Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:53450
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain users.sourceforge.net from 77.240.252.9 cause and error)
Message-ID: <39875.5975f3c3.1308604731.nsm@avilys.eik.lt>
In-Reply-To: <BANLkTi=oC_FQ=o0caf5pWu8YPa33kS4SRA@mail.gmail.com>
References: <BANLkTinH6-ioQ8WB9Vd8f50HdwendCZDYQ@mail.gmail.com>
    <alpine.DEB.2.02.1106201304590.3774@whisky>
    <BANLkTi=z_-3254MijvDRQdvoqcPt=wQwWw@mail.gmail.com>
    <1308584208.6296.9.camel@guybrush>
    <BANLkTimQTa+3+uUB7xcR_c66myiBs0yOWA@mail.gmail.com>
    <1308586150.6296.13.camel@guybrush>
    <BANLkTimOsO+v4SnjKHVV7Lf126PgpoTbfQ@mail.gmail.com>
    <1308589044.8394.27.camel@inspiron> <4DFF7E2A.50506@sugarcrm.com>
    <1308591260.8394.47.camel@inspiron>
    <BANLkTi=oC_FQ=o0caf5pWu8YPa33kS4SRA@mail.gmail.com>
Date: Tue, 21 Jun 2011 00:18:51 +0300 (EEST)
To: internals@lists.php.net
User-Agent: NaSMail/1.7.1
MIME-Version: 1.0
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: 8bit
Importance: Normal
Subject: =?utf-8?Q?Re:_[PHP-DEV]_Re:_foreach=28=29_for_strings?=
From: tokul@users.sourceforge.net ("Tomas Kuliavas")

2011.06.20 21:38 Robert Eisele rašė:
> I really like the ideas shared here. It's a thing of consideration that
> array-functions should also work with strings. Maybe this would be the way
> to go, but I'm more excited about the OOP implementation of TextIterator
> and
> ByteIterator, which solves the whole problem at once (and is easier to
> implement, as mentioned by Stas). As Jonathan  said, Database results with
> a
> certain encoding could get iterated, too. The only way to workaround the
> Text/Byte problem would be, offsetting >EVERY< string with 1-2 byte
> "string-type" information or an additional type flag in the
> zval-strcuture.
> Handling everything with zval's instead of objects would have the
> advantage,
> that database-layers like mysqlnd could write the database-encoding
> directly
> into the zval and the user had no need to decide what encoding is used.
>
> A new casting operator (binary) could then cast the string to a 1-byte
> array. But this is syntactical sugar over OOP-implementations - I don't
> know
> which one is the better choice.
>
> For example:
>
> $utf8_string = "Jägermeister"; // information of utf8 ist stored in the
> zval
>
> foreach ($utf8_string as $k => $v) // would iterate in byte mode
>
> foreach ((binary)$utf8_string as $k => $v) // would iterate in text mode
>
> over this:
> $utf8_obj = new ByteIterator("Jägermeister");
>
> foreach ($utf8_obj as $k => $v)
>
> foreach ($utf8_obj->toText() as $k => $v)
>
>
> I think the first one is easier and would be nicer to average developers
> (and lazy programmers like me ;o) )

You assume that string is in utf-8. It can be some iso-8859-x,
iso-2022-xx, utf-7, utf-16 or any other multibyte character set.

If you want to parse string in symbols, use mb_substr and mb_strlen, set
charset correctly and make sure that your string is in correct character
set, if PHP bug about inconsistent symbol position calculation is still
unfixed.

-- 
Tomas