Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:6031
Return-Path: <php@hristov.com>
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Delivered-To: mailing list internals@lists.php.net
Received: (qmail 13377 invoked by uid 1010); 3 Dec 2003 00:28:50 -0000
Delivered-To: ezmlm-scan-internals@lists.php.net
Delivered-To: ezmlm-internals@lists.php.net
Received: (qmail 13353 invoked from network); 3 Dec 2003 00:28:50 -0000
Received: from unknown (HELO stella.fastforwardnetwork.com) (209.51.153.50)
  by pb1.pair.com with SMTP; 3 Dec 2003 00:28:50 -0000
Received: (qmail 8043 invoked from network); 3 Dec 2003 01:31:19 -0000
Received: from p508ab217.dip.t-dialin.net (HELO hristov.com) (80.138.178.23)
  by d.goto.bg with SMTP; 3 Dec 2003 01:31:19 -0000
Message-ID: <3FCD2E06.6090809@hristov.com>
Date: Wed, 03 Dec 2003 01:27:50 +0100
User-Agent: Mozilla/5.0 (X11; U; Linux i586; en-US; rv:1.6b) Gecko/20031202
X-Accept-Language: en-us, en
MIME-Version: 1.0
To:  internals@lists.php.net
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: BC broken for strrpos()/strripos()
From: php@hristov.com (Andrey Hristov)

Hi devs,
this night i saw that the BC of strrpos()/strripos()
is broken by a patch commited by pollita 7 months ago :
http://cvs.php.net/diff.php/php-src/ext/standard/string.c?login=2&r1=1.370&r2=1.371&ty=u
The documenation of both function states if the second parameter is an
integer instead of string then its value is used as a ord of the character
to be used during the search. This no more true with HEAD.
A script to reproduce :
<?php
var_dump(strrpos("ABCDEF","DEF"));
var_dump(strrpos("ABCDEF","DAF"));
var_dump(strrpos("ABCDEF",ord("D")));
?>
HEAD :
int(3)
bool(false)
bool(false)
PHP 4.3.5-dev (cli) (built: Dec  2 2003 00:13:37) (DEBUG):
int(3)
int(3)
int(3)
What do you think? Is it ok to break the BC here or not? As you see the 
other change introduced with the patch is to use the whole needle string 
while searching, not only the first character.

Comments are welcome,
Andrey