Javascript replace null

Question

var p = "null"
var q = null;
(p == q) //false. as Expected.

p.replace(null, "replaced") // outputs replaced. Not expected.
p.replace("null", "replaced") //outputs replaced. Expected.

q.replace(null, "replaced") // error.  Expected.
q.replace("null", "replaced") //error. Expected.

Why? Does replace not differentiate between "null" and null?

I ask because I ran into a bug in angularjs:

replace((pctEncodeSpaces ? null : /%20/g), '+');

If for example, someone had a username of "null" and it was used as url, it would be replaced with "+" on any $http calls. e.g. GET /user/null.

Not that this scenario would occur often, but I'm more curious why replace treats null and "null" as the same thing. Does replace do a .tostring on null before it does the replacement? Is this just a quirk of Javascript?

I verified this on both IE and Chrome's implementations of replace.

Answer 1

Yes, this is expected according to the spec for replace (bolded relevant line, or page 146 of the ECMA-262 final draft). The first argument is checked to see if it is a regex and if not, it has toString() called on it (well, converted to a string somehow).

15.5.4.11 String.prototype.replace(searchValue, replaceValue)

Let string denote the result of converting the this value to a string.

Cut for brevity

IfsearchValue is not a regular expression, let searchString be ToString(searchValue) and search string for the first occurrence of searchString. Let m be 0.

Cut for brevity

Answer 2

In the ES5 specification for String.prototype.replace:

15.5.4.11 String.prototype.replace (searchValue, replaceValue)

...

If searchValue is not a regular expression, let searchString be ToString(searchValue) and search string for the first occurrence of searchString

So, "".replace(null, XXX) will indeed convert the null to the string "null"

Note that ToString() does not mean null.toString() - it's an internal defined operation within the JavaScript interpreter.