Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:21785
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Message-ID: <43E333FA.1030309@cschneid.com>
Date: Fri, 03 Feb 2006 11:44:10 +0100
User-Agent: Thunderbird 1.5 (Macintosh/20051201)
MIME-Version: 1.0
To: Marcus Boerger <helly@php.net>
Cc: Andrei Zmievski <andrei@gravitonic.com>,
	PHP Developers Mailing List <internals@lists.php.net>
References: <f245197d739464c7d81bec6836e192ee@gravitonic.com> <43E2AF9F.60607@cschneid.com> <134017377.20060203091233@marcus-boerger.de>
In-Reply-To: <134017377.20060203091233@marcus-boerger.de>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [PHP-DEV] Re: Unicode string iterator performance
From: cschneid@cschneid.com (Christian Schneider)

First of all I was simply proposing a very generic concept without 
bothering about the implementation on purpose. If it's not feasible then 
simply ignore it.

Marcus Boerger wrote:
>   caching? There is nothing to cache. And even if we would do that we would
> make every string an object since we would need to invalidate the position
> cache on write operations. Also i agree with the others that most common

Tracking changes to the string could be tricky, agreed. I don't know 
enough about the internal handling of strings, from the user perspective 
PHP strings look somewhat immutable but that could be very wrong 
internally, you know better than me. Changing Unicode strings in place 
sounds kinda tricky to me too so I'd have expected that to the 
encapsulated somewhere.

> And *I* never had code where I used the same position twice. Besides the all

You don't need to access the exact same position. If you know the last 
array index plus the Unicode offset then you can step by Unicode 
characters from there which would result to one single Unicode step for 
iterating over a string. But would also work for $a[$i += 2] as opposed 
to the originally proposed TextIterator. And if there's a way to step 
backwards then $a[$i -= 2] could work too.

> So i am convinced that the cache would only blow up the code, make everything
> much more complex and in the end slow down php.

Could well be. It was just an idea, feel free to ignore it ;-)

- Chris