Good morning, afternoon, evening, night. I have some things to share with you about the new keyword in Javascript. Important things.
I'll start w...
For further actions, you may consider blocking this person and/or reporting abuse
In the edge case you should return
this
Okay, changed it. Is it correct now?
Well, now I think you do not completely understand it yourself.
In the new2 function you apply the constructor to newObject, which makes 'this' inside the constructor to reference newObject, therefore after modifying 'this' in the constructor it should return 'this'.
After that new2 can just return the result of 'constructor.apply' which is the reference to newObject.
P.S. Cannot really edit with markdown from a mobile.
You are correct -- I'm certainly having difficulty with understanding how to handle the edge case(s). Always learning! Thanks for taking the time to discuss it with me.
So, it sounds like my initial implementation of
new2
was correct to returnconstructor.apply(newObject, constructorArgs) || newObject
, but you are saying the Constructor function itself (in the edge case) should returnthis
rather than the object it is currently returning? I believe that would actually be a new edge case (in addition to the one presented above).I tested my implementation in a REPL and it is handling the edge case of a Constructor returning an object correctly.
Would you mind providing a code snippet or copying and correcting the code in the article to convey your insight?
Thanks again for contributing!
If
constructor
returnsit is a completely different from
newObject
object reference and after that ifnew2
returns the result ofconstructor
thenewObject
would be just lost in thenew2
scope.So it have to be like
Ah okay, I see what you're saying now. I guess part of what I was trying to convey is that the constructor function can in fact return any object. In which case, the new function should just return that object as is.
What you've presented here is a more realistic edge case. I will think about a way to cover it in the article.
Great insights -- thank you.
Nice catch Elias -- thank you for bringing it up. I'll fix it tonight!
Fantastic article. I was just wondering about this today.
Nice article overall but the most interesting bit..."give a couple arguments for avoiding this approach(...)"
... has never happened! :-)
You're right! The article ended up getting quite long, so I decided to leave that part for a follow-up. I guess it's a cliffhanger for now ;)
Great article. I put this link in our news letter so other can find it as well: betterdev.link/issues/6