Code Monkey home page Code Monkey logo

react-refetch's Introduction

React Refetch

A simple, declarative, and composable way to fetch data for React components.

build status npm version npm downloads

Installation

Requires React 0.14 or later.

npm install --save react-refetch

This assumes that you’re using npm package manager with a module bundler like Webpack or Browserify to consume CommonJS modules.

Introduction

See Introducing React Refetch on the Heroku Engineering Blog for background and a quick introduction to this project.

Motivation

This project was inspired by (and forked from) React Redux. Redux/Flux is a wonderful library/pattern for applications that need to maintain complicated client-side state; however, if your application is mostly fetching and rendering read-only data from a server, it can over-complicate the architecture to fetch data in actions, reduce it into the store, only to select it back out again. The other approach of fetching data inside the component and dumping it in local state is also messy and makes components smarter and more mutable than they need to be. This module allows you to wrap a component in a connect() decorator like react-redux, but instead of mapping state to props, this lets you map props to URLs to props. This lets you keep your components completely stateless, describe data sources in a declarative manner, and delegate the complexities of data fetching to this module. Advanced options are also supported to lazy load data, poll for new data, and post data to the server.

Example

If you have a component called Profile that has a userId prop, you can wrap it in connect() to map userId to one or more requests and assign them to new props called userFetch and likesFetch:

import React, { Component, PropTypes } from 'react'
import { connect, PromiseState } from 'react-refetch'

export default class Profile extends Component {
  render() {
    // see below
  }
}

connect((props) => ({
  userFetch:  `/users/${props.userId}`,
  likesFetch: `/users/${props.userId}/likes`
}))(Profile)

When the component mounts, the requests will be calculated, fetched, and the result will be passed into the component as the props specified. The result is represented as a PromiseState, which is a synchronous representation of the fetch Promise. It will either be pending, fulfilled, or rejected. This makes it simple to reason about the fetch state at the point in time the component is rendered:

render() {
  const { userFetch, likesFetch } = this.props 

  if (userFetch.pending) {
    return <LoadingAnimation/>
  } else if (userFetch.rejected) {
    return <Error error={userFetch.reason}/>
  } else if (userFetch.fulfilled) {
    return <User user={userFetch.value}/>
  }
  
  // similar for `likesFetch`
}

See the composing responses to see how to handle userFetch and likesFetch together. Although not included in this library because of application-specific defaults, see an example PromiseStateContainer and its example usage for a way to abstract and simplify the rendering of PromiseStates.

Refetching

When new props are received, the requests are re-calculated, and if they changed, the data is refetched and passed into the component as new PromiseStates. Using something like React Router to derive the props from the URL in the browser, the application can control state changes just by changing the URL. When the URL changes, the props change, which recalculates the requests, new data is fetched, and it is reinjected into the components:

react-refetch-flow

By default, the requests are compared using their URL, headers, and body; however, if you want to use a custom value for the comparison, set the comparison attribute on the request. This can be helpful when the request should or should not be refetched in response to a prop change that is not in the request itself. A common situation where this occurs is when two different requests should be refetched together even though one of the requests does not actually include the prop. Note, this is using the request object syntax for userStatsFetch instead of just a plain URL string. This syntax allows for more advanced options. See the API documentation for details:

connect((props) => ({
  usersFetch:  `/users?status=${props.status}&page=${props.page}`,
  userStatsFetch: { url: `/users/stats`, comparison: `${props.status}:${props.page}` }
}))(UsersList)

In this example, usersFetch is refetched every time props.status or props.page changes because they changes its URL. However, userStatsFetch does not contain these props in its URL, so would not normally be refetched, but because we added comparison: ${props.status}:${props.page}, it will be refetched along with usersFetch. In general, you should only rely on changes to the requests themselves to control when data is refetched, but this technique can be helpful when finer-grained control is needed.

If you always want data to be refetched when any new props are received, set the force: true option on the request. This will take precedence over any custom comparison and the default request comparison. For example:

connect((props) => ({
  usersFetch:  `/users?status=${props.status}&page=${props.page}`,
  userStatsFetch: { url: `/users/stats`, force: true }
}))(UsersList)

Setting force: true should be avoid if at all possible because it could result in extraneous data fetching and rendering of the component. Try to use the default comparison or custom comparison option instead.

Automatic Refreshing

If the refreshInterval option is provided along with a URL, the data will be refreshed that many milliseconds after the last successful response. If a request was ever rejected, it will not be refreshed or otherwise retried. In this example, likesFetch will be refreshed every minute. Note, this is using the request object syntax for likeFetch instead of just a plain URL string. This syntax allows for more advanced options. See the API documentation for details.

connect((props) => ({
  userFetch:  `/users/${props.userId}`,
  likesFetch: { url: `/users/${props.userId}/likes`, refreshInterval: 60000 }
}))(Profile)

When refreshing, the PromiseState will be the same as a the previous fulfilled state, but with the refreshing attribute set. That is, pending will remain unset and the existing value will be left in tact. When the refresh completes, refreshing will be unset and the value will be updated with the latest data. If the refresh is rejected, the PromiseState will move into a rejected and not attempt to refresh again.

Fetch Functions

Instead of mapping the props directly to a URL string or request object, you can also map the props to a function that returns a URL string or request object. When the component receives props, instead of the data being fetched immediately and injected as a PromiseState, the function is bound to the props and injected into the component as functional prop to be called later (usually in response to a user action). This can be used to either lazy load data, post data to the server, or refresh data. These are best shown with examples:

Lazy Loading

Here is a simple example of lazy loading the likesFetch with a function:

connect((props) => ({
  userFetch:  `/users/${props.userId}`,
  lazyFetchLikes: (max) => ({
    likesFetch: `/users/${props.userId}/likes?max=${max}`
  })
}))(Profile)

In this example, userFetch is fetched normally when the component receives props, but lazyFetchLikes is a function that returns likesFetch, so nothing is fetched immediately. Instead lazyFetchLikes is injected into the component as a function to be called later inside the component:

this.props.lazyFetchLikes(10)

When this function is called, the request is calculated using both the bound props and any passed in arguments, and the likesFetch result is injected into the component normally as a PromiseState.

Posting Data

Functions can also be used for post data to the server in response to a user action. For example:

connect((props) => ({
  postLike: (subject) => ({
    postLikeResponse: {
      url: `/users/${props.userId}/likes`
      method: 'POST'
      body: JSON.stringify({subject: subject})
    }
  })
}))(Profile)

The postLike function is injected in as a prop, which can then be tied to a button:

<button onClick={() => this.props.postLike(someSubject)}>Like!</button>

When the user clicks the button, someSubject is posted to the URL and the response is injected as a new postLikeResponse prop as a PromiseState to show progress and feedback to the user.

Manually Refreshing Data

Functions can also be used to manually refresh data by overwriting an existing PromiseState:

connect((props) => ({
  userFetch: `/users/${props.userId}`,
  refreshUser: () => ({
    userFetch: `/users/${props.userId}`
  })
}))(Profile)

The userFetch data is first loaded normally when the component receives props, but the refreshUser function is also injected into the component. When this.props.refreshUser() is called, the data is fetched again and overwrites the existing userFetch. Note, you may wish to set the refreshing: true option to avoid the existing PromiseState being cleared while refresh is in progress. This should generally only be used for user-invoked refreshes; see above for automatically refreshing on an interval.

Posting + Refreshing Data

The two examples above can be combined to post data to the server and refresh an existing PromiseState. This is a common pattern when a responding to a user action to update a resource and reflect that update in the component. For example, if PATCH /users/:user_id responds with the updated user, it can be used to overwrite the existing userFetch when the user updates her name:

connect((props) => ({
  userFetch: `/users/${props.userId}`,
  updateUser: (firstName, lastName) => ({
    userFetch: {
       url: `/users/${props.userId}`
       method: 'PATCH'
       body: JSON.stringify({firstName: firstName, lastName: lastName})
     }
   })
}))(Profile)

Composing Responses

If a component needs data from more than one URL, the PromiseStates can be combined with PromiseState.all() to be pending until all the PromiseStates have been fulfilled. For example:

render() {
  const { userFetch, likesFetch } = this.props 
  
  // compose multiple PromiseStates together to wait on them as a whole 
  const allFetches = PromiseState.all([userFetch, likesFetch])

  // render the different promise states
  if (allFetches.pending) {
    return <LoadingAnimation/>
  } else if (allFetches.rejected) {
    return <Error error={allFetches.reason}/>
  } else if (allFetches.fulfilled) {
    // decompose the PromiseState back into individual
    const [user, likes] = allFetches.value
    return (
      <div>
          <User data={user}/>
          <Likes data={likes}/>
      </div>
    )
  }
}

Similarly, PromiseState.race() can be used to return the first settled PromiseState. Like their asynchronous Promise counterparts, PromiseStates can be chained with then() and catch(); however, the handlers are run immediately to transform the existing state. This can be helpful to handle errors or transform values as part of a composition. For example, to provide a fallback value to likesFetch in the case of failure:

PromiseState.all([userFetch, likesFetch.catch((reason) => [])])

Chaining Requests

Inside of connect(), requests can be chained using then(), catch(), andThen() and andCatch() to trigger additional requests after a previous request is fulfilled. These are not to be confused with the similar sounding functions on PromiseState, which are on the response side, are synchronous, and are executed for every change of the PromiseState.

then() is helpful for cases where multiple requests are required to get the data needed by the component and the subsequent request relies on data from the previous request. For example, if you need to make a request to /foos/${name} to look up foo.id and then make a second request to /bar-for-foos-by-id/${foo.id} and return the whole thing as barFetch (the component will not have access to the intermediate foo):

connect(({ name }) => ({
  barFetch: {
    url: `/foos/${name}`,
    then: (foo) => `/bar-for-foos-by-id/${foo.id}`
  }
}))

andThen() is similar, but is intended for side effect requests where you still need access to the result of the first request and/or need to fanout to multiple requests:

connect(({ name }) => ({
  fooFetch: {
    url: `/foos/${name}`,
    andThen: (foo) => { 
      barFetch: `/bar-for-foos-by-id/${foo.id}` 
    }
  }
}))

This is also helpful for cases where a fetch function is changing data that is in some other fetch that is a collection. For example, if you have a list of foos and you create a new foo and the list needs to be refreshed:

 connect(({ name }) => ({
    foosFetch: '/foos',
    createFoo: (name) => {
       method: 'POST',
       url: '/foos',
       andThen: () => { 
         foosFetch: { 
           url: '/foos', 
           refreshing: true 
         }
       }
    }
  }))

catch and andCatch are similar, but for error cases.

Accessing Headers & Metadata

Both request and response headers and other metadata are accessible. Custom request headers can be set on the request as an object:

connect((props) => ({
  userFetch: {
    url: `/users/${props.userId}`,
    headers: {
        FOO: 'foo',
        BAR: 'bar',
    }
   }
}))(Profile)

The raw Request and Response can be accessed via the meta attribute on the PromiseState. For example, to access the a response header:

userFetch.meta.response.headers.get('FOO')

Do not attempt to read bodies directly from meta.request or meta.response. They are provided for metadata purposes only.

Complete Example

This is a complex example demonstrating various feature at once:

import React, { Component, PropTypes } from 'react'
import { connect, PromiseState } from 'react-refetch'

class Profile extends React.Component {
  static propTypes = {
    params: PropTypes.shape({
      userId: PropTypes.string.isRequired,
    }).isRequired,
    userFetch: PropTypes.instanceOf(PromiseState).isRequired
    likesFetch: PropTypes.instanceOf(PromiseState).isRequired
    updateStatus: PropTypes.func.isRequired
    updateStatusResponse: PropTypes.instanceOf(PromiseState) // will not be set until after `updateStatus()` is called
  }
  
  render() {
    const { userFetch, likesFetch } = this.props 
    
    // compose multiple PromiseStates together to wait on them as a whole 
    const allFetches = PromiseState.all([userFetch, likesFetch])
  
    // render the different promise states
    if (allFetches.pending) {
      return <LoadingAnimation/>
    } else if (allFetches.rejected) {
      return <Error error={allFetches.reason}/>
    } else if (allFetches.fulfilled) {
      // decompose the PromiseState back into individual
      const [user, likes] = allFetches.value
      return (
        <div>
            <User data={user}/>
            <Likes data={likes}/>
        </div>
      )
    }
    
    // call `updateState()` on button click
    <button onClick={() => { this.props.updateStatus("Hello World")} }>Update Status</button>
    
    if (updateStatusResponse) {
      // render the different promise states, but will be `null` until `updateState()` is called
    }
  }
}

// declare the requests for fetching the data, assign them props, and connect to the component.
export default connect((props) => {
  return {
    // simple GET from a URL injected as `userFetch` prop
    // if `userId` changes, data will be refetched
    userFetch: `/users/${props.params.userId}`,                             
    
    // similar to `userFetch`, but using object syntax 
    // specifies a refresh interval to poll for new data
    likesFetch: { 
      url: `/users/${props.userId}/likes`, 
      refreshInterval: 60000 
    },
    
    // declaring a request as a function
    // not immediately fetched, but rather bound to the `userId` prop and injected as `updateStatus` prop
    // when `updateStatus` is called, the `status` is posted and the response is injected as `updateStatusResponse` prop.
    updateStatus: (status) => {
        updateStatusResponse: {
            url: `/users/${props.params.userId}/status`,
            method: 'POST',
            body: status
        }
    }
  }
})(Profile)

API Documentation

Support

This software is provided "as is", without warranty or support of any kind, express or implied. See license for details.

License

MIT

react-refetch's People

Contributors

agraboso avatar djkirby avatar ellbee avatar erikras avatar esamattis avatar gaearon avatar gnoff avatar grrowl avatar hnordt avatar ide avatar istarkov avatar jhollingworth avatar jihchi avatar jlogsdon avatar joaojeronimo avatar mattydoincode avatar mikekidder avatar mindjuice avatar moretti avatar moroshko avatar oliviertassinari avatar omnidan avatar pspsynedra avatar ryanbrainard avatar sirreal avatar timdorr avatar timtyrrell avatar udfalkso avatar wbuchwalter avatar zeke avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.