I have a javascript file that reads another file which may contain javascript fragments that need to be eval()-ed. The script fragments are supposed to conform to a strict subset of javascript that limits what they can do and which variables they can change, but I want to know if there is some way to enforce this by preventing the eval from seeing variables in the global scope. Something like the following:
function safeEval( fragment )
{
var localVariable = g_Variable;
{
// do magic scoping here so that the eval fragment can see localVariable
// but not g_Variable or anything else outside function scope
eval( fragment );
}
}
The actual code doesn't need to look like this--I'm open to any and all weird tricks with closures, etc. But I do want to know if this is even possible.
Short answer: No. If it's in the global scope, it's available to anything.
Long answer: if you're eval()ing untrusted code that really wants to read or mess with your execution environment, you're screwed. But if you own and trust all code being executed, including that being eval()ed, you can fake it by overriding the execution context:
function maskedEval(scr)
{
// set up an object to serve as the context for the code
// being evaluated.
var mask = {};
// mask global properties
for (p in this)
mask[p] = undefined;
// execute script in private context
(new Function( "with(this) { " + scr + "}")).call(mask);
}
Again, I must stress:
This will only serve to shield trusted code from the context in which it is executed. If you don't trust the code, DO NOT eval() it (or pass it to new Function(), or use it in any other way that behaves like eval()).
Shog9♦'s Answer is great. But if your code is just an expression, the code will be executed and nothing will be returned. For expressions, use
function evalInContext(context, js) {
return eval('with(context) { ' + js + ' }');
}
Here is how to use it:
var obj = {key: true};
evalInContext(obj, 'key ? "YES" : "NO"');
It will return "YES".
If you are not sure if the code to be executed is expressions or statements, you can combine them:
function evalInContext(context, js) {
var value;
try {
// for expressions
value = eval('with(context) { ' + js + ' }');
} catch (e) {
if (e instanceof SyntaxError) {
try {
// for statements
value = (new Function('with(this) { ' + js + ' }')).call(context);
} catch (e) {}
}
}
return value;
}
Similar to the dynamic function wrapping script in a with block approach above, this allows you to add pseudo-globals to the code you want to execute. You can "hide" specific things by adding them to the context.
function evalInContext(source, context) {
source = '(function(' + Object.keys(context).join(', ') + ') {' + source + '})';
var compiled = eval(source);
return compiled.apply(context, values());
// you likely don't need this - use underscore, jQuery, etc
function values() {
var result = [];
for (var property in context)
if (context.hasOwnProperty(property))
result.push(context[property]);
return result;
}
}
See http://jsfiddle.net/PRh8t/ for an example. Note that Object.keys is not supported in all browsers.
You cant limit the scope of eval
btw see this post
There may be some other way to accomplish what it is you want accomplish in the grand scheme of things but you cannot limit the scope of eval in any way. You may be able to hide certain variables as pseudo private variables in javascript, but I dont think this is what you're going for.
There is a project called Google Caja. You can "sandbox" third party javascript using Caja. https://developers.google.com/caja/
Here's an idea. What if you used a static analyzer (something you could build with esprima, for example) to determine which outside variables the eval'd code uses, and alias them. By "outside code" i mean variables the eval'd code uses but does not declare. Here's an example:
eval(safeEval(
"var x = window.theX;"
+"y = Math.random();"
+"eval('window.z = 500;');"))
where safeEval returns the javascript string modified with a context that blocks access to outside variables:
";(function(y, Math, window) {"
+"var x = window.theX;"
+"y = Math.random();"
+"eval(safeEval('window.z = 500;');"
"})();"
There are a couple things you can do now with this:
You can ensure that eval'd code can't read the values of outside variables, nor write to them (by passing undefined as the function arguments, or not passing arguments). Or you could simply throw an exception in cases where variables are being unsafely accessed.
You also ensure that variables created by eval don't affect the surrounding scope
You could allow eval to create variables in the surrounding scope by declaring those variables outside the closure instead of as function parameters
You could allow read-only access by copying values of outside variables and using them as arguments to the function
You could allow read-write access to specific variables by telling safeEval to not alias those particular names
You can detect cases where the eval does not modify a particular variable and allow it to be automatically excluded from being aliased (eg. Math in this case, is not being modified)
You could give the eval a context in which to run, by passing in argument values that may be different than the surrounding context
You could capture context changes by also returning the function arguments from the function so you can examine them outside the eval.
Note that the use of eval is a special case, since by its nature, it effectively can't be wrapped in another function (which is why we have to do eval(safeEval(...))).
Of course, doing all this work may slow down your code, but there are certainly places where the hit won't matter. Hope this helps someone. And if anyone creates a proof of concept, I'd love to see a link to it here ; )
Don't execute code you don't trust. Globals will always be accessible.
If you do trust the code, you can execute it with particular variables in it's scope as follows:
(new Function("a", "b", "alert(a + b);"))(1, 2);
this is equivalent to:
(function (a, b) {
alert(a + b);
})(1, 2);
I accidentally found out I can use Proxy to restrict the scope object, it seems be a lot easier to mask the variable out of the scope. I'm not sure if this method have disadvantages, but so far its work well for me.
function maskedEval(src, ctx = {})
{
ctx = new Proxy(ctx, {
has: () => true
})
// execute script in private context
let func = (new Function("with(this) { " + src + "}"));
func.call(ctx);
}
a = 1;
maskedEval("console.log(a)", { console });
maskedEval("console.log(a)", { console, a: 22});
maskedEval("a = 1", { a: 22 })
console.log(a)
Don't use eval. There's an alternative, js.js: JS interpreter written in JS, so that you can run JS programs in any environment you've managed to setup. Here's an example of its API from the project page:
var jsObjs = JSJS.Init();
var rval = JSJS.EvaluateScript(jsObjs.cx, jsObjs.glob, "1 + 1");
var d = JSJS.ValueToNumber(jsObjs.cx, rval);
window.alert(d); // 2
JSJS.End(jsObjs);
Nothing scary, as you can see.
Related
In javascript, when would you want to use this:
(function(){
//Bunch of code...
})();
over this:
//Bunch of code...
It's all about variable scoping. Variables declared in the self executing function are, by default, only available to code within the self executing function. This allows code to be written without concern of how variables are named in other blocks of JavaScript code.
For example, as mentioned in a comment by Alexander:
(function() {
var foo = 3;
console.log(foo);
})();
console.log(foo);
This will first log 3 and then throw an error on the next console.log because foo is not defined.
Simplistic. So very normal looking, its almost comforting:
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
However, what if I include a really handy javascript library to my page that translates advanced characters into their base level representations?
Wait... what?
I mean, if someone types in a character with some kind of accent on it, but I only want 'English' characters A-Z in my program? Well... the Spanish 'ñ' and French 'é' characters can be translated into base characters of 'n' and 'e'.
So someone nice person has written a comprehensive character converter out there that I can include in my site... I include it.
One problem: it has a function in it called 'name' same as my function.
This is what's called a collision. We've got two functions declared in the same scope with the same name. We want to avoid this.
So we need to scope our code somehow.
The only way to scope code in javascript is to wrap it in a function:
function main() {
// We are now in our own sound-proofed room and the
// character-converter library's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
That might solve our problem. Everything is now enclosed and can only be accessed from within our opening and closing braces.
We have a function in a function... which is weird to look at, but totally legal.
Only one problem. Our code doesn't work.
Our userName variable is never echoed into the console!
We can solve this issue by adding a call to our function after our existing code block...
function main() {
// We are now in our own sound-proofed room and the
// character-converter libarary's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
main();
Or before!
main();
function main() {
// We are now in our own sound-proofed room and the
// character-converter libarary's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
A secondary concern: What are the chances that the name 'main' hasn't been used yet? ...so very, very slim.
We need MORE scoping. And some way to automatically execute our main() function.
Now we come to auto-execution functions (or self-executing, self-running, whatever).
((){})();
The syntax is awkward as sin. However, it works.
When you wrap a function definition in parentheses, and include a parameter list (another set or parentheses!) it acts as a function call.
So lets look at our code again, with some self-executing syntax:
(function main() {
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
)();
So, in most tutorials you read, you will now be bombarded with the term 'anonymous self-executing' or something similar.
After many years of professional development, I strongly urge you to name every function you write for debugging purposes.
When something goes wrong (and it will), you will be checking the backtrace in your browser. It is always easier to narrow your code issues when the entries in the stack trace have names!
Self-invocation (also known as
auto-invocation) is when a function
executes immediately upon its
definition. This is a core pattern and
serves as the foundation for many
other patterns of JavaScript
development.
I am a great fan :) of it because:
It keeps code to a minimum
It enforces separation of behavior from presentation
It provides a closure which prevents naming conflicts
Enormously – (Why you should say its good?)
It’s about defining and executing a function all at once.
You could have that self-executing function return a value and pass the function as a param to another function.
It’s good for encapsulation.
It’s also good for block scoping.
Yeah, you can enclose all your .js files in a self-executing function and can prevent global namespace pollution. ;)
More here.
Namespacing. JavaScript's scopes are function-level.
I can't believe none of the answers mention implied globals.
The (function(){})() construct does not protect against implied globals, which to me is the bigger concern, see http://yuiblog.com/blog/2006/06/01/global-domination/
Basically the function block makes sure all the dependent "global vars" you defined are confined to your program, it does not protect you against defining implicit globals. JSHint or the like can provide recommendations on how to defend against this behavior.
The more concise var App = {} syntax provides a similar level of protection, and may be wrapped in the function block when on 'public' pages. (see Ember.js or SproutCore for real world examples of libraries that use this construct)
As far as private properties go, they are kind of overrated unless you are creating a public framework or library, but if you need to implement them, Douglas Crockford has some good ideas.
I've read all answers, something very important is missing here, I'll KISS. There are 2 main reasons, why I need Self-Executing Anonymous Functions, or better said "Immediately-Invoked Function Expression (IIFE)":
Better namespace management (Avoiding Namespace Pollution -> JS Module)
Closures (Simulating Private Class Members, as known from OOP)
The first one has been explained very well. For the second one, please study following example:
var MyClosureObject = (function (){
var MyName = 'Michael Jackson RIP';
return {
getMyName: function () { return MyName;},
setMyName: function (name) { MyName = name}
}
}());
Attention 1: We are not assigning a function to MyClosureObject, further more the result of invoking that function. Be aware of () in the last line.
Attention 2: What do you additionally have to know about functions in Javascript is that the inner functions get access to the parameters and variables of the functions, they are defined within.
Let us try some experiments:
I can get MyName using getMyName and it works:
console.log(MyClosureObject.getMyName());
// Michael Jackson RIP
The following ingenuous approach would not work:
console.log(MyClosureObject.MyName);
// undefined
But I can set an another name and get the expected result:
MyClosureObject.setMyName('George Michael RIP');
console.log(MyClosureObject.getMyName());
// George Michael RIP
Edit: In the example above MyClosureObject is designed to be used without the newprefix, therefore by convention it should not be capitalized.
Scope isolation, maybe. So that the variables inside the function declaration don't pollute the outer namespace.
Of course, on half the JS implementations out there, they will anyway.
Is there a parameter and the "Bunch of code" returns a function?
var a = function(x) { return function() { document.write(x); } }(something);
Closure. The value of something gets used by the function assigned to a. something could have some varying value (for loop) and every time a has a new function.
Here's a solid example of how a self invoking anonymous function could be useful.
for( var i = 0; i < 10; i++ ) {
setTimeout(function(){
console.log(i)
})
}
Output: 10, 10, 10, 10, 10...
for( var i = 0; i < 10; i++ ) {
(function(num){
setTimeout(function(){
console.log(num)
})
})(i)
}
Output: 0, 1, 2, 3, 4...
Short answer is : to prevent pollution of the Global (or higher) scope.
IIFE (Immediately Invoked Function Expressions) is the best practice for writing scripts as plug-ins, add-ons, user scripts or whatever scripts are expected to work with other people's scripts. This ensures that any variable you define does not give undesired effects on other scripts.
This is the other way to write IIFE expression. I personally prefer this following method:
void function() {
console.log('boo!');
// expected output: "boo!"
}();
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/void
From the example above it is very clear that IIFE can also affect efficiency and performance, because the function that is expected to be run only once will be executed once and then dumped into the void for good. This means that function or method declaration does not remain in memory.
One difference is that the variables that you declare in the function are local, so they go away when you exit the function and they don't conflict with other variables in other or same code.
First you must visit MDN IIFE , Now some points about this
this is Immediately Invoked Function Expression. So when your javascript file invoked from HTML this function called immediately.
This prevents accessing variables within the IIFE idiom as well as polluting the global scope.
Self executing function are used to manage the scope of a Variable.
The scope of a variable is the region of your program in which it is defined.
A global variable has global scope; it is defined everywhere in your JavaScript code and can be accessed from anywhere within the script, even in your functions. On the other hand, variables declared within a function are defined only within the body of the function.
They are local variables, have local scope and can only be accessed within that function. Function parameters also count as local variables and are defined only within the body of the function.
As shown below, you can access the global variables variable inside your function and also note that within the body of a function, a local variable takes precedence over a global variable with the same name.
var globalvar = "globalvar"; // this var can be accessed anywhere within the script
function scope() {
alert(globalvar);
var localvar = "localvar"; //can only be accessed within the function scope
}
scope();
So basically a self executing function allows code to be written without concern of how variables are named in other blocks of javascript code.
Since functions in Javascript are first-class object, by defining it that way, it effectively defines a "class" much like C++ or C#.
That function can define local variables, and have functions within it. The internal functions (effectively instance methods) will have access to the local variables (effectively instance variables), but they will be isolated from the rest of the script.
Self invoked function in javascript:
A self-invoking expression is invoked (started) automatically, without being called. A self-invoking expression is invoked right after its created. This is basically used for avoiding naming conflict as well as for achieving encapsulation. The variables or declared objects are not accessible outside this function. For avoiding the problems of minimization(filename.min) always use self executed function.
(function(){
var foo = {
name: 'bob'
};
console.log(foo.name); // bob
})();
console.log(foo.name); // Reference error
Actually, the above function will be treated as function expression without a name.
The main purpose of wrapping a function with close and open parenthesis is to avoid polluting the global space.
The variables and functions inside the function expression became private (i.e) they will not be available outside of the function.
Given your simple question: "In javascript, when would you want to use this:..."
I like #ken_browning and #sean_holding's answers, but here's another use-case that I don't see mentioned:
let red_tree = new Node(10);
(async function () {
for (let i = 0; i < 1000; i++) {
await red_tree.insert(i);
}
})();
console.log('----->red_tree.printInOrder():', red_tree.printInOrder());
where Node.insert is some asynchronous action.
I can't just call await without the async keyword at the declaration of my function, and i don't need a named function for later use, but need to await that insert call or i need some other richer features (who knows?).
It looks like this question has been answered all ready, but I'll post my input anyway.
I know when I like to use self-executing functions.
var myObject = {
childObject: new function(){
// bunch of code
},
objVar1: <value>,
objVar2: <value>
}
The function allows me to use some extra code to define the childObjects attributes and properties for cleaner code, such as setting commonly used variables or executing mathematic equations; Oh! or error checking. as opposed to being limited to nested object instantiation syntax of...
object: {
childObject: {
childObject: {<value>, <value>, <value>}
},
objVar1: <value>,
objVar2: <value>
}
Coding in general has a lot of obscure ways of doing a lot of the same things, making you wonder, "Why bother?" But new situations keep popping up where you can no longer rely on basic/core principals alone.
You can use this function to return values :
var Test = (function (){
const alternative = function(){ return 'Error Get Function '},
methods = {
GetName: alternative,
GetAge:alternative
}
// If the condition is not met, the default text will be returned
// replace to 55 < 44
if( 55 > 44){
// Function one
methods.GetName = function (name) {
return name;
};
// Function Two
methods.GetAge = function (age) {
return age;
};
}
return methods;
}());
// Call
console.log( Test.GetName("Yehia") );
console.log( Test.GetAge(66) );
Use of this methodology is for closures. Read this link for more about closures.
IIRC it allows you to create private properties and methods.
I am trying to figure out how to extend :
var outdatedBrowser = function(options) {
function startStylesAndEvents() {
console.log("bleh");
}
}
I am trying to overwrite the function startStylesAndEvents without touching the source code of the library : https://github.com/burocratik/outdated-browser/blob/develop/outdatedbrowser/outdatedbrowser.js
So when I call:
outdatedBrowser({
bgColor: '#f25648',
color: '#ffffff',
lowerThan: 'transform',
languagePath: 'your_path/outdatedbrowser/lang/en.html'
})
and it uses the startStylesAndEvents function, it uses mine instead of theirs...
Thanks!
Without modifying the original source? You can't.
All of JavaScript scoping is based on functions (ignoring let, const and class for the moment). If a value is declared inside of a function, it cannot be accessed outside of that function unless it is returned from the function or modifies some external value.
For example, imagine a function like this:
function doStuff() {
var times = 10;
for (var i = 0; i < times; i++) {
doThing(i);
}
}
Your question is semantically identical to asking how to change times. It just can't be done.
The inner function is contained within a closure, which you don't have access to. Unfortunately (contrary to the "nothing is impossible" ideology) this can't be done at runtime.
That is what you might call a 'private' function. The function is stored just like any other variable. In JavaScript that generally means it has to be 'local scope' (instead of a member variable) so it cannot be overridden. (if only they had used this.functionName... then you could override more easily)
The good news is, there is a hack which seems to be cross-browser compatible. (tested in IE 11, with emulation options back to IE 5!) In JavaScript, you can replace the actual 'source code' of the function itself. (a bit different than a proper override)
var newInnerFunction = function () { // intentionally anonymous for use in eval
}
var overriddenFunction = eval(
outdatedBrowser.toString() // expecting outdatedBrowser to be anonymous for use in eval
.replace('{', '{var startStylesAndEvents=' + newInnerFunction.toString() + ';')
)
Note that outdatedBrowser is an anonymous function (simply function()). If it were named, then the use of eval would have the side effect of adding the new function to the namespace under its original name. An additional replace could take care of that if it were an issue.
Most probably, you can't. But it's not completely impossible.
For example, if you call startStylesAndEvents inside a with statement whose scope object has been leaked to the outside
var scope = Object.create(null);
var outdatedBrowser = function(options) {
function startStylesAndEvents() {
console.log("bleh");
}
with(scope) {
startStylesAndEvents(); // You expect this to be the private function above
}
}
outdatedBrowser(); // "bleh"
Then, you can hijack calls to startStylesAndEvents:
scope.startStylesAndEvents = function() {
console.log("blah");
};
outdatedBrowser(); // "blah"
Don't do this, of course. It's evil, slow and not allowed in strict mode.
In javascript, when would you want to use this:
(function(){
//Bunch of code...
})();
over this:
//Bunch of code...
It's all about variable scoping. Variables declared in the self executing function are, by default, only available to code within the self executing function. This allows code to be written without concern of how variables are named in other blocks of JavaScript code.
For example, as mentioned in a comment by Alexander:
(function() {
var foo = 3;
console.log(foo);
})();
console.log(foo);
This will first log 3 and then throw an error on the next console.log because foo is not defined.
Simplistic. So very normal looking, its almost comforting:
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
However, what if I include a really handy javascript library to my page that translates advanced characters into their base level representations?
Wait... what?
I mean, if someone types in a character with some kind of accent on it, but I only want 'English' characters A-Z in my program? Well... the Spanish 'ñ' and French 'é' characters can be translated into base characters of 'n' and 'e'.
So someone nice person has written a comprehensive character converter out there that I can include in my site... I include it.
One problem: it has a function in it called 'name' same as my function.
This is what's called a collision. We've got two functions declared in the same scope with the same name. We want to avoid this.
So we need to scope our code somehow.
The only way to scope code in javascript is to wrap it in a function:
function main() {
// We are now in our own sound-proofed room and the
// character-converter library's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
That might solve our problem. Everything is now enclosed and can only be accessed from within our opening and closing braces.
We have a function in a function... which is weird to look at, but totally legal.
Only one problem. Our code doesn't work.
Our userName variable is never echoed into the console!
We can solve this issue by adding a call to our function after our existing code block...
function main() {
// We are now in our own sound-proofed room and the
// character-converter libarary's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
main();
Or before!
main();
function main() {
// We are now in our own sound-proofed room and the
// character-converter libarary's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
A secondary concern: What are the chances that the name 'main' hasn't been used yet? ...so very, very slim.
We need MORE scoping. And some way to automatically execute our main() function.
Now we come to auto-execution functions (or self-executing, self-running, whatever).
((){})();
The syntax is awkward as sin. However, it works.
When you wrap a function definition in parentheses, and include a parameter list (another set or parentheses!) it acts as a function call.
So lets look at our code again, with some self-executing syntax:
(function main() {
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
)();
So, in most tutorials you read, you will now be bombarded with the term 'anonymous self-executing' or something similar.
After many years of professional development, I strongly urge you to name every function you write for debugging purposes.
When something goes wrong (and it will), you will be checking the backtrace in your browser. It is always easier to narrow your code issues when the entries in the stack trace have names!
Self-invocation (also known as
auto-invocation) is when a function
executes immediately upon its
definition. This is a core pattern and
serves as the foundation for many
other patterns of JavaScript
development.
I am a great fan :) of it because:
It keeps code to a minimum
It enforces separation of behavior from presentation
It provides a closure which prevents naming conflicts
Enormously – (Why you should say its good?)
It’s about defining and executing a function all at once.
You could have that self-executing function return a value and pass the function as a param to another function.
It’s good for encapsulation.
It’s also good for block scoping.
Yeah, you can enclose all your .js files in a self-executing function and can prevent global namespace pollution. ;)
More here.
Namespacing. JavaScript's scopes are function-level.
I can't believe none of the answers mention implied globals.
The (function(){})() construct does not protect against implied globals, which to me is the bigger concern, see http://yuiblog.com/blog/2006/06/01/global-domination/
Basically the function block makes sure all the dependent "global vars" you defined are confined to your program, it does not protect you against defining implicit globals. JSHint or the like can provide recommendations on how to defend against this behavior.
The more concise var App = {} syntax provides a similar level of protection, and may be wrapped in the function block when on 'public' pages. (see Ember.js or SproutCore for real world examples of libraries that use this construct)
As far as private properties go, they are kind of overrated unless you are creating a public framework or library, but if you need to implement them, Douglas Crockford has some good ideas.
I've read all answers, something very important is missing here, I'll KISS. There are 2 main reasons, why I need Self-Executing Anonymous Functions, or better said "Immediately-Invoked Function Expression (IIFE)":
Better namespace management (Avoiding Namespace Pollution -> JS Module)
Closures (Simulating Private Class Members, as known from OOP)
The first one has been explained very well. For the second one, please study following example:
var MyClosureObject = (function (){
var MyName = 'Michael Jackson RIP';
return {
getMyName: function () { return MyName;},
setMyName: function (name) { MyName = name}
}
}());
Attention 1: We are not assigning a function to MyClosureObject, further more the result of invoking that function. Be aware of () in the last line.
Attention 2: What do you additionally have to know about functions in Javascript is that the inner functions get access to the parameters and variables of the functions, they are defined within.
Let us try some experiments:
I can get MyName using getMyName and it works:
console.log(MyClosureObject.getMyName());
// Michael Jackson RIP
The following ingenuous approach would not work:
console.log(MyClosureObject.MyName);
// undefined
But I can set an another name and get the expected result:
MyClosureObject.setMyName('George Michael RIP');
console.log(MyClosureObject.getMyName());
// George Michael RIP
Edit: In the example above MyClosureObject is designed to be used without the newprefix, therefore by convention it should not be capitalized.
Scope isolation, maybe. So that the variables inside the function declaration don't pollute the outer namespace.
Of course, on half the JS implementations out there, they will anyway.
Is there a parameter and the "Bunch of code" returns a function?
var a = function(x) { return function() { document.write(x); } }(something);
Closure. The value of something gets used by the function assigned to a. something could have some varying value (for loop) and every time a has a new function.
Here's a solid example of how a self invoking anonymous function could be useful.
for( var i = 0; i < 10; i++ ) {
setTimeout(function(){
console.log(i)
})
}
Output: 10, 10, 10, 10, 10...
for( var i = 0; i < 10; i++ ) {
(function(num){
setTimeout(function(){
console.log(num)
})
})(i)
}
Output: 0, 1, 2, 3, 4...
Short answer is : to prevent pollution of the Global (or higher) scope.
IIFE (Immediately Invoked Function Expressions) is the best practice for writing scripts as plug-ins, add-ons, user scripts or whatever scripts are expected to work with other people's scripts. This ensures that any variable you define does not give undesired effects on other scripts.
This is the other way to write IIFE expression. I personally prefer this following method:
void function() {
console.log('boo!');
// expected output: "boo!"
}();
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/void
From the example above it is very clear that IIFE can also affect efficiency and performance, because the function that is expected to be run only once will be executed once and then dumped into the void for good. This means that function or method declaration does not remain in memory.
One difference is that the variables that you declare in the function are local, so they go away when you exit the function and they don't conflict with other variables in other or same code.
First you must visit MDN IIFE , Now some points about this
this is Immediately Invoked Function Expression. So when your javascript file invoked from HTML this function called immediately.
This prevents accessing variables within the IIFE idiom as well as polluting the global scope.
Self executing function are used to manage the scope of a Variable.
The scope of a variable is the region of your program in which it is defined.
A global variable has global scope; it is defined everywhere in your JavaScript code and can be accessed from anywhere within the script, even in your functions. On the other hand, variables declared within a function are defined only within the body of the function.
They are local variables, have local scope and can only be accessed within that function. Function parameters also count as local variables and are defined only within the body of the function.
As shown below, you can access the global variables variable inside your function and also note that within the body of a function, a local variable takes precedence over a global variable with the same name.
var globalvar = "globalvar"; // this var can be accessed anywhere within the script
function scope() {
alert(globalvar);
var localvar = "localvar"; //can only be accessed within the function scope
}
scope();
So basically a self executing function allows code to be written without concern of how variables are named in other blocks of javascript code.
Since functions in Javascript are first-class object, by defining it that way, it effectively defines a "class" much like C++ or C#.
That function can define local variables, and have functions within it. The internal functions (effectively instance methods) will have access to the local variables (effectively instance variables), but they will be isolated from the rest of the script.
Self invoked function in javascript:
A self-invoking expression is invoked (started) automatically, without being called. A self-invoking expression is invoked right after its created. This is basically used for avoiding naming conflict as well as for achieving encapsulation. The variables or declared objects are not accessible outside this function. For avoiding the problems of minimization(filename.min) always use self executed function.
(function(){
var foo = {
name: 'bob'
};
console.log(foo.name); // bob
})();
console.log(foo.name); // Reference error
Actually, the above function will be treated as function expression without a name.
The main purpose of wrapping a function with close and open parenthesis is to avoid polluting the global space.
The variables and functions inside the function expression became private (i.e) they will not be available outside of the function.
Given your simple question: "In javascript, when would you want to use this:..."
I like #ken_browning and #sean_holding's answers, but here's another use-case that I don't see mentioned:
let red_tree = new Node(10);
(async function () {
for (let i = 0; i < 1000; i++) {
await red_tree.insert(i);
}
})();
console.log('----->red_tree.printInOrder():', red_tree.printInOrder());
where Node.insert is some asynchronous action.
I can't just call await without the async keyword at the declaration of my function, and i don't need a named function for later use, but need to await that insert call or i need some other richer features (who knows?).
It looks like this question has been answered all ready, but I'll post my input anyway.
I know when I like to use self-executing functions.
var myObject = {
childObject: new function(){
// bunch of code
},
objVar1: <value>,
objVar2: <value>
}
The function allows me to use some extra code to define the childObjects attributes and properties for cleaner code, such as setting commonly used variables or executing mathematic equations; Oh! or error checking. as opposed to being limited to nested object instantiation syntax of...
object: {
childObject: {
childObject: {<value>, <value>, <value>}
},
objVar1: <value>,
objVar2: <value>
}
Coding in general has a lot of obscure ways of doing a lot of the same things, making you wonder, "Why bother?" But new situations keep popping up where you can no longer rely on basic/core principals alone.
You can use this function to return values :
var Test = (function (){
const alternative = function(){ return 'Error Get Function '},
methods = {
GetName: alternative,
GetAge:alternative
}
// If the condition is not met, the default text will be returned
// replace to 55 < 44
if( 55 > 44){
// Function one
methods.GetName = function (name) {
return name;
};
// Function Two
methods.GetAge = function (age) {
return age;
};
}
return methods;
}());
// Call
console.log( Test.GetName("Yehia") );
console.log( Test.GetAge(66) );
Use of this methodology is for closures. Read this link for more about closures.
IIRC it allows you to create private properties and methods.
I'm building a library (https://github.com/OscarGodson/storageLocker), a localStorage wrapper to be more exact, but because this is my first try at OO JavaScript I'm still learning and I have a couple questions.
I've seen in other libraries that sometimes they wrap them in a anonymous function. Do I, or should I, do that with this? And if so, how without breaking anything?
For the internal API (basically, the internal functions) how should I write them? Should add them to the main object e.g. storageLocker.prototype.myInternalFunction() or just myInternalFunction() randomly in my script? I didn't want the functions to be global though... One of the functions for example just checks a bunch of items in the JSON, sees if their objects, and then checks what the object type is (like Date()) and then converts it.
How/where should I add global, to my script, vars? e.g. i have a var called patterns that is something like var patterns = {"date":/\/Date\(([0-9]+)\)\//} how should I add that into my script?
Thanks a lot. I want to write my script the right way so im asking you guys. I don't know of any JS guys locally that do any OO JS, they're all old school types...
I'd say:
1) the purpose of this technique is not pollute the global namespace. That is a good thing.
In the example below you can see that all your interaction with the library is via one object MyLibrary. Public API is the return value of the anonymous function.
var MyLibrary = function() {
// private
this.InternalVariable = 'some value';
function internalFunction(x,y) {
return x + y;
}
function getInternalVariable() {
return this.InternalVariable;
}
// public
return {
publicVariable : '1.0',
publicFunction : function(x,y) {
return x + y
},
accessInternalVariable : function() {
return getInternalVariable();
}
}
}();
2) see also the example above on how to place your "internal" functions
3) if you global variable is some kind of a configuration option, I'd just make public setter/getter and kept the variable "private"
http://nefariousdesigns.co.uk/archive/2010/10/object-oriented-javascript-follow-up-part-2-technical/
has a decent section on namespacing worth reading.
http://yuiblog.com/blog/2007/06/12/module-pattern/
is also another good overview.
For more great material on good javascript practices, check out
http://javascript.crockford.com/
After our discussion in the comments, I've changed the example to this:
var storageLocker = function (selector) {
var _selector = selector || "default value";
function myPrivateFunction() {
}
var public = {
get: function () {
return _selector;
},
uppercase : function () {
_selector = _selector.toUpperCase()
return this;
}
}
return public;
};
// use:
var test = storageLocker("search for this").uppercase().get();;
alert(test);
While this isn't exactly an example of a library/module (because you're accessing the code by calling the constructor directly), it is an example of keeping the same object in scope for further chaining of methods. Actually it's not returning the storageLocker object, it's returning the 'public' object, but that object has access to the storageLocker's scope through closure.
There could be other better ways to do this by perhaps returning the storageLocker object itself, but that would require a bit more thinking through.
In javascript, when would you want to use this:
(function(){
//Bunch of code...
})();
over this:
//Bunch of code...
It's all about variable scoping. Variables declared in the self executing function are, by default, only available to code within the self executing function. This allows code to be written without concern of how variables are named in other blocks of JavaScript code.
For example, as mentioned in a comment by Alexander:
(function() {
var foo = 3;
console.log(foo);
})();
console.log(foo);
This will first log 3 and then throw an error on the next console.log because foo is not defined.
Simplistic. So very normal looking, its almost comforting:
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
However, what if I include a really handy javascript library to my page that translates advanced characters into their base level representations?
Wait... what?
I mean, if someone types in a character with some kind of accent on it, but I only want 'English' characters A-Z in my program? Well... the Spanish 'ñ' and French 'é' characters can be translated into base characters of 'n' and 'e'.
So someone nice person has written a comprehensive character converter out there that I can include in my site... I include it.
One problem: it has a function in it called 'name' same as my function.
This is what's called a collision. We've got two functions declared in the same scope with the same name. We want to avoid this.
So we need to scope our code somehow.
The only way to scope code in javascript is to wrap it in a function:
function main() {
// We are now in our own sound-proofed room and the
// character-converter library's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
That might solve our problem. Everything is now enclosed and can only be accessed from within our opening and closing braces.
We have a function in a function... which is weird to look at, but totally legal.
Only one problem. Our code doesn't work.
Our userName variable is never echoed into the console!
We can solve this issue by adding a call to our function after our existing code block...
function main() {
// We are now in our own sound-proofed room and the
// character-converter libarary's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
main();
Or before!
main();
function main() {
// We are now in our own sound-proofed room and the
// character-converter libarary's name() function can exist at the
// same time as ours.
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
A secondary concern: What are the chances that the name 'main' hasn't been used yet? ...so very, very slim.
We need MORE scoping. And some way to automatically execute our main() function.
Now we come to auto-execution functions (or self-executing, self-running, whatever).
((){})();
The syntax is awkward as sin. However, it works.
When you wrap a function definition in parentheses, and include a parameter list (another set or parentheses!) it acts as a function call.
So lets look at our code again, with some self-executing syntax:
(function main() {
var userName = "Sean";
console.log(name());
function name() {
return userName;
}
}
)();
So, in most tutorials you read, you will now be bombarded with the term 'anonymous self-executing' or something similar.
After many years of professional development, I strongly urge you to name every function you write for debugging purposes.
When something goes wrong (and it will), you will be checking the backtrace in your browser. It is always easier to narrow your code issues when the entries in the stack trace have names!
Self-invocation (also known as
auto-invocation) is when a function
executes immediately upon its
definition. This is a core pattern and
serves as the foundation for many
other patterns of JavaScript
development.
I am a great fan :) of it because:
It keeps code to a minimum
It enforces separation of behavior from presentation
It provides a closure which prevents naming conflicts
Enormously – (Why you should say its good?)
It’s about defining and executing a function all at once.
You could have that self-executing function return a value and pass the function as a param to another function.
It’s good for encapsulation.
It’s also good for block scoping.
Yeah, you can enclose all your .js files in a self-executing function and can prevent global namespace pollution. ;)
More here.
Namespacing. JavaScript's scopes are function-level.
I can't believe none of the answers mention implied globals.
The (function(){})() construct does not protect against implied globals, which to me is the bigger concern, see http://yuiblog.com/blog/2006/06/01/global-domination/
Basically the function block makes sure all the dependent "global vars" you defined are confined to your program, it does not protect you against defining implicit globals. JSHint or the like can provide recommendations on how to defend against this behavior.
The more concise var App = {} syntax provides a similar level of protection, and may be wrapped in the function block when on 'public' pages. (see Ember.js or SproutCore for real world examples of libraries that use this construct)
As far as private properties go, they are kind of overrated unless you are creating a public framework or library, but if you need to implement them, Douglas Crockford has some good ideas.
I've read all answers, something very important is missing here, I'll KISS. There are 2 main reasons, why I need Self-Executing Anonymous Functions, or better said "Immediately-Invoked Function Expression (IIFE)":
Better namespace management (Avoiding Namespace Pollution -> JS Module)
Closures (Simulating Private Class Members, as known from OOP)
The first one has been explained very well. For the second one, please study following example:
var MyClosureObject = (function (){
var MyName = 'Michael Jackson RIP';
return {
getMyName: function () { return MyName;},
setMyName: function (name) { MyName = name}
}
}());
Attention 1: We are not assigning a function to MyClosureObject, further more the result of invoking that function. Be aware of () in the last line.
Attention 2: What do you additionally have to know about functions in Javascript is that the inner functions get access to the parameters and variables of the functions, they are defined within.
Let us try some experiments:
I can get MyName using getMyName and it works:
console.log(MyClosureObject.getMyName());
// Michael Jackson RIP
The following ingenuous approach would not work:
console.log(MyClosureObject.MyName);
// undefined
But I can set an another name and get the expected result:
MyClosureObject.setMyName('George Michael RIP');
console.log(MyClosureObject.getMyName());
// George Michael RIP
Edit: In the example above MyClosureObject is designed to be used without the newprefix, therefore by convention it should not be capitalized.
Scope isolation, maybe. So that the variables inside the function declaration don't pollute the outer namespace.
Of course, on half the JS implementations out there, they will anyway.
Is there a parameter and the "Bunch of code" returns a function?
var a = function(x) { return function() { document.write(x); } }(something);
Closure. The value of something gets used by the function assigned to a. something could have some varying value (for loop) and every time a has a new function.
Here's a solid example of how a self invoking anonymous function could be useful.
for( var i = 0; i < 10; i++ ) {
setTimeout(function(){
console.log(i)
})
}
Output: 10, 10, 10, 10, 10...
for( var i = 0; i < 10; i++ ) {
(function(num){
setTimeout(function(){
console.log(num)
})
})(i)
}
Output: 0, 1, 2, 3, 4...
Short answer is : to prevent pollution of the Global (or higher) scope.
IIFE (Immediately Invoked Function Expressions) is the best practice for writing scripts as plug-ins, add-ons, user scripts or whatever scripts are expected to work with other people's scripts. This ensures that any variable you define does not give undesired effects on other scripts.
This is the other way to write IIFE expression. I personally prefer this following method:
void function() {
console.log('boo!');
// expected output: "boo!"
}();
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/void
From the example above it is very clear that IIFE can also affect efficiency and performance, because the function that is expected to be run only once will be executed once and then dumped into the void for good. This means that function or method declaration does not remain in memory.
One difference is that the variables that you declare in the function are local, so they go away when you exit the function and they don't conflict with other variables in other or same code.
First you must visit MDN IIFE , Now some points about this
this is Immediately Invoked Function Expression. So when your javascript file invoked from HTML this function called immediately.
This prevents accessing variables within the IIFE idiom as well as polluting the global scope.
Self executing function are used to manage the scope of a Variable.
The scope of a variable is the region of your program in which it is defined.
A global variable has global scope; it is defined everywhere in your JavaScript code and can be accessed from anywhere within the script, even in your functions. On the other hand, variables declared within a function are defined only within the body of the function.
They are local variables, have local scope and can only be accessed within that function. Function parameters also count as local variables and are defined only within the body of the function.
As shown below, you can access the global variables variable inside your function and also note that within the body of a function, a local variable takes precedence over a global variable with the same name.
var globalvar = "globalvar"; // this var can be accessed anywhere within the script
function scope() {
alert(globalvar);
var localvar = "localvar"; //can only be accessed within the function scope
}
scope();
So basically a self executing function allows code to be written without concern of how variables are named in other blocks of javascript code.
Since functions in Javascript are first-class object, by defining it that way, it effectively defines a "class" much like C++ or C#.
That function can define local variables, and have functions within it. The internal functions (effectively instance methods) will have access to the local variables (effectively instance variables), but they will be isolated from the rest of the script.
Self invoked function in javascript:
A self-invoking expression is invoked (started) automatically, without being called. A self-invoking expression is invoked right after its created. This is basically used for avoiding naming conflict as well as for achieving encapsulation. The variables or declared objects are not accessible outside this function. For avoiding the problems of minimization(filename.min) always use self executed function.
(function(){
var foo = {
name: 'bob'
};
console.log(foo.name); // bob
})();
console.log(foo.name); // Reference error
Actually, the above function will be treated as function expression without a name.
The main purpose of wrapping a function with close and open parenthesis is to avoid polluting the global space.
The variables and functions inside the function expression became private (i.e) they will not be available outside of the function.
Given your simple question: "In javascript, when would you want to use this:..."
I like #ken_browning and #sean_holding's answers, but here's another use-case that I don't see mentioned:
let red_tree = new Node(10);
(async function () {
for (let i = 0; i < 1000; i++) {
await red_tree.insert(i);
}
})();
console.log('----->red_tree.printInOrder():', red_tree.printInOrder());
where Node.insert is some asynchronous action.
I can't just call await without the async keyword at the declaration of my function, and i don't need a named function for later use, but need to await that insert call or i need some other richer features (who knows?).
It looks like this question has been answered all ready, but I'll post my input anyway.
I know when I like to use self-executing functions.
var myObject = {
childObject: new function(){
// bunch of code
},
objVar1: <value>,
objVar2: <value>
}
The function allows me to use some extra code to define the childObjects attributes and properties for cleaner code, such as setting commonly used variables or executing mathematic equations; Oh! or error checking. as opposed to being limited to nested object instantiation syntax of...
object: {
childObject: {
childObject: {<value>, <value>, <value>}
},
objVar1: <value>,
objVar2: <value>
}
Coding in general has a lot of obscure ways of doing a lot of the same things, making you wonder, "Why bother?" But new situations keep popping up where you can no longer rely on basic/core principals alone.
You can use this function to return values :
var Test = (function (){
const alternative = function(){ return 'Error Get Function '},
methods = {
GetName: alternative,
GetAge:alternative
}
// If the condition is not met, the default text will be returned
// replace to 55 < 44
if( 55 > 44){
// Function one
methods.GetName = function (name) {
return name;
};
// Function Two
methods.GetAge = function (age) {
return age;
};
}
return methods;
}());
// Call
console.log( Test.GetName("Yehia") );
console.log( Test.GetAge(66) );
Use of this methodology is for closures. Read this link for more about closures.
IIRC it allows you to create private properties and methods.