Hopefully by reading this post you’ll know the answers to:
- What is Scope?
- What is Global/Local Scope?
- What is a Namespace and how does it differ to Scope?
- What is the keyword and how does Scope affect it?
- What is Function/Lexical Scope?
- What are Closures?
- What is Public/Private Scope?
- How can I understand/create/do all of the above?
What is Scope?
Thinking about scope is easy, are we inside or ?
What is Global Scope?
Global scope is your best friend and your worst nightmare, learning to control your scopes is easy and in doing so, you’‘ll run into no issues with global scope problems (usually namespace clashes). You’ll often hear people saying “Global Scope is bad”, but never really justifying as to why. Global scope isn’t bad, you need it to create Modules/APIs that are accessible across scopes, you must use it to your advantage and not cause issues.
Everyone’s used jQuery before, as soon as you do this…
… we’re accessing jQuery in global scope, we can refer to this access as the . The namespace is sometimes an interchangeable word for scope, but usually the refers to the highest level scope. In this case, is in the global scope, and is also our namespace. The namespace is defined in the global scope, which acts as a namespace for the jQuery library as everything inside it becomes a descendent of that namespace.
What is Local Scope?
A local scope refers to any scope defined past the global scope. There is typically one global scope, and each function defined has its own (nested) local scope. Any function defined within another function has a local scope which is linked to the outer function.
If I define a function and create variables inside it, those variables becomes locally scoped. Take this example:
Any locally scoped items are not visible in the global scope - unless exposed, meaning if I define functions or variables within a new scope, it’s inaccessible outside of that current scope. A simple example of this is the following:
The variable is scoped locally, it isn’t exposed to the parent scope and therefore undefined.
It’s easy to create new scope and create local variables/functions/objects.
Whenever you see a function within another function, the inner function has access to the scope in the outer function, this is called Lexical Scope or Closure - also referred to as Static Scope. The easiest way to demonstrate that again:
You’ll notice that isn’t being called here, it’s simply defined. Its order of call also has effect on how the scoped variables react, here I’ve defined my function and called it under another statement:
Lexical scope is easy to work with, any variables/objects/functions defined in its parent scope, are available in the scope chain. For example:
The only important thing to remember is that Lexical scope does not work backwards. Here we can see how Lexical scope doesn’t work:
I can always return a reference to , but never the variable itself.
Closures ties in very closely with Lexical Scope. A better example of how the closure side of things works, can be seen when returning a function reference - a more practical usage. Inside our scope, we can return things so that they’re available in the parent scope:
The concept we’ve used here makes our scope inside inaccessible to the public scope. Calling the function alone will do nothing as it returns a function:
The function returns a function, which means it needs assignment, and then calling:
Okay, I lied, you can call it, and you may have seen functions like this, but this will call your closure:
AngularJS uses the above technique for its method, where you pass the current scope reference into the closure:
Meaning we could guess that their code would (over-simplified) look like this:
A function doesn’t have to return in order to be called a closure though. Simply accessing variables outside of the immediate lexical scope creates a closure.
Scope and ‘this’
Each scope binds a different value of depending on how the function is invoked. We’ve all used the keyword, but not all of us understand it and how it differs when invoked. By default refers to the outer most global object, the . We can easily show how invoking functions in different ways binds the value differently:
There are also problems that we run into when dealing with the value, for instance if I do this, even inside the same function the scope can be changed and the value can be changed:
So what’s happened here? We’ve created new scope which is not invoked from our event handler, so it defaults to the Object as expected. There are several things we can do if we want to access the proper value which isn’t affected by the new scope. You might have seen this before, where we can cache a reference to the value using a variable and refer to the lexical binding:
This is a neat little trick to be able to use the proper value and resolve problems with newly created scope.
Changing scope with .call(), .apply() and .bind()
The value here doesn’t refer to our elements, we’re not invoking anything or changing the scope. Let’s look at how we can change scope (well, it looks like we change scope, but what we’re really doing is changing the context of how the function is called).
.call() and .apply()
The and methods are really sweet, they allows you to pass in a scope to a function, which binds the correct value. Let’s manipulate the above function to make it so that our value is each element in the array:
You can see I’m passing in the current element in the Array iteration, , which changes the scope of the function so that the value becomes that iterated element. We can then use the binding if we wanted. We can use either or to change the scope, but any further arguments are where the two differ: takes individual arguments, comma separated, whereas takes an Array of arguments.
It’s important to remember that using or actually invokes your function, so instead of doing this:
You’ll let handle it and chain the method:
Unlike the above, using does not invoke a function, it merely binds the values before the function is invoked. It’s a real shame this was introduced in ECMAScript 5 and not earlier as this method is fantastic. As you know we can’t pass parameters into function references, something like this:
We can fix this, by creating a new function inside it:
But again this changes scope and we’re creating a needless function again, which will be costly on performance if we were inside a loop and binding event listeners. This is where shines through, as we can pass in arguments but the functions are not called:
The function isn’t invoked, and the scope can be changed if needed, but arguments are sat waiting to be passed in.
Private and Public Scope
We might then add a few functions for use in our app:
But when we come to calling our function, it would be out of scope:
Success! We’ve created private scope. But what if I want the function to be public? There’s a great pattern (called the Module Pattern [and Revealing Module Pattern]) which allows us to scope our functions correctly, using private and public scope and an . Here I grab my global namespace, called , which contains all of my relevant code for that module:
The statement here is what returns our methods, which are accessible in the global scope - but are . This means our Module takes care of our namespace, and can contain as many methods as we want. We can extend the Module as we wish:
So what about private methods? This is where a lot of developers go wrong and pollute the global namespace by dumping all their functions in the global scope. Functions that help our code work do not need to be in the global scope, only the API calls do - things that need to be accessed globally in order to work. Here’s how we can create private scope, by not returning functions:
This means that can be called, but cannot, as it’s privately scoped! These privately scoped functions are things like helpers, addClass, removeClass, Ajax/XHR calls, Arrays, Objects, anything you can think of.
Here’s an interesting twist though, anything in the same scope has access to anything in the same scope, even after the function has been returned. Which means, our methods have access to our ones, so they can still interact but are unaccessible in the global scope.
Here’s an example of returning an Object, making use of and methods:
One neat naming convention is to begin methods with an underscore, which visually helps you differentiate between public and private:
This helps us when returning an anonymous , which the Module can use in Object fashion as we can simply assign the function references:
It helps to think of as equivalent to (if you set the parameter in that function to ). The benefit of is that it allows you to specify more parameters (e.g. the environment), so I prefer to use over in most cases.
Using and means that "enclosing environments of the supplied environment are searched until the variable 'x' is encountered." In other words, it will keep going through the environments in order until it finds a variable with that name, and it will assign it to that. This can be within the scope of a function, or in the global environment.
In order to understand what these functions do, you need to also understand R environments (e.g. using ).
I regularly use these functions when I'm running a large simulation and I want to save intermediate results. This allows you to create the object outside the scope of the given function or loop. That's very helpful, especially if you have any concern about a large loop ending unexpectedly (e.g. a database disconnection), in which case you could lose everything in the process. This would be equivalent to writing your results out to a database or file during a long running process, except that it's storing the results within the R environment instead.
My primary warning with this: be careful because you're now working with global variables, especially when using . That means that you can end up with situations where a function is using an object value from the environment, when you expected it to be using one that was supplied as a parameter. This is one of the main things that functional programming tries to avoid (see side effects). I avoid this problem by assigning my values to a unique variable names (using paste with a set or unique parameters) that are never used within the function, but just used for caching and in case I need to recover later on (or do some meta-analysis on the intermediate results).