What's the difference between “LexicalEnvironment” and “VariableEnvironment” in spec

Question

I\'m reading the ECMAScript 2015 specification, and the terms "LexicalEnvironment" and "VariableEnvironment" are used many times. They are defined in tab

Answer 1

This is a hard one. I will try to explain with some simple examples. So one important thing, in this question is also to understand the execution context.

Lexical Environment

Means where you write something in the code is important. Not all programming languages are like that, but javascript is.

So if you have a function like

function hello() {
    var myVar = 'hello';
}

Now the variable myVar sits lexically inside the function. That's physically the code that you're writing. In short, if talking about lexical environment means where it is written and what surrounds it.

Variable Environment Every time you call a function a new execution context will be created. So even myVar is declared 3 times (see next example) they do not touch each other. That's when you talk about Variable Environment

function b() {
    var myVar;
    console.log('three', myVar) // three undefined 
                                // cause myVar is newly declared in b()
                                // but has no value
}

function a() {
    var myVar = 2;
    console.log('two', myVar) // two 2
    b();
}

var myVar = 1;
console.log('one', myVar) // one 1
a();
console.log('four', myVar) // one 1

Now you where asking for the difference which I guess it is just the theoretical talk about two things. But also the lexical environment kinda knows where the variables are sitting in memory.

So that is actually the answer to your question. But I will show some more examples just to make sure where things can go wrong with misunderstanding.

Because there is also this thing called hoisting in javascript which can give you errors if you write code at the wrong place. And it can have strange behaviour. The next examples are actually very simple but all depend on Lexical Environemnt, Variable Environment, Execution Context and hoisting

console.log(myVar); // undefined
var myVar = 'hello';
console.log(myVar); // hello

but

function a() {
    console.log(myVar) // gives Error myVar is not defined
}
a();

but again:

function a() {
    console.log(myVar); // undefined no Error
                        // cause it would take myVar from global
                        // execution context 
                        // but also no error cause we define it again
                        // in this function (hoisting)
    var myVar = 0;      // define myVar newly in this lexical environment
    console.log(myVar); // 0
}

var myVar = 'hello';
a();
console.log(myVar);     // hello

But again if we do like so

function a() {
    myVar = 0;           // overwrite global myVar
    console.log(myVar); // 0
}

var myVar = 'hello';
a();
console.log(myVar);     // 0 did you expect myVar to be 0 ?

Answer 2

I post the question to offical ECMA262 organization on github, this is the answer of littledan:

A LexicalEnvironment is a local lexical scope, e.g., for let-defined variables. If you define a variable with let in a catch block, it is only visible within the catch block, and to implement that in the spec, we use a LexicalEnvironment. VariableEnvironment is the scope for things like var-defined variables. vars can be thought of as "hoisting" to the top of the function. To implement this in the spec, we give functions a new VariableEnvironment, but say that blocks inherit the enclosing VariableEnvironment.