How is a JavaScript hash map implemented?

Question

I currently work with OpenLayers and have a huge set of data to draw into a vector layer (greater than 100000 vectors).

I'm now trying to put all these vectors into a JavaScript hash map to analyze the performance. I want to know how is the hash map in JavaScript implemented, is it a real hash function or just a wrapped function that uses a simple data structure and a search algorithm?

Answer 1

every javascript object is a simple hashmap which accepts a string or a Symbol as its key, so you could write your code as:

var map = {}; // add a item map[key1] = value1; // or remove it delete map[key1]; // or determine whether a key exists key1 in map; 

javascript object is a real hashmap on its implementation, so the complexity on search is O(1), but there is no dedicated hashcode() function for javascript strings, it is implemented internally by javascript engine (V8, SpiderMonkey, JScript.dll, etc...)

2020 Update:

javascript today supports other datatypes as well: Map and WeakMap. They behave more closely as hash maps than traditional objects.

Answer 2

JavaScript objects cannot be implemented purely on top of hash maps.

Try this in your browser console:

var foo = {     a: true,     b: true,     z: true,     c: true }  for (var i in foo) {     console.log(i); } 

...and you'll recieve them back in insertion order, which is de facto standard behaviour.

Hash maps inherently do not maintain ordering, so JavaScript implementations may use hash maps somehow, but if they do, it'll require at least a separate index and some extra book-keeping for insertions.

Here's a video of Lars Bak explaining why v8 doesn't use hash maps to implement objects.