Name: Using Type Inference in TypeScript
Uploaded: 2016-04-27T20:34:51-05:00
Description: This lesson explains how the Typescript compiler uses inference to guess what types we’re using, so we don’t have to annotate all the things.

Introduction to Static Typing

Using Type Inference in TypeScript

Union Types and Type Aliases in TypeScript

Distinguishing between types of Strings in TypeScript

Using Interfaces to Describe Types in TypeScript

Creating a Class in TypeScript 

Sharing Class Behavior with Inheritance in TypeScript 

Using Assertion to Convert Types in TypeScript

The Basics of Generics in TypeScript

Practical Generics in TypeScript

Use Types Effectively in TypeScript

This lesson explains how the Typescript compiler uses inference to guess what types we’re using, so we don’t have to annotate all the things.

https://app.egghead.io/api/v1/lessons/typescript-using-type-inference-in-typescript/subtitles

[00:00] TypeScript gives us static types. When we run the compiler, there's no errors because the compiler knows that username is a string. Typescript also gives us inference, which just means when there is no annotation the compiler will guess what type we want.

[00:14] If we set a new value for username, even though username has no annotation, both the IDE and the compiler are mad because the compiler inferred, or guessed, that username is a string because it was initialized with a string. Array of strings is not assignable to typestring.

[00:32] We have to be careful. If we don't annotate username, initialize it as undefined, and set a value later, like we've done here, when we run the compiler, we get no errors because the compiler can't infer what the type is.

[00:46] The compiler can also infer function return types. I'm just making a function that accepts a string and a number. Here's the return type. When we run the compiler, we get no errors. If we remove the annotation and run the compiler, again, no errors.

[01:00] If we annotate the return type with the wrong type, the compiler gets mad because we told it the return type is a number. It knows that a string plus a number is a string. That's not assignable to the number return type that we specified.

[01:15] When we remove the return type and run the compiler no errors. This kind of inference happens from the bottom up. The compiler knows what the return type is, bottom, based on its arguments, top. There's also contextual inference, which is top down.

[01:31] Let's say we have an HTML element and we want to use an onclick event and alert on that element. The compiler knows the type signature for the HTML element and its onclick method. When we run the compiler, there's no errors. If we remove the annotations and run the compiler, there's no errors.

[01:50] If we change the event handler argument type, the ID will show us that ironically, there is no button parameter on an HTML button element. The compiler will tell us that we can't define the event handler's type as an HTML button, because it knows it should be a mouse event.

[02:05] We've typed the argument passed through the function as an HTML button element. It's supposed to be a mouse event. We say that the HTML button element is incompatible with the event that's supposed to be passed, which is the mouse event. A button property does not exist on an HTML button element.

[02:23] To review, contextual inference is top down. Top down. The compiler knows what the target is and its onclick method. It's guessing what the event handler argument type is based on the target onclick signature.

[02:40] Vanilla inference is bottom up. The compiler is inferring what the function return type is based on its arguments. If you initialize your variables undefined, the compiler won't be able to infer what the type is.

Using Type Inference in TypeScript

Share this video with your friends

Course

Use Types Effectively in TypeScript