Question

Usually when you do something like 'test'.match(/(e)/) you would receive an array ['e', 'e'], where the first element is the match itself and the second from the selector (braces), but when using the global modifier as in 'test'.match(/(e)/g) it will omit the match, while it doesn’t in case I don’t use selectors at all.

I wonder if and where the following behavior is specified (using Chromium for this test).

Answer 1

If the global flag (g) is not set, Element zero of the array contains the entire match, while elements 1 through n contain any submatches. This behavior is the same as the behavior of the exec Method (Regular Expression) (JavaScript) when the global flag is not set. If the global flag is set, elements 0 through n contain all matches that occurred.

http://msdn.microsoft.com/en-us/library/ie/7df7sf4x(v=vs.94).aspx

In other words, when g is provided, match collects only topmost matches, ignoring any capturing groups.

Example:

> s = "Foo Bar"
"Foo Bar"
> s.match(/([A-Z])([a-z]+)/)
["Foo", "F", "oo"]
> s.match(/([A-Z])([a-z]+)/g)
["Foo", "Bar"]

There’s no built-in that would collect all groups from all matches, like python findall does, but it’s easy to write using exec:

function matchAll(re, str) {
    var p, r = [];
    while(p = re.exec(str))
        r.push(p);
    return r;
}
matchAll(/([A-Z])([a-z]+)/g, "Foo Bar")

Result:

[
Array[3]
0: "Foo"
1: "F"
2: "oo"
index: 0
input: "Foo Bar"
length: 3
__proto__: Array[0]
, 
Array[3]
0: "Bar"
1: "B"
2: "ar"
index: 4
input: "Foo Bar"
length: 3
__proto__: Array[0]
]