inu-engine

1.0.0-pre.0 • Public • Published

Follow your dreams
inu-engine

:dog: :rocket: composable user interface state and effects manager pull streams

table of contents
  • features
  • demos
  • example
  • concepts
  • api
  • install
  • inspiration
  • features

    • minimal size: inu-engine + yo-yo + pull-stream weighs only ~8kb
    • engine is a data structure: only need to learn 4 functions, automatically supports plugins
    • architecture is fractal: compose one engine from many smaller engine
    • single source of truth: the state of your engine is a single object tree
    • state is read-only: update state by dispatching an action, an object describing what happened
    • update with pure functions: updates are handled by a pure function, no magic
    • first-class side effects: initial state or updates can include an effect, an object describing what will happen
    • omakase: consistent flavoring with pull streams all the way down

    demos

    if you want to share anything using inu-engine, add your thing here!

    example

    const start = require('inu-engine')
    const html = require('yo-yo')
    const pull = require('pull-stream')
    const delay = require('pull-delay')
     
    const engine = {
     
      init: () => ({
        model: 0,
        effect: 'SCHEDULE_TICK' // start perpetual motion
      }),
     
      update: (model, action) => {
        switch (action) {
          case 'TICK':
            return {
              model: (model + 1) % 60,
              effect: 'SCHEDULE_TICK'
            }
          default:
            return { model }
        }
      },
     
      view: (model, dispatch) => html`
        <div class='clock'>
          Seconds Elapsed: ${model}
        </div>
      `,
     
      run: (effect, sources) => {
        switch (effect) {
          case 'SCHEDULE_TICK':
            return pull(
              pull.values(['TICK']),
              delay(1000)
            )
        }
      }
    }
     
    const main = document.querySelector('.main')
    const { views } = start(engine)
     
    pull(
      views(),
      pull.drain(function (view) {
        html.update(main, view)
      })
    )

    for a full example of composing multiple engines together, see source and demo.

    concepts

    imagine your app’s current state is described as a plain object. for example, the initial state of a todo app might look like this:

    var initState = {
      model: {
        todos: [{
          text: 'Eat food',
          completed: true
        }, { 
          text: 'Exercise',
          completed: false
        }],
        visibilityFilter: 'SHOW_COMPLETED'
      },
      effect: 'FETCH_TODOS'
    }

    this state object describes the model (a list of todo items and an option for how to filter these items) and any optional effect (we immediately want to fetch for any new todo items).

    to change something in the state, we need to dispatch an action. an action is a plain JavaScript object (notice how we don’t introduce any magic?) that describes what happened. here are a few example actions:

    { type: 'ADD_TODO', text: 'Go to swimming pool' }
    { type: 'TOGGLE_TODO', index: 1 }
    { type: 'SET_VISIBILITY_FILTER', filter: 'SHOW_ALL' }
    { type: 'LOAD_TODOS' }

    enforcing that every change is described as an action lets us have a clear understanding of what’s going on in the app. if something changed, we know why it changed. actions are like breadcrumbs of what has happened.

    finally, to tie state and actions together, we write an update function. again, nothing magic about it — it’s just a function that takes the model and action as arguments, and returns the next state of the app.

    it would be hard to write such a function for a big app, so we write smaller functions managing parts of the state:

    function visibilityFilter (model, action) {
      if (action.type === 'SET_VISIBILITY_FILTER') {
        return action.filter
      } else {
        return { model }
      }
    }
     
    function todos (model, action) {
      switch (action.type) {
        case 'ADD_TODO':
          return { model: model.concat([{ text: action.text, completed: false }]) }
        case 'TOGGLE_TODO':
          return {
            model: model.map((todo, index) =>
              action.index === index ?
                { text: todo.text, completed: !todo.completed } :
                todo
            )
          }
        case 'LOAD_TODOS':
          return { model, effect: 'FETCH_TODOS' }
        default:
          return { model }
      }
    }

    and we write another update function that manages the complete state of our app by calling those two update functions for the corresponding state keys:

    function appUpdate (model, action) {
      const todosState = todos(model.todos, action)
      const visibilityFilterState = visibilityFilter(model.visibilityFilter, action)
     
      return {
        model: {
          todos: todosState.model,
          visibilityFilter: visibilityFilter.model
        },
        effect: todosState.effect
      }
    }

    if any effect is returned by an update function, we want to run it. this run functions is able to listen to any future changes and return a stream of any new actions.

    here's how we handle our effect to fetch any todos, using pull-stream as pull:

    function appRun (effect, sources) {
      if (effect === 'FETCH_TODOS') {
        return pull(
          fetchTodos(),
          pull.map(todo => {
            return {
              type: 'ADD_TODO',
              text: todo.text
            }
          })
        )
      }
    }

    now that we have our state, changes, and side effects managed in a predictable (and easy-to-test) way, we want to view our epic todo list.

    here's a simplified view using yo-yo as html:

    function appView (model, dispatch) {
      return html`
        <div class='todos'>
          ${model.todos.map((todo, index) => html`
            <div class='todo'>
              ${todo.text}
              <button onclick=${toggleTodo(index)}
            </div>
          `)}
        </div>
      `
     
      function toggleTodo (index) {
        return (ev) => dispatch({ 'TOGGLE_TODO', })
      }
    }

    put it all together and we have an inu-engine engine!

    const app = {
      init: () => initState,
      update: appUpdate,
      view: appView,
      run: appRun
    }

    that's it for inu-engine. note that we're only using plain functions and objects. inu-engine (and inu) come with a few utilities to facilitate this pattern, but the main idea is that you describe how your state is updated over time in response to action objects, and 90% of the code you write is just plain JavaScript, with no use of inu-engine itself, its APIs, or any magic.

    (credit @gaearon of redux for initial source of this intro)

    api

    where state is an object with a required key model and an optional key effect,

    an inu-engine engine is defined by an object with the following (optional) keys:

    • init: a function returning the initial state
    • update: a update(model, action) pure function, returns the new state
    • view: a view(model, dispatch) pure function, returns the user interface declaration
    • run: a run(effect, sources) function, returns an optional pull source stream of future actions

    start = require('inu-engine')

    the top-level inu-engine module is a grab bag of all inu-engine/* modules.

    you can also require each module separately like require('inu-engine/start').

    sources = start(app)

    sources is an object with the following keys:

    streams flow diagram

    * in this context, state-ful means that the pull source stream will always start with the last value (if any) first.

    install

    npm install --save inu-engine

    inspiration

    license

    The Apache License

    Copyright © 2017 Michael Williams

    Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0
    

    Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

    Readme

    Keywords

    none

    Package Sidebar

    Install

    npm i inu-engine

    Weekly Downloads

    1

    Version

    1.0.0-pre.0

    License

    Apache-2.0

    Last publish

    Collaborators

    • ahdinosaur