import {
foo,
bar,
} from './someService'
const InnerOrderModal: FC = observer(() => {
const handleFormClick = (value: unknown, item: any) => {
foo(item.key)
bar()
return <></>
}
I have a navigation toolbar, with H1 in it.
I also various sub content for each page, as child component.
How can I pass the title information, from the child page component, to the parent navigation?
I've tried to use Context, but it only propagate from Parent to Childs.
Her is a simplified exemple:
const App = React.createClass({
render() {
return (
<div>
<toolbar>
<h1>{myTitleAccordingToPage}</h1>
</toolbar>
<main>
{this.props.children}
</main>
</div>
)
}
})
const A = React.createClass({
render() {
return (
<div>
Content for A page
</div>
)
}
})
const B = React.createClass({
render() {
return (
<div>
Content for B page
</div>
)
}
})
Extract state for your navigation either into a store, or somewhere else, and propagate that state to your application via props.
var page = {
title: 'foo'
otherData: ''
}
ReactDOM.render( <App page={ page }, element )
Now your top-level component knows about the state and can filter accordingly down to its children, for example, you might change your App render function to pass different bits of the application state down to different children i.e.
(I'm passing props explicitly here to show what is going on, and using a pure function to render, use a class and render method if that is more comfortable for you, the advantage of only ever passing props is that it is easier to make your functions pure, which has many advantages)
const App = props => {
<div>
<Toolbar title={ props.page.title } />
<Main otherData={ props.page.otherData />
</div>
}
Now you have created top-down data flow, making your application more predictable and possibly paving the way to use the pure render function to maybe give you performance boosts for free. The key with top-down data-flow is really making your application more predictable, which should translate into testability, and thus reliability, and makes hacking on your application easier.
The react-router module can help with creating structures like this. There is a small learning phase when adopting the module but the benefits for an app like the one you are creating should outweigh this and the speed of development once learnt may very well outweigh any learning outlay you invest into it.
Possible it's time to use Flex https://facebook.github.io/flux/docs/dispatcher.html#content
Usable approuch to two way data binding http://voidcanvas.com/react-tutorial-two-way-data-binding/
Original documentation to two way data binding, but mixin now is now supported https://facebook.github.io/react/docs/two-way-binding-helpers.html
( For mixin possible to use react-mixin npm package )
EDIT:
There is an opinion at comments that: "two-way data-binding is often a bad idea, and if your application makes more sense with two-way data binding then React should probably not be the framework to employ". So you need to rearrange your application architecture to avoid using two-way data binding if possible. Use flux instead.
I have created a very big(500loc, It its very big and difficult to reason about) react class. Its an autocomplete. Whats the recommended way to split this up with react/reflux. Add the logic to som services? What is best practise. I have Stores but as I understand they shouldn't contain view logic.
It's difficult to be specific to your case given that you haven't provided the code behind your component, but if I were to develop an autocomplete component I would do it as follows.
Facebook's Thinking in React guidelines suggest to break down your UI into components that represent a single aspect of your data model:
In your case, you could implement the following hierarchy:
AutoComplete
|
--AutoCompleteInput
|
AutoCompleteResults (list of results)
|
--AutoCompleteResult (individual result)
So at a very high level...
AutoComplete.jsx:
[...]
render() {
return (
<div>
<AutoCompleteInput />
<AutoCompleteResults />
</div>
);
}
AutoCompleteInput.jsx:
[...]
updateQuery(e) {
// FYI - you should ideally throttle this logic
autoCompleteActions.updateQuery(e.target.value);
}
render() {
return <input onChange={this.updateQuery} />
}
AutoCompleteResults.jsx:
[...]
onStoreUpdated() { // hypothetically invoked when store receives new data
this.setState({
results: productListStore.getResults()
});
}
render() {
const { results } = this.state;
const children = results.map(result => {
return <AutoCompleteResult result={result} />
});
return (
<ul>
{children}
</ul>
);
}
You're correct to state that stores should not contain view logic, but they are allowed to store both general app state and data that's resulted from an XHR, for example. In the above example, the autoCompleteActions.updateQuery would consume a data service method to get the autocomplete options for the specific query. The success callback would then be responsible for passing the returned data to the store. The store should ideally notify subscribed components that it has received new data.
Although this is a simple overview, the result is that the components driving the AutoComplete functionality are broken down to respect single responsibilities, which should simplify your testing.