The most common gotcha I encounter in Javascript has to do with callbacks. I see this problem all the time in my own code and that of my colleagues: callbacks that have been registered either more or less times than the author of the code expected. Usually more times.
This can manifest itself in many ways. Usually in the form of a UI element being displayed more than once, but often more subtly such as a HTTP request being sent more than once. Often the bugs only manifest themselves after a certain sequence of actions, for example: click this UI element, check this radio box, then hit submit and watch the net tab.
It's so often the root cause of Javascript bugs that if I or somebody else on my team has a Javascript problem, I often start by asking myself if it looks like it could be this. Lately I have begun to wonder if there isn't some slight truth to that inflammatory "Callbacks are the new goto" story that was on here a couple of months ago [1].
I see this problem all the time in my own code and that of my colleagues: callbacks that have been registered either more or less times than the author of the code expected. Usually more times.
I've yet to run into this problem before, and I'm an even split between JS / Ruby these days. Are you building on top of any particular libraries or frameworks? Could you give some examples of where you've encountered this?
I've encountered it once with a form, although it was a while ago and if I saw the code again I might be able to identify what was wrong. In that case it was probably a poor choice of element to attach to that caused the callback to be fired independently multiple times on one click, and stopping the bubbling had no effect.
The other time was in a Rails app, where a script was added to a partial for a layout, and then added again later on (by a different developer) to the layout itself. That caused a particularly nasty bug since it used 'toggle', and thus the second callback negated the first.
Multiple registration of callbacks and "too many similar events firing" (and even circular event firing : event A triggers event B, event B triggers event A) can all be solved by using event filtering no!?
I'm not saying it's a good thing to work around the issue by using event filtering, but events filtering is a great technique.
You can also use your event filtering to "fail fast" during development: detect a duplicated event called back? Throw an exception / make your application fail.
a basic explanation is that by preceding a variable declaration with var creates a property on the nearest containing function
This is not true, the var hoist the variable and scopes it to the function, but does not create it as a property of the function. A function is defined in its outer function scope, an it's properties will persist across calls.
function foo() {
var bar = 1;
foo.baz = 2;
}
foo();
foo.bar; // is undefined
foo.baz; // is 2
Adding properties to functions is useful for memoization, but is different than what the var statement does.
Your summary of that section is a better explanation:
all variables are scoped to a function (which is itself an object), and where you declare those variables with var determines the function they are scoped to.
You might add, if you never declare the variable with var it is implicitly declared global.
Using closures to encapsulate private variables is a good idea, but the way this article recommends doing it doesn't demonstrate any of the benefit as it doesn't really close over anything. I don't know if it's an example of good JS code, but something like the basic closure definition of a counter might have better demonstrated the ability to hide state:
var counter = (function () {
var i = 0;
return function () {
i += 1;
return i;
}
}());
Where the internal variable, i, is inaccessible to anything but the counter function itself.
author here, it wasn't about hiding private variables. i was demonstrating using closures to avoid global namespace collisions which i think is generally a big problem for beginners.
i believe you mean using functions, not closures. your post doesn't seem to have the same misnomer, and simply explains lexical function scope. it doesn't seem to touch on the closure, while your parent comment does.
not exactly. closures close over scope. functions alone simply create a new scope. these concepts are not synonymous.
if your function doesn't return another function, the references from that function are deallocated once the function has executed.
when your function returns a function (a "closure"), the inner function closes over the variables of the outer function, maintaining references to it's context as long as you maintain a reference to the closure.
This can manifest itself in many ways. Usually in the form of a UI element being displayed more than once, but often more subtly such as a HTTP request being sent more than once. Often the bugs only manifest themselves after a certain sequence of actions, for example: click this UI element, check this radio box, then hit submit and watch the net tab.
It's so often the root cause of Javascript bugs that if I or somebody else on my team has a Javascript problem, I often start by asking myself if it looks like it could be this. Lately I have begun to wonder if there isn't some slight truth to that inflammatory "Callbacks are the new goto" story that was on here a couple of months ago [1].
[1] http://elm-lang.org/learn/Escape-from-Callback-Hell.elm