Template Syntax

The Angular application manages what the user sees and can do, achieving this through the interaction of a component class instance (the component) and its user-facing template.

You may be familiar with the component/template duality from your experience with model-view-controller (MVC) or model-view-viewmodel (MVVM). In Angular, the component plays the part of the controller/viewmodel, and the template represents the view.

This page is a comprehensive technical reference to the Angular template language. It explains basic principles of the template language and describes most of the syntax that you'll encounter elsewhere in the documentation.

Many code snippets illustrate the points and concepts, all of them available in the .

HTML in templates

HTML is the language of the Angular template. Almost all HTML syntax is valid template syntax. The <script> element is a notable exception; it is forbidden, eliminating the risk of script injection attacks. In practice, <script> is ignored and a warning appears in the browser console. See the Security page for details.

Some legal HTML doesn't make much sense in a template. The <html>, <body>, and <base> elements have no useful role. Pretty much everything else is fair game.

You can extend the HTML vocabulary of your templates with components and directives that appear as new elements and attributes. In the following sections, you'll learn how to get and set DOM (Document Object Model) values dynamically through data binding.

Begin with the first form of data binding—interpolation—to see how much richer template HTML can be.

Interpolation and Template Expressions

Interpolation allows you to incorporate calculated strings into the text between HTML element tags and within attribute assignments. Template expressions are what you use to calculate those strings.

The interpolation live example demonstrates all of the syntax and code snippets described in this section.

Interpolation `{{...}}`

Interpolation refers to embedding expressions into marked up text. By default, interpolation uses as its delimiter the double curly braces, {{ and }}.

In the following snippet, {{ currentCustomer }} is an example of interpolation.

<h3>Current customer: {{ currentCustomer }}</h3>

The text between the braces is often the name of a component property. Angular replaces that name with the string value of the corresponding component property.

<p>{{title}}</p>
<div><img src="{{itemImageUrl}}"></div>

In the example above, Angular evaluates the title and itemImageUrl properties and fills in the blanks, first displaying some title text and then an image.

More generally, the text between the braces is a template expression that Angular first evaluates and then converts to a string. The following interpolation illustrates the point by adding two numbers:

<!-- "The sum of 1 + 1 is 2" -->
<p>The sum of 1 + 1 is {{1 + 1}}.</p>

The expression can invoke methods of the host component such as getVal() in the following example:

<!-- "The sum of 1 + 1 is not 4" -->
<p>The sum of 1 + 1 is not {{1 + 1 + getVal()}}.</p>

Angular evaluates all expressions in double curly braces, converts the expression results to strings, and links them with neighboring literal strings. Finally, it assigns this composite interpolated result to an element or directive property.

You appear to be inserting the result between element tags and assigning it to attributes.

However, interpolation is a special syntax that Angular converts into a property binding.

If you'd like to use something other than {{ and }}, you can configure the interpolation delimiter via the interpolation option in the Component metadata.

Template expressions

A template expression produces a value and appears within the double curly braces, {{ }}. Angular executes the expression and assigns it to a property of a binding target; the target could be an HTML element, a component, or a directive.

The interpolation braces in {{1 + 1}} surround the template expression 1 + 1. In the property binding, a template expression appears in quotes to the right of the = symbol as in [property]="expression".

In terms of syntax, template expressions are similar to JavaScript. Many JavaScript expressions are legal template expressions, with a few exceptions.

You can't use JavaScript expressions that have or promote side effects, including:

Assignments (=, +=, -=, ...)
Operators such as new, typeof, instanceof, etc.
Chaining expressions with ; or ,
The increment and decrement operators ++ and --
Some of the ES2015+ operators

Other notable differences from JavaScript syntax include:

No support for the bitwise operators such as | and &
New template expression operators, such as |, ?. and !

Expression context

The expression context is typically the component instance. In the following snippets, the recommended within double curly braces and the itemImageUrl2 in quotes refer to properties of the AppComponent.

<h4>{{recommended}}</h4>
<img [src]="itemImageUrl2">

An expression may also refer to properties of the template's context such as a template input variable,

let customer, or a template reference variable, #customerInput.

<ul>
  <li *ngFor="let customer of customers">{{customer.name}}</li>
</ul>

<label>Type something:
  <input #customerInput>{{customerInput.value}}
</label>

The context for terms in an expression is a blend of the template variables, the directive's context object (if it has one), and the component's members. If you reference a name that belongs to more than one of these namespaces, the template variable name takes precedence, followed by a name in the directive's context, and, lastly, the component's member names.

The previous example presents such a name collision. The component has a customer property and the *ngFor defines a customer template variable.

The customer in {{customer.name}} refers to the template input variable, not the component's property.

Template expressions cannot refer to anything in the global namespace, except undefined. They can't refer to window or document. Additionally, they can't call console.log() or Math.max() and they are restricted to referencing members of the expression context.

Expression guidelines

When using template expressions follow these guidelines:

No visible side effects

A template expression should not change any application state other than the value of the target property.

This rule is essential to Angular's "unidirectional data flow" policy. You should never worry that reading a component value might change some other displayed value. The view should be stable throughout a single rendering pass.

An idempotent expression is ideal because it is free of side effects and improves Angular's change detection performance.

In Angular terms, an idempotent expression always returns exactly the same thing until one of its dependent values changes.

Dependent values should not change during a single turn of the event loop. If an idempotent expression returns a string or a number, it returns the same string or number when called twice in a row. If the expression returns an object, including an array, it returns the same object reference when called twice in a row.

There is one exception to this behavior that applies to *ngFor. *ngFor has trackBy functionality that can deal with referential inequality of objects that when iterating over them.

For more information, see the *ngFor with trackBy section of this guide.

Quick execution

Angular executes template expressions after every change detection cycle. Change detection cycles are triggered by many asynchronous activities such as promise resolutions, HTTP results, timer events, key presses and mouse moves.

Expressions should finish quickly or the user experience may drag, especially on slower devices. Consider caching values when their computation is expensive.

Simplicity

Although it's possible to write complex template expressions, it's a better practice to avoid them.

A property name or method call should be the norm, but an occasional Boolean negation, !, is OK. Otherwise, confine application and business logic to the component, where it is easier to develop and test.

Template statements

A template statement responds to an event raised by a binding target such as an element, component, or directive. You'll see template statements in the event binding section, appearing in quotes to the right of the = symbol as in (event)="statement".

<button (click)="deleteHero()">Delete hero</button>

A template statement has a side effect. That's the whole point of an event. It's how you update application state from user action.

Responding to events is the other side of Angular's "unidirectional data flow". You're free to change anything, anywhere, during this turn of the event loop.

Like template expressions, template statements use a language that looks like JavaScript. The template statement parser differs from the template expression parser and specifically supports both basic assignment (=) and chaining expressions (with ; or ,).

However, certain JavaScript syntax is not allowed:

new
increment and decrement operators, ++ and --
operator assignment, such as += and -=
the bitwise operators | and &
the template expression operators

Statement context

As with expressions, statements can refer only to what's in the statement context such as an event handling method of the component instance.

The statement context is typically the component instance. The deleteHero in (click)="deleteHero()" is a method of the data-bound component.

<button (click)="deleteHero()">Delete hero</button>

The statement context may also refer to properties of the template's own context. In the following examples, the template $event object, a template input variable (let hero), and a template reference variable (#heroForm) are passed to an event handling method of the component.

<button (click)="onSave($event)">Save</button>
<button *ngFor="let hero of heroes" (click)="deleteHero(hero)">{{hero.name}}</button>
<form #heroForm (ngSubmit)="onSubmit(heroForm)"> ... </form>

Template context names take precedence over component context names. In deleteHero(hero) above, the hero is the template input variable, not the component's hero property.

Template statements cannot refer to anything in the global namespace. They can't refer to window or document. They can't call console.log or Math.max.

Statement guidelines

As with expressions, avoid writing complex template statements. A method call or simple property assignment should be the norm.

Now that you have a feel for template expressions and statements, you're ready to learn about the varieties of data binding syntax beyond interpolation.

Binding syntax: An overview

Data binding is a mechanism for coordinating what users see, with application data values. While you could push values to and pull values from HTML, the application is easier to write, read, and maintain if you turn these chores over to a binding framework. You simply declare bindings between binding sources and target HTML elements and let the framework do the work.

Angular provides many kinds of data binding. This guide covers most of them, after a high-level view of Angular data binding and its syntax.

Binding types can be grouped into three categories distinguished by the direction of data flow: from the source-to-view, from view-to-source, and in the two-way sequence: view-to-source-to-view:

Data direction	Syntax	Type
One-wayfrom data sourceto view target	{{expression}} [target]="expression" bind-target="expression"	Interpolation Property Attribute Class Style
One-wayfrom view targetto data source	(target)="statement" on-target="statement"	Event
Two-way	[(target)]="expression" bindon-target="expression"	Two-way

Binding types other than interpolation have a target name to the left of the equal sign, either surrounded by punctuation ([], ()) or preceded by a prefix (bind-, on-, bindon-).

The target name is the name of a property. It may look like the name of an attribute but it never is. To appreciate the difference, you must develop a new way to think about template HTML.

A new mental model

With all the power of data binding and the ability to extend the HTML vocabulary with custom markup, it is tempting to think of template HTML as HTML Plus.

It really is HTML Plus. But it's also significantly different than the HTML you're used to. It requires a new mental model.

In the normal course of HTML development, you create a visual structure with HTML elements, and you modify those elements by setting element attributes with string constants.

<div class="special">Mental Model</div>
<img src="assets/images/hero.png">
<button disabled>Save</button>

You still create a structure and initialize attribute values this way in Angular templates.

Then you learn to create new elements with components that encapsulate HTML and drop them into templates as if they were native HTML elements.

<!-- Normal HTML -->
<div class="special">Mental Model</div>
<!-- Wow! A new element! -->
<app-hero-detail></app-hero-detail>

That's HTML Plus.

Then you learn about data binding. The first binding you meet might look like this:

<!-- Bind button disabled state to `isUnchanged` property -->
<button [disabled]="isUnchanged">Save</button>

You'll get to that peculiar bracket notation in a moment. Looking beyond it, your intuition suggests that you're binding to the button's disabled attribute and setting it to the current value of the component's isUnchanged property.

Your intuition is incorrect! Your everyday HTML mental model is misleading. In fact, once you start data binding, you are no longer working with HTML attributes. You aren't setting attributes. You are setting the properties of DOM elements, components, and directives.

HTML attribute vs. DOM property

The distinction between an HTML attribute and a DOM property is crucial to understanding how Angular binding works.

Attributes are defined by HTML. Properties are defined by the DOM (Document Object Model).

A few HTML attributes have 1:1 mapping to properties. id is one example.

Some HTML attributes don't have corresponding properties. colspan is one example.

Some DOM properties don't have corresponding attributes. textContent is one example.

Many HTML attributes appear to map to properties ... but not in the way you might think!

That last category is confusing until you grasp this general rule:

Attributes initialize DOM properties and then they are done. Property values can change; attribute values can't.

For example, when the browser renders <input type="text" value="Bob">, it creates a corresponding DOM node with a value property initialized to "Bob".

When the user enters "Sally" into the input box, the DOM element value property becomes "Sally". But the HTML value attribute remains unchanged as you discover if you ask the input element about that attribute: input.getAttribute('value') returns "Bob".

The HTML attribute value specifies the initial value; the DOM value property is the current value.

The disabled attribute is another peculiar example. A button's disabled property is false by default so the button is enabled. When you add the disabled attribute, its presence alone initializes the button's disabled property to true so the button is disabled.

Adding and removing the disabled attribute disables and enables the button. The value of the attribute is irrelevant, which is why you cannot enable a button by writing <button disabled="false">Still Disabled</button>.

Setting the button's disabled property (say, with an Angular binding) disables or enables the button. The value of the property matters.

The HTML attribute and the DOM property are not the same thing, even when they have the same name.

This fact bears repeating: Template binding works with properties and events, not attributes.

A world without attributes

In the world of Angular, the only role of attributes is to initialize element and directive state. When you write a data binding, you're dealing exclusively with properties and events of the target object. HTML attributes effectively disappear.

With this model firmly in mind, read on to learn about binding targets.

Binding targets

The target of a data binding is something in the DOM. Depending on the binding type, the target can be an (element | component | directive) property, an (element | component | directive) event, or (rarely) an attribute name. The following table summarizes:

Type	Target	Examples
Property	Element property Component property Directive property	<img [src]="heroImageUrl"> <app-hero-detail [hero]="currentHero"></app-hero-detail> <div [ngClass]="{'special': isSpecial}"></div>
Event	Element event Component event Directive event	<button (click)="onSave()">Save</button> <app-hero-detail (deleteRequest)="deleteHero()"></app-hero-detail> <div (myClick)="clicked=$event" clickable>click me</div>
Two-way	Event and property	<input [(ngModel)]="name">
Attribute	Attribute (the exception)	<button [attr.aria-label]="help">help</button>
Class	`class` property	<div [class.special]="isSpecial">Special</div>
Style	`style` property	<button [style.color]="isSpecial ? 'red' : 'green'">

With this broad view in mind, you're ready to look at binding types in detail.

Property binding ( [property] )

Write a template property binding to set a property of a view element. The binding sets the property to the value of a template expression.

The most common property binding sets an element property to a component property value. An example is binding the src property of an image element to a component's heroImageUrl property:

<img [src]="heroImageUrl">

Another example is disabling a button when the component says that it isUnchanged:

<button [disabled]="isUnchanged">Cancel is disabled</button>

Another is setting a property of a directive:

<div [ngClass]="classes">[ngClass] binding to the classes property</div>

Yet another is setting the model property of a custom component (a great way for parent and child components to communicate):

<app-hero-detail [hero]="currentHero"></app-hero-detail>

One-way `in`

People often describe property binding as one-way data binding because it flows a value in one direction, from a component's data property into a target element property.

You cannot use property binding to pull values out of the target element. You can't bind to a property of the target element to read it. You can only set it.

Similarly, you cannot use property binding to call a method on the target element.

If the element raises events, you can listen to them with an event binding.

If you must read a target element property or call one of its methods, you'll need a different technique. See the API reference for ViewChild and ContentChild.

Binding target

An element property between enclosing square brackets identifies the target property. The target property in the following code is the image element's src property.

<img [src]="heroImageUrl">

Some people prefer the bind- prefix alternative, known as the canonical form:

<img bind-src="heroImageUrl">

The target name is always the name of a property, even when it appears to be the name of something else. You see src and may think it's the name of an attribute. No. It's the name of an image element property.

Element properties may be the more common targets, but Angular looks first to see if the name is a property of a known directive, as it is in the following example:

<div [ngClass]="classes">[ngClass] binding to the classes property</div>

Technically, Angular is matching the name to a directive input, one of the property names listed in the directive's inputs array or a property decorated with @Input(). Such inputs map to the directive's own properties.

If the name fails to match a property of a known directive or element, Angular reports an “unknown directive” error.

Avoid side effects

As mentioned previously, evaluation of a template expression should have no visible side effects. The expression language itself does its part to keep you safe. You can't assign a value to anything in a property binding expression nor use the increment and decrement operators.

Of course, the expression might invoke a property or method that has side effects. Angular has no way of knowing that or stopping you.

The expression could call something like getFoo(). Only you know what getFoo() does. If getFoo() changes something and you happen to be binding to that something, you risk an unpleasant experience. Angular may or may not display the changed value. Angular may detect the change and throw a warning error. In general, stick to data properties and to methods that return values and do no more.

Return the proper type

The template expression should evaluate to the type of value expected by the target property. Return a string if the target property expects a string. Return a number if the target property expects a number. Return an object if the target property expects an object.

The hero property of the HeroDetail component expects a Hero object, which is exactly what you're sending in the property binding:

<app-hero-detail [hero]="currentHero"></app-hero-detail>

Remember the brackets

The brackets tell Angular to evaluate the template expression. If you omit the brackets, Angular treats the string as a constant and initializes the target property with that string. It does not evaluate the string!

Don't make the following mistake:

<!-- ERROR: HeroDetailComponent.hero expects a
     Hero object, not the string "currentHero" -->
  <app-hero-detail hero="currentHero"></app-hero-detail>

One-time string initialization

You should omit the brackets when all of the following are true:

The target property accepts a string value.
The string is a fixed value that you can bake into the template.
This initial value never changes.

You routinely initialize attributes this way in standard HTML, and it works just as well for directive and component property initialization. The following example initializes the prefix property of the HeroDetailComponent to a fixed string, not a template expression. Angular sets it and forgets about it.

<app-hero-detail prefix="You are my" [hero]="currentHero"></app-hero-detail>

The [hero] binding, on the other hand, remains a live binding to the component's currentHero property.

Property binding or interpolation?

You often have a choice between interpolation and property binding. The following binding pairs do the same thing:

<p><img src="{{heroImageUrl}}"> is the <i>interpolated</i> image.</p>
<p><img [src]="heroImageUrl"> is the <i>property bound</i> image.</p>

<p><span>"{{title}}" is the <i>interpolated</i> title.</span></p>
<p>"<span [innerHTML]="title"></span>" is the <i>property bound</i> title.</p>

Interpolation is a convenient alternative to property binding in many cases.

When rendering data values as strings, there is no technical reason to prefer one form to the other. You lean toward readability, which tends to favor interpolation. You suggest establishing coding style rules and choosing the form that both conforms to the rules and feels most natural for the task at hand.

When setting an element property to a non-string data value, you must use property binding.

Content security

Imagine the following malicious content.

evilTitle = 'Template <script>alert("evil never sleeps")</script>Syntax';

Fortunately, Angular data binding is on alert for dangerous HTML. It sanitizes the values before displaying them. It will not allow HTML with script tags to leak into the browser, neither with interpolation nor property binding.

<!--
  Angular generates warnings for these two lines as it sanitizes them
  WARNING: sanitizing HTML stripped some content (see http://g.co/ng/security#xss).
 -->
<p><span>"{{evilTitle}}" is the <i>interpolated</i> evil title.</span></p>
<p>"<span [innerHTML]="evilTitle"></span>" is the <i>property bound</i> evil title.</p>

Interpolation handles the script tags differently than property binding but both approaches render the content harmlessly.

Attribute, class, and style bindings

The template syntax provides specialized one-way bindings for scenarios less well suited to property binding.

Attribute binding

You can set the value of an attribute directly with an attribute binding.

This is the only exception to the rule that a binding sets a target property. This is the only binding that creates and sets an attribute.

This guide stresses repeatedly that setting an element property with a property binding is always preferred to setting the attribute with a string. Why does Angular offer attribute binding?

You must use attribute binding when there is no element property to bind.

Consider the ARIA, SVG, and table span attributes. They are pure attributes. They do not correspond to element properties, and they do not set element properties. There are no property targets to bind to.

This fact becomes painfully obvious when you write something like this.

<tr><td colspan="{{1 + 1}}">Three-Four</td></tr>

And you get this error:

Template parse errors:
  Can't bind to 'colspan' since it isn't a known native property

As the message says, the <td> element does not have a colspan property. It has the "colspan" attribute, but interpolation and property binding can set only properties, not attributes.

You need attribute bindings to create and bind to such attributes.

Attribute binding syntax resembles property binding. Instead of an element property between brackets, start with the prefix attr, followed by a dot (.) and the name of the attribute. You then set the attribute value, using an expression that resolves to a string.

Bind [attr.colspan] to a calculated value:

<table border=1>
  <!--  expression calculates colspan=2 -->
  <tr><td [attr.colspan]="1 + 1">One-Two</td></tr>

  <!-- ERROR: There is no `colspan` property to set!
    <tr><td colspan="{{1 + 1}}">Three-Four</td></tr>
  -->

  <tr><td>Five</td><td>Six</td></tr>
</table>

Here's how the table renders:

One-Two
Five	Six

One of the primary use cases for attribute binding is to set ARIA attributes, as in this example:

<!-- create and set an aria attribute for assistive technology -->
<button [attr.aria-label]="actionName">{{actionName}} with Aria</button>

Class binding

You can add and remove CSS class names from an element's class attribute with a class binding.

Class binding syntax resembles property binding. Instead of an element property between brackets, start with the prefix class, optionally followed by a dot (.) and the name of a CSS class: [class.class-name].

The following examples show how to add and remove the application's "special" class with class bindings. Here's how to set the attribute without binding:

<!-- standard class attribute setting  -->
<div class="bad curly special">Bad curly special</div>

You can replace that with a binding to a string of the desired class names; this is an all-or-nothing, replacement binding.

<!-- reset/override all class names with a binding  -->
<div class="bad curly special"
     [class]="badCurly">Bad curly</div>

Finally, you can bind to a specific class name. Angular adds the class when the template expression evaluates to truthy. It removes the class when the expression is falsy.

<!-- toggle the "special" class on/off with a property -->
<div [class.special]="isSpecial">The class binding is special</div>

<!-- binding to `class.special` trumps the class attribute -->
<div class="special"
     [class.special]="!isSpecial">This one is not so special</div>

While this is a fine way to toggle a single class name, the NgClass directive is usually preferred when managing multiple class names at the same time.

Style binding

You can set inline styles with a style binding.

Style binding syntax resembles property binding. Instead of an element property between brackets, start with the prefix style, followed by a dot (.) and the name of a CSS style property: [style.style-property].

<button [style.color]="isSpecial ? 'red': 'green'">Red</button>
<button [style.background-color]="canSave ? 'cyan': 'grey'" >Save</button>

Some style binding styles have a unit extension. The following example conditionally sets the font size in “em” and “%” units .

<button [style.font-size.em]="isSpecial ? 3 : 1" >Big</button>
<button [style.font-size.%]="!isSpecial ? 150 : 50" >Small</button>

While this is a fine way to set a single style, the NgStyle directive is generally preferred when setting several inline styles at the same time.

Note that a style property name can be written in either dash-case, as shown above, or camelCase, such as fontSize.

Event binding `(event)`

Event binding allows you to listen for certain events such as keystrokes, mouse movements, clicks, and touches. For an example demonstrating all of the points in this section, see the event binding example.

Angular event binding syntax consists of a target event name within parentheses on the left of an equal sign, and a quoted template statement on the right. The following event binding listens for the button's click events, calling the component's onSave() method whenever a click occurs:

Target event

As above, the target is the button's click event.

<button (click)="onSave($event)">Save</button>

Alternatively, use the on- prefix, known as the canonical form:

<button on-click="onSave($event)">on-click Save</button>

Element events may be the more common targets, but Angular looks first to see if the name matches an event property of a known directive, as it does in the following example:

<h4>myClick is an event on the custom ClickDirective:</h4>
<button (myClick)="clickMessage=$event" clickable>click with myClick</button>
{{clickMessage}}

If the name fails to match an element event or an output property of a known directive, Angular reports an “unknown directive” error.

`$event` and event handling statements

In an event binding, Angular sets up an event handler for the target event.

When the event is raised, the handler executes the template statement. The template statement typically involves a receiver, which performs an action in response to the event, such as storing a value from the HTML control into a model.

The binding conveys information about the event. This information can include data values such as an event object, string, or number named $event.

The target event determines the shape of the $event object. If the target event is a native DOM element event, then $event is a DOM event object, with properties such as target and target.value.

Consider this example:

<input [value]="currentItem.name"
       (input)="currentItem.name=$event.target.value" >
without NgModel

This code sets the <input> value property by binding to the name property. To listen for changes to the value, the code binds to the input event of the <input> element. When the user makes changes, the input event is raised, and the binding executes the statement within a context that includes the DOM event object, $event.

To update the name property, the changed text is retrieved by following the path $event.target.value.

If the event belongs to a directive—recall that components are directives—$event has whatever shape the directive produces.

Custom events with `EventEmitter`

Directives typically raise custom events with an Angular EventEmitter. The directive creates an EventEmitter and exposes it as a property. The directive calls EventEmitter.emit(payload) to fire an event, passing in a message payload, which can be anything. Parent directives listen for the event by binding to this property and accessing the payload through the $event object.

Consider an ItemDetailComponent that presents item information and responds to user actions. Although the ItemDetailComponent has a delete button, it doesn't know how to delete the hero. It can only raise an event reporting the user's delete request.

Here are the pertinent excerpts from that ItemDetailComponent:

<img src="{{itemImageUrl}}" [style.display]="displayNone">
<span [style.text-decoration]="lineThrough">{{ item.name }}
</span>
<button (click)="delete()">Delete</button>

// This component makes a request but it can't actually delete a hero.
@Output() deleteRequest = new EventEmitter<Item>();

delete() {
  this.deleteRequest.emit(this.item.name);
  this.displayNone = this.displayNone ? '' : 'none';
  this.lineThrough = this.lineThrough ? '' : 'line-through';
}

The component defines a deleteRequest property that returns an EventEmitter. When the user clicks delete, the component invokes the delete() method, telling the EventEmitter to emit an Item object.

Now imagine a hosting parent component that binds to the deleteRequest event of the ItemDetailComponent.

<app-item-detail (deleteRequest)="deleteItem($event)" [item]="currentItem"></app-item-detail>

When the deleteRequest event fires, Angular calls the parent component's deleteItem() method, passing the item-to-delete (emitted by ItemDetail) in the $event variable.

Template statements have side effects

Though template expressions shouldn't have side effects, template statements usually do. The deleteItem() method does have a side effect: it deletes an item.

Deleting an item updates the model, and depending on your code, triggers other changes including queries and saving to a remote server. These changes propagate through the system and ultimately display in this and other views.

Two-way binding ( [(...)] )

You often want to both display a data property and update that property when the user makes changes.

On the element side that takes a combination of setting a specific element property and listening for an element change event.

Angular offers a special two-way data binding syntax for this purpose, [(x)]. The [(x)] syntax combines the brackets of property binding, [x], with the parentheses of event binding, (x).

[( )] = banana in a box

Visualize a banana in a box to remember that the parentheses go inside the brackets.

The [(x)] syntax is easy to demonstrate when the element has a settable property called x and a corresponding event named xChange. Here's a SizerComponent that fits the pattern. It has a size value property and a companion sizeChange event:

import { Component, EventEmitter, Input, Output } from '@angular/core';

@Component({
  selector: 'app-sizer',
  template: `
  <div>
    <button (click)="dec()" title="smaller">-</button>
    <button (click)="inc()" title="bigger">+</button>
    <label [style.font-size.px]="size">FontSize: {{size}}px</label>
  </div>`
})
export class SizerComponent {
  @Input()  size: number | string;
  @Output() sizeChange = new EventEmitter<number>();

  dec() { this.resize(-1); }
  inc() { this.resize(+1); }

  resize(delta: number) {
    this.size = Math.min(40, Math.max(8, +this.size + delta));
    this.sizeChange.emit(this.size);
  }
}

The initial size is an input value from a property binding. Clicking the buttons increases or decreases the size, within min/max values constraints, and then raises (emits) the sizeChange event with the adjusted size.

Here's an example in which the AppComponent.fontSizePx is two-way bound to the SizerComponent:

<app-sizer [(size)]="fontSizePx"></app-sizer>
<div [style.font-size.px]="fontSizePx">Resizable Text</div>

The AppComponent.fontSizePx establishes the initial SizerComponent.size value. Clicking the buttons updates the AppComponent.fontSizePx via the two-way binding. The revised AppComponent.fontSizePx value flows through to the style binding, making the displayed text bigger or smaller.

The two-way binding syntax is really just syntactic sugar for a property binding and an event binding. Angular desugars the SizerComponent binding into this:

<app-sizer [size]="fontSizePx" (sizeChange)="fontSizePx=$event"></app-sizer>

The $event variable contains the payload of the SizerComponent.sizeChange event. Angular assigns the $event value to the AppComponent.fontSizePx when the user clicks the buttons.

Clearly the two-way binding syntax is a great convenience compared to separate property and event bindings.

It would be convenient to use two-way binding with HTML form elements like <input> and <select>. However, no native HTML element follows the x value and xChange event pattern.

Fortunately, the Angular NgModel directive is a bridge that enables two-way binding to form elements.

Built-in directives

Earlier versions of Angular included over seventy built-in directives. The community contributed many more, and countless private directives have been created for internal applications.

You don't need many of those directives in Angular. You can often achieve the same results with the more capable and expressive Angular binding system. Why create a directive to handle a click when you can write a simple binding such as this?

<button (click)="onSave()">Save</button>

You still benefit from directives that simplify complex tasks. Angular still ships with built-in directives; just not as many. You'll write your own directives, just not as many.

This segment reviews some of the most frequently used built-in directives, classified as either attribute directives or structural directives.

Built-in `attribute` directives

Attribute directives listen to and modify the behavior of other HTML elements, attributes, properties, and components. They are usually applied to elements as if they were HTML attributes, hence the name.

Many details are covered in the Attribute Directives guide. Many NgModules such as the RouterModule and the FormsModule define their own attribute directives. This section is an introduction to the most commonly used attribute directives:

NgClass - add and remove a set of CSS classes
NgStyle - add and remove a set of HTML styles
NgModel - two-way data binding to an HTML form element

NgClass

You typically control how elements appear by adding and removing CSS classes dynamically. You can bind to the ngClass to add or remove several classes simultaneously.

A class binding is a good way to add or remove a single class.

<!-- toggle the "special" class on/off with a property -->
<div [class.special]="isSpecial">The class binding is special</div>

To add or remove many CSS classes at the same time, the NgClass directive may be the better choice.

Try binding ngClass to a key:value control object. Each key of the object is a CSS class name; its value is true if the class should be added, false if it should be removed.

Consider a setCurrentClasses component method that sets a component property, currentClasses with an object that adds or removes three classes based on the true/false state of three other component properties:

currentClasses: {};
setCurrentClasses() {
  // CSS classes: added/removed per current state of component properties
  this.currentClasses =  {
    'saveable': this.canSave,
    'modified': !this.isUnchanged,
    'special':  this.isSpecial
  };
}

Adding an ngClass property binding to currentClasses sets the element's classes accordingly:

<div [ngClass]="currentClasses">This div is initially saveable, unchanged, and special</div>

It's up to you to call setCurrentClasses(), both initially and when the dependent properties change.

NgStyle

You can set inline styles dynamically, based on the state of the component. With NgStyle you can set many inline styles simultaneously.

A style binding is an easy way to set a single style value.

<div [style.font-size]="isSpecial ? 'x-large' : 'smaller'" >
  This div is x-large or smaller.
</div>

To set many inline styles at the same time, the NgStyle directive may be the better choice.

Try binding ngStyle to a key:value control object. Each key of the object is a style name; its value is whatever is appropriate for that style.

Consider a setCurrentStyles component method that sets a component property, currentStyles with an object that defines three styles, based on the state of three other component properties:

currentStyles: {};
setCurrentStyles() {
  // CSS styles: set per current state of component properties
  this.currentStyles = {
    'font-style':  this.canSave      ? 'italic' : 'normal',
    'font-weight': !this.isUnchanged ? 'bold'   : 'normal',
    'font-size':   this.isSpecial    ? '24px'   : '12px'
  };
}

Adding an ngStyle property binding to currentStyles sets the element's styles accordingly:

<div [ngStyle]="currentStyles">
  This div is initially italic, normal weight, and extra large (24px).
</div>

It's up to you to call setCurrentStyles(), both initially and when the dependent properties change.

NgModel - Two-way binding to form elements with [(ngModel)]

When developing data entry forms, you often both display a data property and update that property when the user makes changes.

Two-way data binding with the NgModel directive makes that easy. Here's an example:

<input [(ngModel)]="currentHero.name">

FormsModule is required to use ngModel

Before using the ngModel directive in a two-way data binding, you must import the FormsModule and add it to the NgModule's imports list. Learn more about the FormsModule and ngModel in the Forms guide.

Here's how to import the FormsModule to make [(ngModel)] available.

import { NgModule } from '@angular/core';
import { BrowserModule }  from '@angular/platform-browser';
import { FormsModule } from '@angular/forms'; // <--- JavaScript import from Angular

/* Other imports */

@NgModule({
  imports: [
    BrowserModule,
    FormsModule  // <--- import into the NgModule
  ],
  /* Other module metadata */
})
export class AppModule { }

Inside [(ngModel)]

Looking back at the name binding, note that you could have achieved the same result with separate bindings to the <input> element's value property and input event.

<input [value]="currentHero.name"
       (input)="currentHero.name=$event.target.value" >

That's cumbersome. Who can remember which element property to set and which element event emits user changes? How do you extract the currently displayed text from the input box so you can update the data property? Who wants to look that up each time?

That ngModel directive hides these onerous details behind its own ngModel input and ngModelChange output properties.

<input
  [ngModel]="currentHero.name"
  (ngModelChange)="currentHero.name=$event">

The ngModel data property sets the element's value property and the ngModelChange event property listens for changes to the element's value.

The details are specific to each kind of element and therefore the NgModel directive only works for an element supported by a ControlValueAccessor that adapts an element to this protocol. The <input> box is one of those elements. Angular provides value accessors for all of the basic HTML form elements and the Forms guide shows how to bind to them.

You can't apply [(ngModel)] to a non-form native element or a third-party custom component until you write a suitable value accessor, a technique that is beyond the scope of this guide.

You don't need a value accessor for an Angular component that you write because you can name the value and event properties to suit Angular's basic two-way binding syntax and skip NgModel altogether. The sizer shown above is an example of this technique.

Separate ngModel bindings is an improvement over binding to the element's native properties. You can do better.

You shouldn't have to mention the data property twice. Angular should be able to capture the component's data property and set it with a single declaration, which it can with the [(ngModel)] syntax:

<input [(ngModel)]="currentHero.name">

Is [(ngModel)] all you need? Is there ever a reason to fall back to its expanded form?

The [(ngModel)] syntax can only set a data-bound property. If you need to do something more or something different, you can write the expanded form.

The following contrived example forces the input value to uppercase:

<input
  [ngModel]="currentHero.name"
  (ngModelChange)="setUppercaseName($event)">

Here are all variations in action, including the uppercase version:

Built-in `structural` directives

Structural directives are responsible for HTML layout. They shape or reshape the DOM's structure, typically by adding, removing, and manipulating the host elements to which they are attached.

The deep details of structural directives are covered in the Structural Directives guide where you'll learn:

why you prefix the directive name with an asterisk (*).
to use <ng-container> to group elements when there is no suitable host element for the directive.
how to write your own structural directive.
that you can only apply one structural directive to an element.

This section is an introduction to the common structural directives:

NgIf - conditionally add or remove an element from the DOM
NgSwitch - a set of directives that switch among alternative views
NgForOf - repeat a template for each item in a list

NgIf

You can add or remove an element from the DOM by applying an NgIf directive to that element (called the host element). Bind the directive to a condition expression like isActive in this example.

<app-hero-detail *ngIf="isActive"></app-hero-detail>

Don't forget the asterisk (*) in front of ngIf.

When the isActive expression returns a truthy value, NgIf adds the HeroDetailComponent to the DOM. When the expression is falsy, NgIf removes the HeroDetailComponent from the DOM, destroying that component and all of its sub-components.

Show/hide is not the same thing

You can control the visibility of an element with a class or style binding:

<!-- isSpecial is true -->
<div [class.hidden]="!isSpecial">Show with class</div>
<div [class.hidden]="isSpecial">Hide with class</div>

<!-- HeroDetail is in the DOM but hidden -->
<app-hero-detail [class.hidden]="isSpecial"></app-hero-detail>

<div [style.display]="isSpecial ? 'block' : 'none'">Show with style</div>
<div [style.display]="isSpecial ? 'none'  : 'block'">Hide with style</div>

Hiding an element is quite different from removing an element with NgIf.

When you hide an element, that element and all of its descendents remain in the DOM. All components for those elements stay in memory and Angular may continue to check for changes. You could be holding onto considerable computing resources and degrading performance, for something the user can't see.

When NgIf is false, Angular removes the element and its descendents from the DOM. It destroys their components, potentially freeing up substantial resources, resulting in a more responsive user experience.

The show/hide technique is fine for a few elements with few children. You should be wary when hiding large component trees; NgIf may be the safer choice.

Guard against null

The ngIf directive is often used to guard against null. Show/hide is useless as a guard. Angular will throw an error if a nested expression tries to access a property of null.

Here we see NgIf guarding two <div>s. The currentHero name will appear only when there is a currentHero. The nullHero will never be displayed.

<div *ngIf="currentHero">Hello, {{currentHero.name}}</div>
<div *ngIf="nullHero">Hello, {{nullHero.name}}</div>

See also the safe navigation operator described below.

NgForOf

NgForOf is a repeater directive — a way to present a list of items. You define a block of HTML that defines how a single item should be displayed. You tell Angular to use that block as a template for rendering each item in the list.

Here is an example of NgForOf applied to a simple <div>:

<div *ngFor="let hero of heroes">{{hero.name}}</div>

You can also apply an NgForOf to a component element, as in this example:

<app-hero-detail *ngFor="let hero of heroes" [hero]="hero"></app-hero-detail>

Don't forget the asterisk (*) in front of ngFor.

The text assigned to *ngFor is the instruction that guides the repeater process.

*ngFor microsyntax

The string assigned to *ngFor is not a template expression. It's a microsyntax — a little language of its own that Angular interprets. The string "let hero of heroes" means:

Take each hero in the heroes array, store it in the local hero looping variable, and make it available to the templated HTML for each iteration.

Angular translates this instruction into a <ng-template> around the host element, then uses this template repeatedly to create a new set of elements and bindings for each hero in the list.

Learn about the microsyntax in the Structural Directives guide.

Template input variables

The let keyword before hero creates a template input variable called hero. The NgForOf directive iterates over the heroes array returned by the parent component's heroes property and sets hero to the current item from the array during each iteration.

You reference the hero input variable within the NgForOf host element (and within its descendants) to access the hero's properties. Here it is referenced first in an interpolation and then passed in a binding to the hero property of the <hero-detail> component.

<div *ngFor="let hero of heroes">{{hero.name}}</div>
<app-hero-detail *ngFor="let hero of heroes" [hero]="hero"></app-hero-detail>

Learn more about template input variables in the Structural Directives guide.

ngFor with index*

The index property of the NgForOf directive context returns the zero-based index of the item in each iteration. You can capture the index in a template input variable and use it in the template.

The next example captures the index in a variable named i and displays it with the hero name like this.

<div *ngFor="let hero of heroes; let i=index">{{i + 1}} - {{hero.name}}</div>

NgFor is implemented by the NgForOf directive. Read more about the other NgForOf context values such as last, even, and odd in the NgForOf API reference.

ngFor with trackBy*

The NgForOf directive may perform poorly, especially with large lists. A small change to one item, an item removed, or an item added can trigger a cascade of DOM manipulations.

For example, re-querying the server could reset the list with all new hero objects.

Most, if not all, are previously displayed heroes. You know this because the id of each hero hasn't changed. But Angular sees only a fresh list of new object references. It has no choice but to tear down the old DOM elements and insert all new DOM elements.

Angular can avoid this churn with trackBy. Add a method to the component that returns the value NgForOf should track. In this case, that value is the hero's id.

trackByHeroes(index: number, hero: Hero): number { return hero.id; }

In the microsyntax expression, set trackBy to this method.

<div *ngFor="let hero of heroes; trackBy: trackByHeroes">
  ({{hero.id}}) {{hero.name}}
</div>

Here is an illustration of the trackBy effect. "Reset heroes" creates new heroes with the same hero.ids. "Change ids" creates new heroes with new hero.ids.

With no trackBy, both buttons trigger complete DOM element replacement.
With trackBy, only changing the id triggers element replacement.

The `NgSwitch` directives

NgSwitch is like the JavaScript switch statement. It can display one element from among several possible elements, based on a switch condition. Angular puts only the selected element into the DOM.

NgSwitch is actually a set of three, cooperating directives: NgSwitch, NgSwitchCase, and NgSwitchDefault as seen in this example.

<div [ngSwitch]="currentHero.emotion">
  <app-happy-hero    *ngSwitchCase="'happy'"    [hero]="currentHero"></app-happy-hero>
  <app-sad-hero      *ngSwitchCase="'sad'"      [hero]="currentHero"></app-sad-hero>
  <app-confused-hero *ngSwitchCase="'confused'" [hero]="currentHero"></app-confused-hero>
  <app-unknown-hero  *ngSwitchDefault           [hero]="currentHero"></app-unknown-hero>
</div>

NgSwitch is the controller directive. Bind it to an expression that returns the switch value. The emotion value in this example is a string, but the switch value can be of any type.

Bind to [ngSwitch]. You'll get an error if you try to set *ngSwitch because NgSwitch is an attribute directive, not a structural directive. It changes the behavior of its companion directives. It doesn't touch the DOM directly.

Bind to *ngSwitchCase and *ngSwitchDefault. The NgSwitchCase and NgSwitchDefault directives are structural directives because they add or remove elements from the DOM.

NgSwitchCase adds its element to the DOM when its bound value equals the switch value.
NgSwitchDefault adds its element to the DOM when there is no selected NgSwitchCase.

The switch directives are particularly useful for adding and removing component elements. This example switches among four "emotional hero" components defined in the hero-switch.components.ts file. Each component has a hero input property which is bound to the currentHero of the parent component.

Switch directives work as well with native elements and web components too. For example, you could replace the <confused-hero> switch case with the following.

<div *ngSwitchCase="'confused'">Are you as confused as {{currentHero.name}}?</div>

Template reference variables ( #var )

A template reference variable is often a reference to a DOM element within a template. It can also be a reference to an Angular component or directive or a web component.

Use the hash symbol (#) to declare a reference variable. The #phone declares a phone variable on an <input> element.

<input #phone placeholder="phone number">

You can refer to a template reference variable anywhere in the template. The phone variable declared on this <input> is consumed in a <button> on the other side of the template

<input #phone placeholder="phone number">

<!-- lots of other elements -->

<!-- phone refers to the input element; pass its `value` to an event handler -->
<button (click)="callPhone(phone.value)">Call</button>

How a reference variable gets its value

In most cases, Angular sets the reference variable's value to the element on which it was declared. In the previous example, phone refers to the phone number <input> box. The phone button click handler passes the input value to the component's callPhone method. But a directive can change that behavior and set the value to something else, such as itself. The NgForm directive does that.

The following is a simplified version of the form example in the Forms guide.

<form (ngSubmit)="onSubmit(heroForm)" #heroForm="ngForm">
  <div class="form-group">
    <label for="name">Name
      <input class="form-control" name="name" required [(ngModel)]="hero.name">
    </label>
  </div>
  <button type="submit" [disabled]="!heroForm.form.valid">Submit</button>
</form>
<div [hidden]="!heroForm.form.valid">
  {{submitMessage}}
</div>

A template reference variable, heroForm, appears three times in this example, separated by a large amount of HTML. What is the value of heroForm?

If Angular hadn't taken it over when you imported the FormsModule, it would be the HTMLFormElement. The heroForm is actually a reference to an Angular NgForm directive with the ability to track the value and validity of every control in the form.

The native <form> element doesn't have a form property. But the NgForm directive does, which explains how you can disable the submit button if the heroForm.form.valid is invalid and pass the entire form control tree to the parent component's onSubmit method.

Template reference variable warning notes

A template reference variable (#phone) is not the same as a template input variable (let phone) such as you might see in an *ngFor. Learn the difference in the Structural Directives guide.

The scope of a reference variable is the entire template. Do not define the same variable name more than once in the same template. The runtime value will be unpredictable.

You can use the ref- prefix alternative to #. This example declares the fax variable as ref-fax instead of #fax.

<input ref-fax placeholder="fax number">
<button (click)="callFax(fax.value)">Fax</button>

Input and Output properties

An Input property is a settable property annotated with an @Input decorator. Values flow into the property when it is data bound with a property binding

An Output property is an observable property annotated with an @Output decorator. The property almost always returns an Angular EventEmitter. Values flow out of the component as events bound with an event binding.

You can only bind to another component or directive through its Input and Output properties.

Remember that all components are directives.

The following discussion refers to components for brevity and because this topic is mostly a concern for component authors.

Discussion

You are usually binding a template to its own component class. In such binding expressions, the component's property or method is to the right of the (=).

<img [src]="iconUrl"/>
<button (click)="onSave()">Save</button>

The iconUrl and onSave are members of the AppComponent class. They are not decorated with @Input() or @Output. Angular does not object.

You can always bind to a public property of a component in its own template. It doesn't have to be an Input or Output property

A component's class and template are closely coupled. They are both parts of the same thing. Together they are the component. Exchanges between a component class and its template are internal implementation details.

Binding to a different component

You can also bind to a property of a different component. In such bindings, the other component's property is to the left of the (=).

In the following example, the AppComponent template binds AppComponent class members to properties of the HeroDetailComponent whose selector is 'app-hero-detail'.

<app-hero-detail [hero]="currentHero" (deleteRequest)="deleteHero($event)">
</app-hero-detail>

The Angular compiler may reject these bindings with errors like this one:

Uncaught Error: Template parse errors:
Can't bind to 'hero' since it isn't a known property of 'app-hero-detail'

You know that HeroDetailComponent has hero and deleteRequest properties. But the Angular compiler refuses to recognize them.

The Angular compiler won't bind to properties of a different component unless they are Input or Output properties.

There's a good reason for this rule.

It's OK for a component to bind to its own properties. The component author is in complete control of those bindings.

But other components shouldn't have that kind of unrestricted access. You'd have a hard time supporting your component if anyone could bind to any of its properties. Outside components should only be able to bind to the component's public binding API.

Angular asks you to be explicit about that API. It's up to you to decide which properties are available for binding by external components.

TypeScript public doesn't matter

You can't use the TypeScript public and private access modifiers to shape the component's public binding API.

All data bound properties must be TypeScript public properties. Angular never binds to a TypeScript private property.

Angular requires some other way to identify properties that outside components are allowed to bind to. That other way is the @Input() and @Output() decorators.

Declaring Input and Output properties

In the sample for this guide, the bindings to HeroDetailComponent do not fail because the data bound properties are annotated with @Input() and @Output() decorators.

@Input()  hero: Hero;
@Output() deleteRequest = new EventEmitter<Hero>();

Alternatively, you can identify members in the inputs and outputs arrays of the directive metadata, as in this example:
@Component({
  inputs: ['hero'],
  outputs: ['deleteRequest'],
})

Input or output?

Input properties usually receive data values. Output properties expose event producers, such as EventEmitter objects.

The terms input and output reflect the perspective of the target directive.

HeroDetailComponent.hero is an input property from the perspective of HeroDetailComponent because data flows into that property from a template binding expression.

HeroDetailComponent.deleteRequest is an output property from the perspective of HeroDetailComponent because events stream out of that property and toward the handler in a template binding statement.

Aliasing input/output properties

Sometimes the public name of an input/output property should be different from the internal name.

This is frequently the case with attribute directives. Directive consumers expect to bind to the name of the directive. For example, when you apply a directive with a myClick selector to a <div> tag, you expect to bind to an event property that is also called myClick.

<div (myClick)="clickMessage=$event" clickable>click with myClick</div>

However, the directive name is often a poor choice for the name of a property within the directive class. The directive name rarely describes what the property does. The myClick directive name is not a good name for a property that emits click messages.

Fortunately, you can have a public name for the property that meets conventional expectations, while using a different name internally. In the example immediately above, you are actually binding through the myClick alias to the directive's own clicks property.

You can specify the alias for the property name by passing it into the input/output decorator like this:

@Output('myClick') clicks = new EventEmitter<string>(); //  @Output(alias) propertyName = ...

You can also alias property names in the inputs and outputs arrays. You write a colon-delimited (:) string with the directive property name on the left and the public alias on the right:
@Directive({
  outputs: ['clicks:myClick']  // propertyName:alias
})

Template expression operators

The template expression language employs a subset of JavaScript syntax supplemented with a few special operators for specific scenarios. The next sections cover two of these operators: pipe and safe navigation operator.

The pipe operator ( | )

The result of an expression might require some transformation before you're ready to use it in a binding. For example, you might display a number as a currency, force text to uppercase, or filter a list and sort it.

Angular pipes are a good choice for small transformations such as these. Pipes are simple functions that accept an input value and return a transformed value. They're easy to apply within template expressions, using the pipe operator (|):

<div>Title through uppercase pipe: {{title | uppercase}}</div>

The pipe operator passes the result of an expression on the left to a pipe function on the right.

You can chain expressions through multiple pipes:

<!-- Pipe chaining: convert title to uppercase, then to lowercase -->
<div>
  Title through a pipe chain:
  {{title | uppercase | lowercase}}
</div>

And you can also apply parameters to a pipe:

<!-- pipe with configuration argument => "February 25, 1970" -->
<div>Birthdate: {{currentHero?.birthdate | date:'longDate'}}</div>

The json pipe is particularly helpful for debugging bindings:

<div>{{currentHero | json}}</div>

The generated output would look something like this

{ "id": 0, "name": "Hercules", "emotion": "happy",
    "birthdate": "1970-02-25T08:00:00.000Z",
    "url": "http://www.imdb.com/title/tt0065832/",
    "rate": 325 }

The Angular safe navigation operator (?.) is a fluent and convenient way to guard against null and undefined values in property paths. Here it is, protecting against a view render failure if the currentHero is null.

The current hero's name is {{currentHero?.name}}

What happens when the following data bound title property is null?

The title is {{title}}

The view still renders but the displayed value is blank; you see only "The title is" with nothing after it. That is reasonable behavior. At least the app doesn't crash.

Suppose the template expression involves a property path, as in this next example that displays the name of a null hero.

The null hero's name is {{nullHero.name}}

JavaScript throws a null reference error, and so does Angular:

TypeError: Cannot read property 'name' of null in [null].

Worse, the entire view disappears.

This would be reasonable behavior if the hero property could never be null. If it must never be null and yet it is null, that's a programming error that should be caught and fixed. Throwing an exception is the right thing to do.

On the other hand, null values in the property path may be OK from time to time, especially when the data are null now and will arrive eventually.

While waiting for data, the view should render without complaint, and the null property path should display as blank just as the title property does.

Unfortunately, the app crashes when the currentHero is null.

You could code around that problem with *ngIf.

<!--No hero, div not displayed, no error -->
<div *ngIf="nullHero">The null hero's name is {{nullHero.name}}</div>

You could try to chain parts of the property path with &&, knowing that the expression bails out when it encounters the first null.

The null hero's name is {{nullHero && nullHero.name}}

These approaches have merit but can be cumbersome, especially if the property path is long. Imagine guarding against a null somewhere in a long property path such as a.b.c.d.

The Angular safe navigation operator (?.) is a more fluent and convenient way to guard against nulls in property paths. The expression bails out when it hits the first null value. The display is blank, but the app keeps rolling without errors.

<!-- No hero, no problem! -->
The null hero's name is {{nullHero?.name}}

It works perfectly with long property paths such as a?.b?.c?.d.

The non-null assertion operator ( ! )

As of Typescript 2.0, you can enforce strict null checking with the --strictNullChecks flag. TypeScript then ensures that no variable is unintentionally null or undefined.

In this mode, typed variables disallow null and undefined by default. The type checker throws an error if you leave a variable unassigned or try to assign null or undefined to a variable whose type disallows null and undefined.

The type checker also throws an error if it can't determine whether a variable will be null or undefined at runtime. You may know that can't happen but the type checker doesn't know. You tell the type checker that it can't happen by applying the post-fix non-null assertion operator (!).

The Angular non-null assertion operator (!) serves the same purpose in an Angular template.

For example, after you use *ngIf to check that hero is defined, you can assert that hero properties are also defined.

<!--No hero, no text -->
<div *ngIf="hero">
  The hero's name is {{hero!.name}}
</div>

When the Angular compiler turns your template into TypeScript code, it prevents TypeScript from reporting that hero.name might be null or undefined.

Unlike the safe navigation operator, the non-null assertion operator does not guard against null or undefined. Rather it tells the TypeScript type checker to suspend strict null checks for a specific property expression.

You'll need this template operator when you turn on strict null checks. It's optional otherwise.

Built-in template functions

The `$any()` type cast function

Sometimes a binding expression triggers a type error during AOT compilation and it is not possible or difficult to fully specify the type. To silence the error, you can use the $any() cast function to cast the expression to the any type as in the following example:

<p>The item's undeclared best by date is: {{$any(item).bestByDate}}</p>

When the Angular compiler turns this template into TypeScript code, it prevents TypeScript from reporting that bestByDate is not a member of the item object when it runs type checking on the template.

The $any() cast function also works with this to allow access to undeclared members of the component.

<p>The item's undeclared best by date is: {{$any(this).bestByDate}}</p>

The $any() cast function works anywhere in a binding expression where a method call is valid.

© 2010–2019 Google, Inc.
Licensed under the Creative Commons Attribution License 4.0.
https://v7.angular.io/guide/template-syntax