mobx-utils
Advanced tools
The mobx-utils package provides a set of utility functions and classes to complement MobX, a state management library for JavaScript applications. These utilities help simplify common patterns and enhance the functionality of MobX.
fromPromise
The `fromPromise` utility converts a promise into an observable object, allowing you to easily track the state of asynchronous operations within MobX.
const { fromPromise } = require('mobx-utils');
const { observable } = require('mobx');
const fetchData = () => new Promise(resolve => setTimeout(() => resolve('data'), 1000));
const observablePromise = fromPromise(fetchData());
observablePromise.case({
pending: () => console.log('Loading...'),
fulfilled: value => console.log('Data:', value),
rejected: error => console.log('Error:', error)
});lazyObservable
The `lazyObservable` utility creates an observable that only starts producing values when it is first observed. This is useful for deferring expensive computations or data fetching until necessary.
const { lazyObservable } = require('mobx-utils');
const lazyData = lazyObservable(sink => {
setTimeout(() => sink('Lazy data'), 1000);
});
lazyData.observe(newValue => console.log('New value:', newValue));keepAlive
The `keepAlive` utility ensures that a computed value remains active and is not garbage collected, even if there are no observers. This can be useful for maintaining the state of certain computations.
const { keepAlive } = require('mobx-utils');
const { observable, computed } = require('mobx');
const data = observable.box(1);
const computedValue = computed(() => data.get() * 2);
keepAlive(computedValue);
console.log(computedValue.get()); // 2
data.set(2);
console.log(computedValue.get()); // 4MobX-State-Tree (MST) is a state management library built on top of MobX. It provides a structured way to manage state with a focus on immutability and snapshots. Unlike mobx-utils, which offers utility functions, MST provides a more opinionated and structured approach to state management.
Redux is a popular state management library for JavaScript applications. It uses a single immutable state tree and actions to update the state. While Redux is more verbose and has a steeper learning curve compared to MobX and mobx-utils, it is widely adopted and has a large ecosystem of middleware and tools.
Recoil is a state management library for React applications developed by Facebook. It provides a more fine-grained approach to state management compared to MobX and mobx-utils, with a focus on performance and scalability. Recoil's API is designed to be simple and intuitive for React developers.
Utility functions and common patterns for MobX
This package provides utility functions and common MobX patterns build on top of MobX. It is encouraged to take a peek under the hood and read the sources of these utilities. Feel free to open a PR with your own utilities. For large new features, please open an issue first.
NPM: npm install mobx-utils --save
CDN: https://unpkg.com/mobx-utils/mobx-utils.umd.js
import {function_name} from 'mobx-utils'
fromPromise takes a Promise, extends it with 2 observable properties that track
the status of the promise and returns it. The returned object has the following observable properties:
value: either the initial value, the value the Promise resolved to, or the value the Promise was rejected with. use .state if you need to be able to tell the difference.state: one of "pending", "fulfilled" or "rejected"And the following methods:
case({fulfilled, rejected, pending}): maps over the result using the provided handlers, or returns undefined if a handler isn't available for the current promise state.then((value: TValue) => TResult1 | PromiseLike<TResult1>, [(rejectReason: any) => any]): chains additional handlers to the provided promise.The returned object implements PromiseLike<TValue>, so you can chain additional Promise handlers using then. You may also use it with await in async functions.
Note that the status strings are available as constants:
mobxUtils.PENDING, mobxUtils.REJECTED, mobxUtil.FULFILLED
fromPromise takes an optional second argument, a previously created fromPromise based observable.
This is useful to replace one promise based observable with another, without going back to an intermediate
"pending" promise state while fetching data. For example:
origPromiseoldPromise IThenable<T> ? The previously observed promisepromise IThenable<T> The promise which will be observed@observer
class SearchResults extends React.Component {
@observable.ref searchResults
componentDidUpdate(nextProps) {
if (nextProps.query !== this.props.query)
this.searchResults = fromPromise(
window.fetch("/search?q=" + nextProps.query),
// by passing, we won't render a pending state if we had a successful search query before
// rather, we will keep showing the previous search results, until the new promise resolves (or rejects)
this.searchResults
)
}
render() {
return this.searchResults.case({
pending: (staleValue) => {
return staleValue || "searching" // <- value might set to previous results while the promise is still pending
},
fulfilled: (value) => {
return value // the fresh results
},
rejected: (error) => {
return "Oops: " + error
}
})
}
}
Observable promises can be created immediately in a certain state using
`fromPromise.reject(reason)` or `fromPromise.resolve(value?)`.
The main advantage of `fromPromise.resolve(value)` over `fromPromise(Promise.resolve(value))` is that the first _synchronously_ starts in the desired state.
It is possible to directly create a promise using a resolve, reject function:
`fromPromise((resolve, reject) => setTimeout(() => resolve(true), 1000))`
const fetchResult = fromPromise(fetch("http://someurl"))
// combine with when..
when(
() => fetchResult.state !== "pending",
() => {
console.log("Got ", fetchResult.value)
}
)
// or a mobx-react component..
const myComponent = observer(({ fetchResult }) => {
switch(fetchResult.state) {
case "pending": return <div>Loading...</div>
case "rejected": return <div>Ooops... {fetchResult.value}</div>
case "fulfilled": return <div>Gotcha: {fetchResult.value}</div>
}
})
// or using the case method instead of switch:
const myComponent = observer(({ fetchResult }) =>
fetchResult.case({
pending: () => <div>Loading...</div>,
rejected: error => <div>Ooops.. {error}</div>,
fulfilled: value => <div>Gotcha: {value}</div>,
}))
// chain additional handler(s) to the resolve/reject:
fetchResult.then(
(result) => doSomeTransformation(result),
(rejectReason) => console.error('fetchResult was rejected, reason: ' + rejectReason)
).then(
(transformedResult) => console.log('transformed fetchResult: ' + transformedResult)
)
Returns IPromiseBasedObservable<T>
Returns true if the provided value is a promise-based observable.
value anyReturns boolean
Moves an item from one position to another, checking that the indexes given are within bounds.
const source = observable([1, 2, 3])
moveItem(source, 0, 1)
console.log(source.map(x => x)) // [2, 1, 3]
Returns ObservableArray<T>
lazyObservable creates an observable around a fetch method that will not be invoked
until the observable is needed the first time.
The fetch method receives a sink callback which can be used to replace the
current value of the lazyObservable. It is allowed to call sink multiple times
to keep the lazyObservable up to date with some external resource.
Note that it is the current() call itself which is being tracked by MobX,
so make sure that you don't dereference to early.
fetchinitialValue T optional initialValue that will be returned from current as long as the sink has not been called at least once (optional, default undefined)const userProfile = lazyObservable(
sink => fetch("/myprofile").then(profile => sink(profile))
)
// use the userProfile in a React component:
const Profile = observer(({ userProfile }) =>
userProfile.current() === undefined
? <div>Loading user profile...</div>
: <div>{userProfile.current().displayName}</div>
)
// triggers refresh the userProfile
userProfile.refresh()
fromResource creates an observable whose current state can be inspected using .current(),
and which can be kept in sync with some external datasource that can be subscribed to.
The created observable will only subscribe to the datasource if it is in use somewhere,
(un)subscribing when needed. To enable fromResource to do that two callbacks need to be provided,
one to subscribe, and one to unsubscribe. The subscribe callback itself will receive a sink callback, which can be used
to update the current state of the observable, allowing observes to react.
Whatever is passed to sink will be returned by current(). The values passed to the sink will not be converted to
observables automatically, but feel free to do so.
It is the current() call itself which is being tracked,
so make sure that you don't dereference to early.
For inspiration, an example integration with the apollo-client on github,
or the implementation of mobxUtils.now
The following example code creates an observable that connects to a dbUserRecord,
which comes from an imaginary database and notifies when it has changed.
subscriberunsubscriber IDisposer (optional, default NOOP)initialValue T the data that will be returned by get() until the sink has emitted its first data (optional, default undefined)function createObservableUser(dbUserRecord) {
let currentSubscription;
return fromResource(
(sink) => {
// sink the current state
sink(dbUserRecord.fields)
// subscribe to the record, invoke the sink callback whenever new data arrives
currentSubscription = dbUserRecord.onUpdated(() => {
sink(dbUserRecord.fields)
})
},
() => {
// the user observable is not in use at the moment, unsubscribe (for now)
dbUserRecord.unsubscribe(currentSubscription)
}
)
}
// usage:
const myUserObservable = createObservableUser(myDatabaseConnector.query("name = 'Michel'"))
// use the observable in autorun
autorun(() => {
// printed everytime the database updates its records
console.log(myUserObservable.current().displayName)
})
// ... or a component
const userComponent = observer(({ user }) =>
<div>{user.current().displayName}</div>
)
Converts an expression to an observable stream (a.k.a. TC 39 Observable / RxJS observable). The provided expression is tracked by mobx as long as there are subscribers, automatically emitting when new values become available. The expressions respect (trans)actions.
expressionfireImmediately boolean (by default false)const user = observable({
firstName: "C.S",
lastName: "Lewis"
})
Rx.Observable
.from(mobxUtils.toStream(() => user.firstname + user.lastName))
.scan(nameChanges => nameChanges + 1, 0)
.subscribe(nameChanges => console.log("Changed name ", nameChanges, "times"))
Returns IObservableStream<T>
createViewModel takes an object with observable properties (model)
and wraps a viewmodel around it. The viewmodel proxies all enumerable properties of the original model with the following behavior:
submit() is called.The viewmodel exposes the following additional methods, besides all the enumerable properties of the model:
submit(): copies all the values of the viewmodel to the model and resets the statereset(): resets the state of the viewmodel, abandoning all local modificationsresetProperty(propName): resets the specified property of the viewmodelisDirty: observable property indicating if the viewModel contains any modificationsisPropertyDirty(propName): returns true if the specified property is dirtychangedValues: returns a key / value map with the properties that have been changed in the model so farmodel: The original model object for which this viewModel was createdYou may use observable arrays, maps and objects with createViewModel but keep in mind to assign fresh instances of those to the viewmodel's properties, otherwise you would end up modifying the properties of the original model.
Note that if you read a non-dirty property, viewmodel only proxies the read to the model. You therefore need to assign a fresh instance not only the first time you make the assignment but also after calling reset() or submit().
model Tclass Todo {
@observable title = "Test"
}
const model = new Todo()
const viewModel = createViewModel(model);
autorun(() => console.log(viewModel.model.title, ",", viewModel.title))
// prints "Test, Test"
model.title = "Get coffee"
// prints "Get coffee, Get coffee", viewModel just proxies to model
viewModel.title = "Get tea"
// prints "Get coffee, Get tea", viewModel's title is now dirty, and the local value will be printed
viewModel.submit()
// prints "Get tea, Get tea", changes submitted from the viewModel to the model, viewModel is proxying again
viewModel.title = "Get cookie"
// prints "Get tea, Get cookie" // viewModel has diverged again
viewModel.reset()
// prints "Get tea, Get tea", changes of the viewModel have been abandoned
MobX normally suspends any computed value that is not in use by any reaction,
and lazily re-evaluates the expression if needed outside a reaction while not in use.
keepAlive marks a computed value as always in use, meaning that it will always fresh, but never disposed automatically.
_1_2target Object an object that has a computed property, created by @computed or extendObservableproperty string the name of the property to keep aliveconst obj = observable({
number: 3,
doubler: function() { return this.number * 2 }
})
const stop = keepAlive(obj, "doubler")
Returns IDisposer stops this keep alive so that the computed value goes back to normal behavior
_1_2computedValue IComputedValue<any> created using the computed functionconst number = observable(3)
const doubler = computed(() => number.get() * 2)
const stop = keepAlive(doubler)
// doubler will now stay in sync reactively even when there are no further observers
stop()
// normal behavior, doubler results will be recomputed if not observed but needed, but lazily
Returns IDisposer stops this keep alive so that the computed value goes back to normal behavior
queueProcessor takes an observable array, observes it and calls processor
once for each item added to the observable array, optionally debouncing the action
observableArray Array<T> observable array instance to trackprocessordebounce number optional debounce time in ms. With debounce 0 the processor will run synchronously (optional, default 0)const pendingNotifications = observable([])
const stop = queueProcessor(pendingNotifications, msg => {
// show Desktop notification
new Notification(msg);
})
// usage:
pendingNotifications.push("test!")
Returns IDisposer stops the processor
chunkProcessor takes an observable array, observes it and calls processor
once for a chunk of items added to the observable array, optionally deboucing the action.
The maximum chunk size can be limited by number.
This allows both, splitting larger into smaller chunks or (when debounced) combining smaller
chunks and/or single items into reasonable chunks of work.
observableArray Array<T> observable array instance to trackprocessordebounce number optional debounce time in ms. With debounce 0 the processor will run synchronously (optional, default 0)maxChunkSize number optionally do not call on full array but smaller chunks. With 0 it will process the full array. (optional, default 0)const trackedActions = observable([])
const stop = chunkProcessor(trackedActions, chunkOfMax10Items => {
sendTrackedActionsToServer(chunkOfMax10Items);
}, 100, 10)
// usage:
trackedActions.push("scrolled")
trackedActions.push("hoveredButton")
// when both pushes happen within 100ms, there will be only one call to server
Returns IDisposer stops the processor
Returns the current date time as epoch number. The date time is read from an observable which is updated automatically after the given interval. So basically it treats time as an observable.
The function takes an interval as parameter, which indicates how often now() will return a new value.
If no interval is given, it will update each second. If "frame" is specified, it will update each time a
requestAnimationFrame is available.
Multiple clocks with the same interval will automatically be synchronized.
Countdown example: https://jsfiddle.net/mweststrate/na0qdmkw/
interval (number | "frame") interval in milliseconds about how often the interval should update (optional, default 1000)const start = Date.now()
autorun(() => {
console.log("Seconds elapsed: ", (mobxUtils.now() - start) / 1000)
})
expr can be used to create temporary computed values inside computed values.
Nesting computed values is useful to create cheap computations in order to prevent expensive computations from needing to run.
In the following example the expression prevents that a component is rerender each time the selection changes;
instead it will only rerenders when the current todo is (de)selected.
expr(func) is an alias for computed(func).get().
Please note that the function given to expr is evaluated twice in the scenario that the overall expression value changes.
It is evaluated the first time when any observables it depends on change.
It is evaluated a second time when a change in its value triggers the outer computed or reaction to evaluate, which recreates and reevaluates the expression.
In the following example, the expression prevents the TodoView component from being re-rendered if the selection changes elsewhere.
Instead, the component will only re-render when the relevant todo is (de)selected, which happens much less frequently.
exprconst Todo = observer((props) => {
const todo = props.todo
const isSelected = mobxUtils.expr(() => props.viewState.selection === todo)
const TodoView = observer(({ todo, editorState }) => {
const isSelected = mobxUtils.expr(() => editorState.selection === todo)
return <div className={isSelected ? "todo todo-selected" : "todo"}>{todo.title}</div>
})
Creates a function that maps an object to a view. The mapping is memoized.
See the transformer section for more details.
transformerarg2onCleanupGiven an object, deeply observes the given object.
It is like observe from mobx, but applied recursively, including all future children.
Note that the given object cannot ever contain cycles and should be a tree.
As benefit: path and root will be provided in the callback, so the signature of the listener is (change, path, root) => void
The returned disposer can be invoked to clean up the listener
deepObserve cannot be used on computed values.
targetlistenerconst disposer = deepObserve(target, (change, path) => {
console.dir(change)
})
Reactively sorts a base observable array into multiple observable arrays based on the value of a
groupBy: (item: T) => G function.
This observes the individual computed groupBy values and only updates the source and dest arrays
when there is an actual change, so this is far more efficient than, for example
base.filter(i => groupBy(i) === 'we'). Call #dispose() to stop tracking.
No guarantees are made about the order of items in the grouped arrays.
The resulting map of arrays is read-only. clear(), set(), delete() are not supported and modifying the group arrays will lead to undefined behavior.
NB: ObservableGroupMap relies on Symbols. If you are targeting a platform which doesn't
support these natively, you will need to provide a polyfill.
base array The array to sort into groups.groupBy function The function used for grouping.options Object with properties:
name: Debug name of this ObservableGroupMap.
keyToName: Function to create the debug names of the observable group arrays.const slices = observable([
{ day: "mo", hours: 12 },
{ day: "tu", hours: 2 },
])
const slicesByDay = new ObservableGroupMap(slices, (slice) => slice.day)
autorun(() => console.log(
slicesByDay.get("mo")?.length ?? 0,
slicesByDay.get("we"))) // outputs 1, undefined
slices[0].day = "we" // outputs 0, [{ day: "we", hours: 12 }]
Base observable array which is being sorted into groups.
The ObservableGroupMap needs to track some state per-item. This is the name/symbol of the property used to attach the state.
The function used to group the items.
This function is used to generate the mobx debug names of the observable group arrays.
Disposes all observers created during construction and removes state added to base array items.
computedFn takes a function with an arbitrary amount of arguments, and memoizes the output of the function based on the arguments passed in.
computedFn(fn) returns a function with the very same signature. There is no limit on the amount of arguments that is accepted. However, the amount of arguments must be constant and default arguments are not supported.
By default the output of a function call will only be memoized as long as the output is being observed.
The function passes into computedFn should be pure, not be an action and only be relying on
observables.
Setting keepAlive to true will cause the output to be forcefully cached forever.
Note that this might introduce memory leaks!
fnkeepAliveOrOptionsconst store = observable({
a: 1,
b: 2,
c: 3,
m: computedFn(function(x) {
return this.a * this.b * x
})
})
const d = autorun(() => {
// store.m(3) will be cached as long as this autorun is running
console.log(store.m(3) * store.c)
})
With createTransformer it is very easy to transform a complete data graph into another data graph.
Transformation functions can be composed so that you can build a tree using lots of small transformations.
The resulting data graph will never be stale, it will be kept in sync with the source by applying small patches to the result graph.
This makes it very easy to achieve powerful patterns similar to sideways data loading, map-reduce, tracking state history using immutable data structures etc.
createTransformer turns a function (that should transform value A into another value B) into a reactive and memoizing function.
In other words, if the transformation function computes B given a specific A, the same B will be returned for all other future invocations of the transformation with the same A.
However, if A changes, the transformation will be re-applied so that B is updated accordingly.
And last but not least, if nobody is using the transformation of a specific A anymore, its entry will be removed from the memoization table.
The optional onCleanup function can be used to get a notification when a transformation of an object is no longer needed.
This can be used to dispose resources attached to the result object if needed.
Always use transformations inside a reaction like observer or autorun.
Transformations will, like any other computed value, fall back to lazy evaluation if not observed by something, which sort of defeats their purpose.
onCleanup?: (result: B, value?: A) => void)createTransformer<A, B>(transformation: (value: A) => B, onCleanup?: (result: B, value?: A) => void): (value: A) => B
This all might still be a bit vague, so here are two examples that explain this whole idea of transforming one data structure into another by using small, reactive functions:
This example is taken from the Reactive2015 conference demo:
/*
The store that holds our domain: boxes and arrows
*/
const store = observable({
boxes: [],
arrows: [],
selection: null,
})
/**
Serialize store to json upon each change and push it onto the states list
*/
const states = []
autorun(() => {
states.push(serializeState(store))
})
const serializeState = createTransformer((store) => ({
boxes: store.boxes.map(serializeBox),
arrows: store.arrows.map(serializeArrow),
selection: store.selection ? store.selection.id : null,
}))
const serializeBox = createTransformer((box) => ({ ...box }))
const serializeArrow = createTransformer((arrow) => ({
id: arrow.id,
to: arrow.to.id,
from: arrow.from.id,
}))
In this example the state is serialized by composing three different transformation functions.
The autorunner triggers the serialization of the store object, which in turn serializes all boxes and arrows.
Let's take closer look at the life of an imaginary example box#3.
map to serializeBox,
the serializeBox transformation is executed and an entry containing box#3 and its serialized representation is added to the internal memoization table of serializeBox.store.boxes list.
This would cause the serializeState function to re-compute, resulting in a complete remapping of all the boxes.
However, all the invocations of serializeBox will now return their old values from the memoization tables since their transformation functions didn't (need to) run again.serializeBox to box#3 to re-compute, just like any other reactive function in MobX.
Since the transformation will now produce a new Json object based on box#3, all observers of that specific transformation will be forced to run again as well.
That's the serializeState transformation in this case.
serializeState will now produce a new value in turn and map all the boxes again. But except for box#3, all other boxes will be returned from the memoization table.store.boxes, serializeState will compute again.
But since it will no longer be using the application of serializeBox to box#3,
that reactive function will go back to non-reactive mode.
This signals the memoization table that the entry can be removed so that it is ready for GC.So effectively we have achieved state tracking using immutable, shared datas structures here.
All boxes and arrows are mapped and reduced into single state tree.
Each change will result in a new entry in the states array, but the different entries will share almost all of their box and arrow representations.
Instead of returning plain values from a transformation function, it is also possible to return observable objects. This can be used to transform an observable data graph into a another observable data graph, which can be used to transform... you get the idea.
Here is a small example that encodes a reactive file explorer that will update its representation upon each change. Data graphs that are built this way will in general react a lot faster and will consist of much more straight-forward code, compared to derived data graph that are updated using your own code. See the performance tests for some examples.
Unlike the previous example, the transformFolder will only run once as long as a folder remains visible;
the DisplayFolder objects track the associated Folder objects themselves.
In the following example all mutations to the state graph will be processed automatically.
Some examples:
path property and the path property of all its descendants.DisplayFolders from the tree.import {extendObservable, observable, createTransformer, autorun} from "mobx"
function Folder(parent, name) {
this.parent = parent;
extendObservable(this, {
name: name,
children: observable.shallow([]),
});
}
function DisplayFolder(folder, state) {
this.state = state;
this.folder = folder;
extendObservable(this, {
collapsed: false,
get name() {
return this.folder.name;
},
get isVisible() {
return !this.state.filter || this.name.indexOf(this.state.filter) !== -1 || this.children.some(child => child.isVisible);
},
get children() {
if (this.collapsed)
return [];
return this.folder.children.map(transformFolder).filter(function(child) {
return child.isVisible;
})
},
get path() {
return this.folder.parent === null ? this.name : transformFolder(this.folder.parent).path + "/" + this.name;
})
});
}
var state = observable({
root: new Folder(null, "root"),
filter: null,
displayRoot: null
});
var transformFolder = createTransformer(function (folder) {
return new DisplayFolder(folder, state);
});
// returns list of strings per folder
var stringTransformer = createTransformer(function (displayFolder) {
var path = displayFolder.path;
return path + "\n" +
displayFolder.children.filter(function(child) {
return child.isVisible;
}).map(stringTransformer).join('');
});
function createFolders(parent, recursion) {
if (recursion === 0)
return;
for (var i = 0; i < 3; i++) {
var folder = new Folder(parent, i + '');
parent.children.push(folder);
createFolders(folder, recursion - 1);
}
}
createFolders(state.root, 2); // 3^2
autorun(function() {
state.displayRoot = transformFolder(state.root);
state.text = stringTransformer(state.displayRoot)
console.log(state.text)
});
state.root.name = 'wow'; // change folder name
state.displayRoot.children[1].collapsed = true; // collapse folder
state.filter = "2"; // search
state.filter = null; // unsearch
FAQs
Utility functions and common patterns for MobX
The npm package mobx-utils receives a total of 0 weekly downloads. As such, mobx-utils popularity was classified as not popular.
We found that mobx-utils demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 4 open source maintainers 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
Elastic’s return to open source with the AGPL license has been met with skepticism, as many developers see it as a strategic move rather than a genuine effort to restore user trust and freedoms.
Security News
A new "revival hijack" supply chain attack targets deleted Python packages, with an estimated 22K packages at risk. Socket can detect and block hijacked packages that have added malicious code.
Product
We're introducing a new Analytics feature in the Socket dashboard so you can view changes in your organization's and repositories' alerts over time.