[PHP-DOC] Re: array_change_key_case()? key_exists()? From: Jeroen van Wolffelaar (jeroen <email protected>)
Date: 11/11/01

On Sun, 11 Nov 2001, Derick Rethans wrote:

> On Sun, 11 Nov 2001, Egon Schmid wrote:
>
> > Please revert that patch from Jeroen. I have something about
> > undocumented functions.
>
> done with pleasure.

The reason for (temporarily at least) commenting them was to prevent
users from learning from that function without a proper discussion
wether they really belong in PHP. Once they are documented, there's just
another BC problem created. So please do not revert it until properly
discussed. I didn't have the chance to respond to various arguments yet.

Egon and Goba: this is just while the discussion is going on, if it's
decided that they should stay as it is now, you can uncomment it again
tomorrow - what does a day make on a function not being documented for
months?

Okay, to the discussion itself:

(array_)key_exists: As Andi said, there is a reason why isset behaves
like it behaves now. IMO, I see two possibilities:
(a): if that reason isn't valid anymore, let's fix isset in stead of
adding another function
(b): if it IS valid, then why would key_exists be useful and good, IMO
key_exists should follow isset(), rendering it superflueous.

array_change_key_case:

Well... there be performance issues, but no matter what, a more
general-use function should be created - if one. For example a
array_walk like function which applies a random function to each key.
(doesn't array_walk already support this?)

So I'm still quite against array_change_key_case, for array_key_exists
there might be good arguments in favor, I didn't yet read them, but
even.

I recommented those functions just to prevent a "It's in the manual now,
you cannot revert it, because of BC" closing of this discussion which is
definitely something undesireable.

--Jeroen

Jeroen van Wolffelaar
Jeroen <email protected>
http://www.A-Eskwadraat.nl/~jeroen