Re: [RFC] ArrayIterator improvements

Hi, Given that this new method seems closely related to ArrayIterator::seek, I (as a userland developer) would very much expect it to handle error conditions in the same way. In the case of ArrayIterator::seek - it throws an OutOfBoundsException. I don’t see how that isn’t appropriate for seekKey? Cheers Stephen > On 23 Nov 2016, at 04:56, Wes <netmo.php@gmail.com> wrote: > > I have decided to go with that because > > 1- I'm a fan of using Exceptions (mostly) for exceptional error conditions; > between these two > public seekKey($key):void throws WhateverException; > public seekKey($key):bool; > I (by far) prefer the latter. > > 2- I think it would require a new type of SPL Exception, because none of > the existing makes sense to use here, in my opinion. > > 2016-11-22 21:43 GMT+01:00 Nikita Popov <nikita.ppv@gmail.com>: > >> On Tue, Nov 22, 2016 at 6:47 PM, Wes <netmo.php@gmail.com> wrote: >> >>> Greetings again PHPeople, >>> >>> I wanted to avoid the discussion for the small improvements I was >>> proposing >>> thinking it would be acceptable to do so ( >>> http://news.php.net/php.internals/97118) but apparently >>> I was wrong >>> (sorry >>> for that), so here's the discussion thread! >>> >>> https://wiki.php.net/rfc/arrayiterator-improvements >>> >>> Again this is my first RFC, and I hope I'm doing nothing wrong this time >>> :P >>> >>> Thanks again, >>> Wes >>> >>> (and special thanks to Room11 for their feedback about the RFC process) >>> >> >> What's the reason for making seekKey() return a boolean? The existing >> seek() method throws an exception if the seek offset does not exist. It >> would make sense to me for seekKey() to throw an exception if the key does >> not exist, to keep things consistent. >> >> Nikita >> >>