Re: [whatwg/url] URLSearchParams: calling .set() and .append() should return itself (#90)

> Another thing I was thinking, for a multimap you might want append() and such to return something indicating whether the multimap already contains the key being appended.

That seems to duplicate what `.has()` and `.contains()` is for - or possibly even `.getAll(item).length`. 

>  It's not clear to me what utility returning this offers and the Twitter poll doesn't really address that point either.

It's what @jussi-kalliokoski said on Twitter: "consistency and least surprise". The current design violates both, because: 

 1.  Consistency: it's inconsistent with the language primitives (map and set).
 2. Least surprise: it's surprising that it doesn't return self, particularly when performing a `reduce()` as per the code below (which goes to argument of utility).  

Consider again what we have today: 

```JS
    .reduce((params, url) => {
       params.append("foo", url);
       return params;
    }, params)
```

So, the utility in returning self is that `.reduce()` expects the "collector" to be returned. Thus, returning `this` would be idiomatic because it would work like regular maps and sets:

```JS
  // Look ma! no return keyword and no "{" or "}"  
  .reduce((params, url) => params.append("foo", url), new URLSearchParams())
```

It's extremely common to want to `.map()` a list of things, and then `.reduce()` those things into a collection type, like `URLSearchParams`, `DOMTokenList`, `Headers`, etc. but it's *surprising and inconsistent* that doing so don't follow ES. 
  


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/url/issues/90#issuecomment-268384396