Scope
All About Scope
The scope of a program in JavaScript is the set of variables that are available for use within the program. If a variable or other expression is not in the current scope, then it is unavailable for use. If we declare a variable, this variable will only be valid in the scope where we declared it. We can have nested scopes, but we'll see that in a little bit.
When we declare a variable in a certain scope, it will evaluate to a specific value in that scope. We have been using the concept of scope in our code all along! Now we are just giving this concept a name.
By the end of this reading you should be able to predict the evaluation of code that utilizes local scope, block scope, lexical scope, and scope chaining
Advantages of utilizing scope
Before we start talking about different types of scope we'll be talking about the two main advantages that scope gives us:
- Security - Scope adds security to our code by ensuring that variables can
only be accessed by pre-defined parts of our programs.
- Reduced Variable Name Collisions - Scope reduces variable name
collisions, also known as namespace collisions, by ensuring you can use the
same variable name multiple times in different scopes without accidentally
overwriting those variable's values.
Different kinds of scope
There are three types of scope in JavaScript: global scope
, local scope
, and block scope
.
Global scope
Let's start by talking about the widest scope there is: global scope. The global scope is represented by the window
object in the browser and the global
object in Node.js. Adding attributes to these objects makes them available throughout the entire program. We can show this with a quick example:
let myName = "Apples";
console.log(myName);
// this myName references the myName variable from this scope,
// so myName will evaluate to "Apples"
The variable myName
above is not inside a function, it is just lying out in the open in our code. The myName
variable is part of global scope. The Global scope is the largest scope that exists, it is the outermost scope that exists.
While useful on occasion, global variables are best avoided. Every time a variable is declared on the global scope, the chance of a name collision increases. If we are unaware of the global variables in our code, we may accidentally overwrite variables.
Local scope
The scope of a function is the set of variables that are available for use within that function. We call the scope within a function: local scope. The local scope of a function includes:
- the function's arguments
- any local variables declared inside the function
- any variables that were already declared when the function was defined
In JavaScript when we enter a new function we enter a new scope:
// global scope
let myName = "global";
function function1() {
// function1's scope
let myName = "func1";
console.log("function1 myName: " + myName);
}
function function2() {
// function2's scope
let myName = "func2";
console.log("function2 myName: " + myName);
}
function1(); // function1 myName: func1
function2(); // function2 myName: func2
console.log("global myName: " + myName); // global myName: global
In the code above we are dealing with three different scopes: the global scope, function1
, and function2
. Since each of the myName
variables were declared in separate scopes, we are allowed to reuse variable names without any issues. This is because each of the myName
variables is bound to their respective functions.
Block scope
A block in JavaScript is denoted by a pair of curly braces ({}
). Examples of block statements in JavaScript are if
conditionals or for
and while
loops.
When using the keywords let
or const
the variables defined within the curly braces will be block scoped. Let's look at an example:
// global scope
let dog = "woof";
// block scope
if (true) {
let dog = "bowwow";
console.log(dog); // will print "bowwow"
}
console.log(dog); // will print "woof"
Scope chaining: variables and scope
A key scoping rule in JavaScript is the fact that an inner scope does have access to variables in the outer scope.
Let's look at a simple example:
let name = "Fiona";
// we aren't passing in or defining and variables
function hungryHippo() {
console.log(name + " is hungry!");
}
hungryHippo(); // => "Fiona is hungry"
So when the hungryHippo
function is declared a new local scope will be created for that function. Continuing on that line of thought what happens when we refer to name
inside of hungryHippo
? If the name
variable is not found in the immediate scope, JavaScript will search all of the accessible outer scopes until it finds a variable name that matches the one we are referencing. Once it finds the first matching variable, it will stop searching. In JavaScript this is called scope chaining.
Now let's look at an example of scope chaining with nested scope. Just like functions in JavaScript, a scope can be nested within another scope. Take a look at the example below:
// global scope
let person = "Rae";
// sayHello function's local scope
function sayHello() {
let person = "Jeff";
// greet function's local scope
function greet() {
console.log("Hi, " + person + "!");
}
greet();
}
sayHello(); // logs 'Hi, Jeff!'
In the example above, the variable person
is referenced by greet
, even though it was never declared within greet
! When this code is executed JavaScript will attempt to run the greet
function - notice there is no person
variable within the scope of the greet
function and move on to seeing if that variable is defined in an outer scope.
Notice that the greet
function prints out Hi, Jeff!
instead of Hi, Rae!
. This is because JavaScript will start at the inner most scope looking for a variable named person
. Then JavaScript will work it's way outward looking for a variable with a matching name of person
. Since the person
variable within sayHello
is in the next level of scope above greet
JavaScript then stops it's scope chaining search and assigns the value of the person
variable.
Functions such as greet
that use (ie. capture) variables like the person variable are called closures. We'll be talking a lot more about closures very soon!
Important An inner scope can reference outer variables, but an outer scope cannot reference inner variables:
function potatoMaker() {
let name = "potato";
console.log(name);
}
potatoMaker(); // => "potato"
console.log(name); // => ReferenceError: name is not defined
Lexical scope
There is one last important concept to talk about when we refer to scope - and that is lexical scope. Whenever you run a piece of JavaScript that code is first parsed before it is actually run. This is known as the lexing time. In the lexing time your parser resolves variable names to their values when functions are nested.
The main take away is that lexical scope is determined at lexing time so we can determine the values of variables without having to run any code. JavaScript is a language without dynamic scoping. This means that by looking at a piece of code we can determine the values of variables just by looking at the different scopes involved.
Let's look at a quick example:
function outer() {
let x = 5;
function inner() {
// here we know the value of x because scope chaining will
// go into the scope above this one looking for variable named x.
// We do not need to run this code in order to determine the value of x!
console.log(x);
}
inner();
}
In the inner
function above we don't need to run the outer
function to know what the value of x
will be because of lexical scoping.
The scope of a program in JavaScript is the set of variables that are available for use within the program. Due to lexical scoping we can determine the value of a variable by looking at various scopes without having to run our code. Scope Chaining allows code within an inner scope to access variables declared in an outer scope.
There are three different scopes:
- global scope - the global space is JavaScript
- local scope - created when a function is defined
- block scope - created by entering a pair of curly braces
Different Kinds of Variables
Variables are used to store information to be referenced and manipulated in a computer program. A variable's sole purpose is to label and store data in computer memory. Up to this point we've been using the let
keyword as our only way of declaring a JavaScript variable. It's now time to expand your tool set to learn about the different kinds of JavaScript variables you can use!
When you finish this reading, you should be able to:
- Identify the three keywords used to declare a variable in JavaScript
- Explain the differences between
const
,let
andvar
- Identify the difference between function and block-scoped variables
- Paraphrase the concept of hoisting in regards to function and block-scoped
variables
Declaring variables
All the code you write in JavaScript is evaluated. A variable always evaluates to the value it contains no matter how you declare it.
The different ways to declare variables
In the beginning there was var
. The var
keyword used to be the only way to declare a JavaScript variable. However, in ECMAScript 2015 JavaScript introduced two new ways of declaring JavaScript variables: let
and const
. Meaning, in JavaScript there are three different ways to declare a variable. Each of these keywords has advantages and disadvantages and we will now talk about each keyword at length.
-
let
: any variables declared with the keywordlet
allows you to reassign
that variable. Variable declared using let
is scoped within a block.
-
const
: any variables declared with the keywordconst
_will not allow you
to reassign_ that variable. Variable declared using const
is scoped within
a block.
-
var
: Avar
declared variable may or may not be reassigned, and the
variable is scoped to a function.
For this course and for your programming career moving forward we recommend you always use let
& const
. These two words allow us to be the most clear with our intentions for the variable we are creating.
Hoisting and scoping with variables
A wonderful definition of hoisting by Mabishi Wakio, "Hoisting is a JavaScript mechanism where variables and function declarations are moved to the top of their scope before code execution."
What this means is that when you run JavaScript code the variables and function declarations will be hoisted to the top of their particular scope. This is important because const
and let
are block-scoped while var
is function-scoped.
Let's start by talking more about all const
, let
, and var
before we dive into why the difference of scopes and hoisting is important.
Function-scoped variables
When JavaScript was young the only available variable was var
. The var
keyword creates function-scoped variables. That means when you use the var
keyword to declare a variable that variable will be confined to the scope of the current function.
Here is a simple example of declaring a var
variable within a function:
function test() {
var a = 10;
console.log(a); // => 10
}
One of the drawbacks of using var
is that it is a less indicative way of defining a variable.
Hoisting with function-scoped variables
Let's take a look at what hoisting does to a function-scoped variable:
function test() {
console.log(hoistedVar); // => undefined
var hoistedVar = 10;
}
test();
Huh - that's weird. You'd expect an error from referring to a variable like hoistedVar
before it's defined, something like: ReferenceError: hoistedVar is not defined
. However this is not the case because of hoisting in JavaScript!
So essentially hoisting will isolate and, in the computer's memory, will declare a variable as the top of it's scope. With a function-scoped variable, var
, the name of the variable will be hoisted to the top of the function. In the above snippet, since hoistedVar
is declared using the var
keyword the hoistedVar
's scope is the test
function. To be clear what is being hoisted is the declaration, not the assignment itself.
In JavaScript, all variables defined with the var
keyword have an initial value of undefined
. Here is a translation of how JavaScript would deal with hoisting in the above test
function:
function test() {
// JavaScript will declare the variable *in computer memory* at the top of it's scope
var hoistedVar;
// since hoisting declared the variable above we now get
// the value of 'undefined'
console.log(hoistedVar); // => undefined
var hoistedVar = 10;
}
Block-scoped variables
When you are declaring a variable with the keyword let
or const
you are declaring a variable that exists within block scope. Blocks in JavaScript are denoted by curly braces({}
). The following examples create a block scope: if
statements, while
loops, switch
statements, and for
loops.
Using the keyword let
We can use let
to declare re-assignable block-scoped variables. You are, of course, very familiar with let
so let's take a look at how let
works within a block scope:
function blockScope() {
let test = "upper scope";
if (true) {
let test = "lower scope";
console.log(test); // "lower scope"
}
console.log(test); // "upper scope"
}
In the example above we can see that the test
variable was declared twice using the keyword let
but since they were declared within different scopes they have different values.
JavaScript will raise a SyntaxError
if you try to declare the same let
variable twice in one block.
if (true) {
let test = "this works!";
let test = "nope!"; // Identifier 'test' has already been declared
}
Whereas if you try the same example with var
:
var test = "this works!";
var test = "nope!";
console.log(test); // prints "nope!"
We can see above that var
will allow you to redeclare a variable twice which can lead to some very confusing and frustrating debugging.
Feel free to peruse the documentation for the keyword let
for more examples.
Using the keyword const
We use const
to declare block-scoped variables that can not be reassigned. In JavaScript variables that cannot be reassigned are called constants. Constants should be used for values that will not be re-declared or re-assigned.
Properties of constants:
- They are block-scoped like
let
. - JavaScript enforces constants by raising an error if you try to reassign them.
- Trying to redeclare a constant with a
var
orlet
by the same name will
also raise an error.
Let's look at a quick example of what happens when trying to reassign a constant:
> const favFood = "cheeseboard pizza"; // Initializes a constant
undefined
> const favFood = "inferior food"; // Re-initialization raises an error
TypeError: Identifier 'favFood' has already been declared
> let favFood = "other inferior food"; // Re-initialization raises an error
TypeError: Identifier 'favFood' has already been declared
> favFood = "deep-dish pizza"; // Re-assignment raises an error
TypeError: Assignment to constant variable.
We cannot reassign a constant, but constants that are assigned to Reference types are mutable. The name binding of a constant is immutable. For example, if we set a constant equal to an Reference type like an object, we can still modify that object:
const animals = {};
animals.big = "beluga whale"; // This works!
animals.small = "capybara"; // This works!
animals = { big: "beluga whale" }; // Will error because of the reassignment
Constants cannot be reassigned but, just like with let
, new constants of the same names can be declared within nested scopes.
Take a look at the following for an example:
const favFood = "cheeseboard pizza";
console.log(favFood);
if (true) {
// This works! Declaration is scoped to the `if` block
const favFood = "noodles";
console.log(favFood); // Prints "noodles"
}
console.log(favFood); // Prints 'cheeseboard pizza'
Just like with let
when you use const
twice in the same block JavaScript will raise a SyntaxError
.
if (true) {
const test = "this works!";
const test = "nope!"; // SyntaxError: Identifier 'test' has already been declared
}
Hoisting with block-scoped variables
When JavaScript ES6 introduced new ways of declaring a variable using let
and const
the idea of block-level hoisting was also introduced. Block scope hoisting allows developers to avoid previous debugging debacles that naturally happened from using var
.
Let's take a look at what hoisting does to a block-scoped variable:
if (true) {
console.log(str); // => Uncaught ReferenceError: Cannot access 'str' before initialization
const str = "apple";
}
Looking at the above we can see that an explicit error is thrown if you attempt to use a block-scoped variable before it was declared. This is the typical behavior in a lot of programming languages - that a variable cannot be referred to until initialized to a value.
However, JavaScript is still performing hoisting with block-scoped declared variables. The difference lies is how it initializes them. Meaning that let
and const
variables are not initialized to the value of undefined
.
The time before a let
or const
variable is declared, but not used is called the Temporal Dead Zone. A very cool name for a simple idea. Variables declared using let
and const
are not initialized until their definitions are evaluated. Meaning, you will get an error if you try to reference a let
or const
declared variable before it is evaluated.
Let's look at one more example that should illuminate the presence of the Temporal Dead Zone:
var str = "not apple";
if (true) {
console.log(str); //Uncaught ReferenceError: Cannot access 'str' before initialization
let str = "apple";
}
In the above example we can see that inside the if
block the let
declared variable, str
, throws an error. Showing that the error thrown by a let
variable in the temporal dead zone takes precedence over any scope chaining that would attempt to go to the outer scope to find a value for the str
variable.
Function scope vs. block scope
Let's now take a deeper look at the comparison of using function vs. block scoped variables.
Let's start with a simple example:
function partyMachine() {
var string = "party";
console.log("this is a " + string);
}
Looks good so far but let's take that example a step farther and see some of the less fun parts of the var
keyword in terms of scope:
function partyMachine() {
var string = "party";
if (true) {
// since var is not block-scoped and not constant
// this assignment sticks!
var string = "bummer";
}
console.log("this is a " + string);
}
partyMachine(); // => "this is a bummer"
We can see in the above example how the flexibility of var
can ultimately be a bad thing. Since var
is function-scoped and can be reassigned and re-declared without error it is very easy to overwrite variable values by accident.
This is the problem that ES6 introduced let
and const
to solve. Since let
and const
are block-scoped it's a lot easier to avoid accidentally overwriting variable values.
Let's take a look at the example function above rewritten using let
and const
:
function partyMachine() {
const string = "party";
if (true) {
// this variable is restricted to the scope of this block
const string = "bummer";
}
console.log("this is a " + string);
}
partyMachine(); // => "this is a party"
Global variables
If you leave off a declaration when initializing a variable, it will become a global. Do not do this. We declare variables using the keywords var
, let
, and const
to ensure that our variables are declared within a proper scope. Any variables declared without these keywords will be declared on the global scope.
JavaScript has a single global scope, which means all of the files from your projects and any libraries you use will all be sharing the same scope. Every time a variable is declared on the global scope, the chance of a name collision increases. If we are unaware of the global variables in our code, we may accidentally overwrite variables.
Let's look at a quick example showing why this is a bad idea:
function good() {
let x = 5;
let y = "yay";
}
function bad() {
y = "Expect the unexpected (eg. globals)";
}
function why() {
console.log(y); // "Expect the unexpected (eg. globals)""
console.log(x); // Raises an error
}
why();
Limiting global variables will help you create code that is much more easily maintainable. Strive to write your functions so that they are self-contained and not reliant on outside variables. This will also be a huge help in allowing us test each function by itself.
One of our jobs as programmers is to write code that can be integrated easily within a team. In order to do that, we need to limit the number of globally declared variables in our code as much as possible, to avoid accidental name collisions.
Sloppy programmers use global variables, and you are not working so hard in order to be a sloppy programmer!
dentify the different ways to declare a variable in JavaScript
- Explain the differences between
const
,let
andvar
- Identify the difference between function and block-scoped variables
- Paraphrase the concept of hoisting in regards to variables
Calculating Closures
What is a closure? This question is one of the most frequent interview questions where JavaScript is involved. If you answer this question quickly and knowledgeably you'll look like a great candidate. We know you want to know it all so let's dive right in!
The official definition of a closure from MDN is, "A closure is the combination of a function and the lexical environment within which that function was declared." The practicality of how a closure is used it simple: a closure is when an inner function uses, or changes, variables in an outer function. Closures in JavaScript are incredibly important in terms of the creativity, flexibility and security of your code.
When you finish this reading you should be able to implement a closure and explain how that closure effects scope.
Closures and scope
Let's look at an example of a simple closure below:
function climbTree(treeType) {
let treeString = "You climbed a ";
function sayClimbTree() {
// this inner function has access to the variables in the outer scope
// in which is was defined - including any defined parameters
return treeString + treeType;
}
return sayClimbTree();
}
// We assign the result to a variable
const sayFunction = climbTree("Pine");
// So we can call it, and indeed the variables have been saved in the closure
// and the sayFunction prints out their values.
console.log(sayFunction); // You climbed a Pine
In the above snippet the sayClimbTree
function captures and uses the treeString
and treeType
variables within its own inner scope.
Let's go over some basic closure rules:
- Closures have access to any variables within its own, as well as any outer
function's, scope when they are declared. This is where the _lexical
environment comes in - the _lexical environment consists of any variables
available within the scope in which the closure was declared (which are the
local inner scope, outer function's scope, and global scope).
- A closure will keep reference to all the variables when it was defined **even
if the outer function has returned**.
Notice above that even though the above climbTree
had run its return
statement the inner function of sayClimbTree
still has access to the variables(treeString
and treeType
) from the outer scope where it was declared. So, even after an outer function has returned, an inner function will still have access to the outer function’s variables.
Let's look at another example of a closure:
function makeAdder(x) {
return function(y) {
return x + y;
};
}
const add5 = makeAdder(5);
console.log(add5(2)); // prints 7
In the above example the function the anonymous function within the makeAdder
function closes over the x
variable and utilizes it within the inner anonymous function. This allows us to do some pretty cool stuff like creating the add5
function above. Closures are your friend ❤️.
Applications of closures
Let's take a look at some of the common and practical applications of closures in JavaScript.
Private State
Information hiding is incredibly important in the world of software engineering. JavaScript as a language does not have a way of declaring a function as exclusively private, as can be done in other programming languages. We can however, use closures to create private state within a function.
The following code illustrates how to use closures to define functions that can emulate private functions and variables:
function createCounter() {
let count = 0;
return function() {
count++;
return count;
};
}
let counter = createCounter();
console.log(counter()); // => 1
console.log(counter()); // => 2
//we cannot reach the count variable!
counter.count; // undefined
let counter2 = createCounter();
console.log(counter2()); // => 1
In the above code we are storing the anonymous inner function inside the createCounter
function onto the variable counter
. The counter
variable is now a closure. The counter
variable closes over the inner count
value inside createCounter
even after createCounter
has returned.
By closing over (or capturing) the count
variable, each function that is return from createCounter
has a private, mutable state that cannot be accessed externally. There is no way any outside function beside the closure itself can access the count
state.
Passing Arguments Implicitly
We can use closures to pass down arguments to helper functions without explicitly passing them into that helper function.
function isPalindrome(string) {
function reverse() {
return string
.split("")
.reverse()
.join("");
}
return string === reverse();
}
How to implement a closure and explain how that closure effects scope.
Context in JavaScript
It's now time to dive into one of the most interesting concepts in JavaScript: the idea of context.
Programmers from the junior to senior level often confuse scope and context as the same thing - but that is not the case! Every function that is invoked has both a scope and a context associated with that function. Scope refers to the visibility and availability of variables, whereas context refers to the value of the this
keyword when code is executed.
When you finish this reading you should be able to:
- Define a method that references
this
on an object - Identify what
this
refers to in a code snippet - Utilize the built in
Function#bind
to maintain the context ofthis
What about this
?
When learning about objects we previously came across the idea of a method. A method is a function that is a value within an object and belongs to an object.
There will be times when you will have to know which object a method belongs to. The keyword this
exists in every function and it evaluates to the object that is currently invoking that function. So the value of this
relies entirely on where a function is invoked.
That may sound pretty abstract, so let's jump into an example:
let dog = {
name: "Bowser",
isSitting: true,
stand: function () {
this.isSitting = false;
return this.isSitting;
},
};
// Bowser starts out sitting
console.log(dog.isSitting); // prints `true`
// Let's make him stand
console.log(dog.stand()); // prints `false`
// He's actually standing now!
console.log(dog.isSitting); // prints `false`
Inside of a method, we can use the keyword this
to refer to the object that is calling that method! So when calling dog.stand()
and we invoke the code of the stand
method, this
will refer to the dog
object.
Still skeptical? Don't take our word for it, check this
(heh) out:
let dog = {
name: "Bowser",
test: function () {
return this === dog;
},
};
console.log(dog.test()); // prints `true`
In short, by using the this
keyword inside a method, we can refer to values within that object.
Let's look at another example of this:
let cat = {
purr: function () {
console.log("meow");
},
purrMore: function () {
this.purr();
},
};
cat.purrMore();
Through the this
variable, the purrMore
method can access the object it was called on. In purrMore
, we use this
to access the cat
object that has a purr
method. In other words, inside of the purrMore
function if we had tried to use purr()
instead of this.purr()
it would not work.
When we invoked the purrMore
function using cat.purrMore
we used a method-style invocation.
Method style invocations follow the format: object.method(args)
. You've already been doing this using built in data type methods! (i.e. Array#push
, String#toUpperCase
, etc.)
Using method-style invocation (note the dot notation) ensures the method will be invoked and that the this
within the method will be the object that method was called upon.
Now that we have gone over what this
refers to - you can have a full understanding of the definition of context. Context refers to the value of this
within a function and this
refers to where a function is invoked.
Issues with scope and context
In the case of context the value of this
is determined by how a function is invoked. In the above section we talked briefly about method-style invocation, where this
is set to the object the method was called upon.
Let's now talk about what this
is when using normal function style invocation.
If you run the following in Node:
function testMe() {
console.log(this); //
}
testMe(); // Object [global] {global: [Circular], etc.}
When you run the above testMe
function in Node you'll see that this
is set to the global
object. To reiterate: each function you invoke will have both a context and a scope. So even running functions in Node that are not defined explicitly on declared objects are run using the global object as their this
and therefore their context.
When methods have an unexpected context
So let's now look at what happens when we try to invoke a method using an unintended context.
Say we have a function that will change the name of a dog object:
let dog = {
name: "Bowser",
changeName: function () {
this.name = "Layla";
},
};
Now say we wanted to take the changeName
function above and call it somewhere else. Maybe we have a callback we'd like to pass it to or another object or something like that.
Let's take a look at what happens when we try to isolate and invoke just the changeName
function:
let dog = {
name: "Bowser",
changeName: function () {
this.name = "Layla";
},
};
// note this is **not invoked** - we are assigning the function itself
let change = dog.changeName;
console.log(change()); // undefined
// our dog still has the same name
console.log(dog); // { name: 'Bowser', changeName: [Function: changeName] }
// instead of changing the dog we changed the global name!!!
console.log(this); // Object [global] {etc, etc, etc, name: 'Layla'}
So in the above code notice how we stored the dog.changeName
function without invoking it to the variable change
. On the next line when we did invoke the change
function we can see that we did not actually change the dog
object like we intended to. We created a new key value pair for name
on the global object! This is because we invoked change without the context of a specific object (like dog
), so JavaScript used the only object available to it, the global object!
The above example might seem like an annoying inconvenience but let's take a look at what happens when calling something in the wrong context can be a big problem.
Using our cat
object from before:
let cat = {
purr: function () {
console.log("meow");
},
purrMore: function () {
this.purr();
},
};
let notACat = cat.purrMore;
console.log(notACat()); // TypeError: this.purr is not a function
So in the above code snippet we attempted to call the purrMore
function without the correct Object for context. Meaning we attempted to call the purrMore
function on the global object! Since the global object does not have a purr
method upon its this
it raised an error. This is a common problem when invoking methods: invoking methods without their proper context.
Let's look at one more example of confusing this
when using a callback. Incorrectly passing context is an inherent problem with callbacks. The global.setTimeout()
method on the global object is a popular way of setting a function to run on a timer. The global.setTimeout()
method accepts a callback and a number of milliseconds to wait before invoking the callback.
Let's look at a simple example:
let hello = function () {
console.log("hello!");
};
// global. is a method of the global object!
global.setTimeout(hello, 5000); // waits 5 seconds then prints "hello!"
Expanding on the global.setTimeout
method now using our cat
from before let's say we wanted our cat
to "meow" in 5 seconds instead of right now:
let cat = {
purr: function () {
console.log("meow");
},
purrMore: function () {
this.purr();
},
};
global.setTimeout(cat.purrMore, 5000); // 5 seconds later: TypeError: this.purr is not a function
So what happened there? We called cat.purrMore
so it should have the right context right? Noooooope. This is because cat.purrMore
is a callback in the above code! Meaning that when the global.setTimeout
function attempts to call the purrMore
function all it has reference to is the function itself. Since setTimeout
is on the global object that means that the global object will be the context for attempting to invoke the cat.purrMore
function.
Strictly protecting the global object
The accidental mutation of the global object when invoking functions in unintended contexts is one of the reasons JavaScript released "strict" mode in ECMAScript version 5. We won't dive too much into JavaScript's strict mode here, but it's important to know how strict mode can be used to protect the global object.
Writing and running code in strict mode is easy and much like writing code in "sloppy mode" (jargon for the normal JavaScript environment). We can run JavaScript in strict mode simply by adding the string "use strict" at the top of our file:
"use strict";
function hello() {
return "Hello!";
}
console.log(hello); // prints "Hello!"
One of the differences of strict mode becomes apparent when trying to access the global object. As we mentioned previously, the global object is the context of invoked functions in Node that are not defined explicitly on declared objects.
So referencing this
within a function using the global object as its context will give us access to the global object:
function hello() {
console.log(this);
}
hello(); // Object [global] {etc, etc, etc }
However, strict mode will no longer allow you access to the global object in functions via the this
keyword and will instead return undefined
:
"use strict";
function hello() {
console.log(this);
}
hello(); // undefined
Using strict mode can help us avoid scenarios where we accidentally would have mutated the global object. Let's take our example from earlier and try it in strict mode:
"use strict";
let dog = {
name: "Bowser",
changeName: function () {
this.name = "Layla";
},
};
// // note this is **not invoked** - we are assigning the function itself
let changeNameFunc = dog.changeName;
console.log(changeNameFunc()); // TypeError: Cannot set property 'name' of undefined
As you can see above, when we attempt to invoke the changeNameFunc
an error is thrown because referencing this
in strict mode will give us undefined
instead of the global object. The above behavior is helpful for catching otherwise tricky bugs.
If you'd like to learn more about strict mode we recommend checking out the documentation.
Changing context using bind
Good thing JavaScript has something that can solve this problem for us: what is known as the binding of a context to a function.
From the Function.prototype.bind()
, "The simplest use of bind()
is to make a function that, no matter how it is called, is called with a particular this
value".
Here is a preview of the syntax we use to bind
:
let aboundFunc = func.bind(context);
So when we call bind
we are returned what is called an exotic function. Which essentially means a function with it's this
bound no matter where that function is invoked.
Let's take a look at example at bind
in action:
let cat = {
purr: function () {
console.log("meow");
},
purrMore: function () {
this.purr();
},
};
let sayMeow = cat.purrMore;
console.log(sayMeow()); // TypeError: this.purr is not a function
// we can now use the built in Function.bind to ensure our context, our `this`,
// is the cat object
let boundCat = sayMeow.bind(cat);
// we still *need* to invoke the function
boundCat(); // prints "meow"
That is the magic of Function#bind
! It allows you choose the context for your function. You don't need to restrict the context you'd like to bind to either - you can bind
functions to any context.
Let's look at another example:
let cat = {
name: "Meowser",
sayName: function () {
console.log(this.name);
},
};
let dog = {
name: "Fido",
};
let sayNameFunc = cat.sayName;
let sayHelloCat = sayNameFunc.bind(cat);
sayHelloCat(); // prints Meowser
let sayHelloDog = sayNameFunc.bind(dog);
sayHelloDog(); // prints Fido
Let's now revisit our above example of losing context in a callback and fix our context! Using the global.setTimeout
function we want to call the cat.purrMore
function with the context bound to the cat object.
Here we go:
let cat = {
purr: function () {
console.log("meow");
},
purrMore: function () {
this.purr();
},
};
// here we will bind the cat.purrMore function to the context of the cat object
const boundPurr = cat.purrMore.bind(cat);
global.setTimeout(boundPurr, 5000); // prints 5 seconds later: meow
Binding with arguments
So far we've talking of one of the the common uses of the bind
function - binding a context to a function. However, bind will not only allow you to bind the context of a function but also to bind arguments to a function.
Here is the syntax for binding arguments to a function:
let aboundFunc = func.bind(context, arg1, arg2, etc...);
Following that train of logic let's look at example of binding arguments to a function, regardless of the context:
const sum = function (a, b) {
return a + b;
};
// here we are creating a new function named add3
// this function will bind the value 3 for the first argument
const add3 = sum.bind(null, 3);
// now when we invoke our new add3 function it will add 3 to
// one incoming argument
console.log(add3(10));
Note that in the above snippet where we bind
with null
we don’t actually use this
in the sum
function. However, since bind
requires a first argument we can put in null
as a place holder.
Above when we created the add3
function we were creating a new bound function where the context was null
, since the context won't matter, and the first argument will always be 3
for that function. Whenever we invoke the add3
function all other arguments will be passed in normally.
Using bind
like this gives you a lot of flexibility with your code. Allowing you to create independent functions that essentially do the same thing while keeping your code very DRY.
Here is another example:
const multiply = function (a, b) {
return a * b;
};
const double = multiply.bind(null, 2);
const triple = multiply.bind(null, 3);
console.log(double(3)); // 6
console.log(triple(3)); // 9
How to define a method that references this
on an object
- Identify what
this
refers to in a code snippet - How to utilize the built in
Function#bind
to maintain the context ofthis
Arrow Functions
Arrow functions, a.k.a. Fat Arrows (=>
), are a more concise way of declaring functions. Arrow functions were introduced in ES2015 as a way of solving many of the inconveniences of the normal callback function syntax.
Two major factors influenced the reason behind the desire for arrow functions: the need for shorter functions and behavior of this
and context.
When you finish this reading you should be able to:
- Define an arrow function
- Given an arrow function, deduce the value of
this
without executing the code
Arrow functions solving problems
Let's start by looking at the arrow function in action!
// function declaration
let average = function(num1, num2) {
let avg = (num1 + num2) / 2;
return avg;
};
// fat arrow function style!
let averageArrow = (num1, num2) => {
let avg = (num1 + num2) / 2;
return avg;
};
Both functions in the example above accomplish the same thing. However, the arrow syntax is a little shorter and easier to follow.
Anatomy of an arrow function
The syntax for a multiple statement arrow function is as follows:
(parameters, go, here) => {
statement1;
statement2;
return <a value>;
}
So let's look at a quick translation between a function declared with a function expression syntax and a fat arrow function. Take notice of the removal of the function
keyword, and the addition of the fat arrow (=>
).
function fullName(fname, lname) {
let str = "Hello " + fname + " " + lname;
return str;
}
// vs.
let fullNameArrow = (fname, lname) => {
let str = "Hello " + fname + " " + lname;
return str;
};
If there is only a single parameter you may omit the ( )
around the parameter declaration:
param1 => {
statement1;
return value;
};
If you have no parameters with an arrow function you must still use the ( )
:
// no parameters will use parenthesis
() => {
statements;
return value;
};
Let's see an example of an arrow function with a single parameter with no parenthesis:
const sayName = name => {
return "Hello " + name;
};
sayName("Jared"); // => "Hello Jared"
Single expression arrow functions
Reminder: In JavaScript, an expression is a line of code that returns a value. Statements are, more generally, any line of code.
One of the most fun things about single expression arrow functions is they allow for something previously unavailable in JavaScript: implicit returns. Meaning, in an arrow function with a single-expression block, the curly braces ({ }
) and the return
are keyword are implied.
argument => expression; // equal to (argument) => { return expression };
Look at the below example you can see how we use this snazzy implicit returns syntax:
const multiply = function(num1, num2) {
return num1 * num2;
};
// do not need to explicitly state return!
const arrowMultiply = (num1, num2) => num1 * num2;
However this doesn't work if the fat arrow uses multiple statements:
const halfMyAge = myAge => {
const age = myAge;
age / 2;
};
console.log(halfMyAge(30)); // "undefined"
To return a value from a fat arrow with multiple statements, you must explicitly return:
const halfMyAge = myAge => {
const age = myAge;
return age / 2;
};
console.log(halfMyAge(30)); // 15
Syntactic ambiguity with arrow functions
In Javascript, {}
can signify either an empty object or an empty block.
const ambiguousFunction = () => {};
Is ambiguousFunction
supposed to return an empty object or an empty code block? Confusing right? JavaScript standards state that the curly braces after a fat arrow evaluate to an empty block (which has the default value of undefined
):
ambiguousFunction(); // undefined
To make a single-expression fat arrow return an empty object, wrap that object within parentheses:
// this will implicitly return an empty object
const clearFunction = () => ({});
clearFunction(); // returns an object: {}
Arrow functions are anonymous
Fat arrows are anonymous, like their lambda
counterparts in other languages.
sayHello(name) => console.log("Hi, " + name); // SyntaxError
(name) => console.log("Hi, " + name); // this works!
If you want to name your function you must assign it to a variable:
const sayHello = name => console.log("Hi, " + name);
sayHello("Curtis"); // => Hi, Curtis
That's about all you need to know for arrow functions syntax-wise. Arrow functions aren't just a different way of writing functions, though. They behave differently too - especially when it comes to context!
Arrow functions with context
Arrow functions, unlike normal functions, carry over context, binding this
lexically. In other words, this
means the same thing inside an arrow function that it does outside of it. Unlike all other functions, the value of this
inside an arrow function is not dependent on how it is invoked.
Let's do a little compare and contrast to illustrate this point:
const testObj = {
name: "The original object!",
createFunc: function() {
return function() {
return this.name;
};
},
createArrowFunc: function() {
// the context within this function is the testObj
return () => {
return this.name;
};
}
};
const noName = testObj.createFunc();
const arrowName = testObj.createArrowFunc();
noName(); // undefined
arrowName(); // The original object!
Let's walk through what just happened - we created a testObj
with two methods that each returned an anonymous function. The difference between these two methods is that the createArrowFunc
function contained an arrow function inside it. When we invoked both methods we created two function - the noName
function creating it's own scope and context while the arrowName
kept the context of the function that created it (createArrowFunc
's context of testObj
).
An arrow function will always have the same context as the function that created it - giving it access to variables available in that context (like this.name
in this case!)
No binding in arrow functions
One thing to know about arrow functions is since they already have a bound context, unlike normal functions, you can't reassign this
. The this
in arrow functions is always what it was at the time that the arrow function was declared.
const returnName = () => this.name;
returnName(); // undefined
// arrow functions can't be bound
let tryToBind = returnName.bind({ name: "Party Wolf" }); // undefined
tryToBind(); // will still be undefined
How to define an arrow function
- how to deduce the value of
this
in an arrow function
Scope Problems
It's time to get some practice using scope in the wild! This task includes a link to download a zip
file with a number of problems.
Complete the problems in the order specified. In addition to the prompts available at the top of each file, Mocha specs are provided to test your work.
To get started, use the following commands:
-
cd
into the project directory -
npm install
to install any dependencies -
mocha
to run the test cases
WhiteBoarding Problem
Write a function named hiddenCounter()
. The hiddenCounter
function will start by declaring a variable that will keep track of a count and will be initially set to 0. Upon first invocation hiddenCounter
will return a function. Every subsequent invocation will increment the previously described count variable.
Explain how the closure you have created affects the scope of both functions.
Examples:
let hidden1 = hiddenCounter(); //returns a function
hidden1(); // returns 1
hidden1(); // returns 2
let hidden2 = hiddenCounter(); // returns a function
hidden2(); // returns 1
The Answer
function hiddenCounter() {
let count = 0;
// here we are returning an inner function that will create a closure by
// closing over the above count variable and changing it each time the
// the inner function is invoked
return () => (count += 1);
}
Top comments (0)