String.replace(…) test

Support for String.replace('…','…')

var foo = 'It\'s the end of the world as we know it, and I feel fine.';
var result = foo.replace('end','BEGINNING');

Expected result: "It's the BEGINNING of the world as we know it, and I feel fine."

Outcome: it's basic, and it works in just about every browser.

Support for String.replace(/…/,'…')

var vowels = /[aeiou]/gi;
result = foo.replace(vowels,'-');

Expected result: "-t's th- -nd -f th- w-rld -s w- kn-w -t, -nd - f--l f-n-."

Outcome: also well supported by browsers.

Support for String.replace(/…/,myFunction)

function Capitalize(s){
	return s.toUpperCase();
}
result = foo.replace(vowels,Capitalize);

Expected result: "It's thE End Of thE wOrld As wE knOw It, And I fEEl fInE."

Outcome: IEMac (5.2.2) and Safari (1.0b2) do not support supplying a function for the second parameter. Instead .toString() is applied to the function, resulting in the raw code of the function being used in the replacement.

Support for String.replace(/…/,myFunction), using RegExp

function Capitalize(){
	return RegExp.$1.toUpperCase()+RegExp.$2;
}
result = foo.replace(/([aeiou])([a-z])/g,Capitalize);

Expected result: "It's the End Of the wOrld As we knOw It, And I fEel fIne."

Outcome: In addition to the IEMac's and Safari's problem mentioned above, IEWin does not seem to populate the RegExp object when a regular expression is used inside .replace(). IEWin has major problems with this. Mozilla alone gets all four tests correct.

Support for String.replace(/…/,myFunction), using parameters

function Capitalize(orig,re1,re2){
	return re1.toUpperCase()+re2;
}
result = foo.replace(/([aeiou])([a-z])/g,Capitalize);

Expected result: "It's the End Of the wOrld As we knOw It, And I fEel fIne."

Outcome: IEWin gets this one right. This is the proper method that works in all browsers which support using a function as the second parameter. (Plus, it's faster than attempting to use RegExp.$x as above.)