Using preg_replace to filter custom textarea

preg_replace expects it’s regex to be surrounded by a character. Usually that’s a slash, like this:

'/.*/'

When you do this:

preg_replace('%%PROPOSAL_LINK_URL%%', $some_text, $replacement);

preg_replace thinks the first two % are your surrounding characters, and fails because of an unrecognized modifier. This is easy to test:

$ php -a
Interactive shell

php > echo preg_replace('%%PROPOSAL_LINK_URL%%', '%%PROPOSAL_LINK_URL%%', 'here'), PHP_EOL;
PHP Warning:  preg_replace(): Unknown modifier 'P' in php shell code on line 1
PHP Stack trace:
PHP   1. {main}() php shell code:0
PHP   2. preg_replace() php shell code:1

Warning: preg_replace(): Unknown modifier 'P' in php shell code on line 1

Call Stack:
   28.9460     235264   1. {main}() php shell code:0
   28.9460     236016   2. preg_replace() php shell code:1


php >

You have two options:

Change your patterns to something like this: /%%PROPOSAL_LINK_URL%%/ and /%%PROPOSAL_LINK%%/

$ php -a
Interactive shell

php > echo preg_replace('/%%PROPOSAL_LINK_URL%%/', '%%PROPOSAL_LINK_URL%%', 'here'), PHP_EOL;
here

Forget all about preg_replace because you don’t need it. Use str_replace or str_ireplace instead.

<?php
str_replace('%%PROPOSAL_LINK_URL%%', $proposal_link_url, $email_content);
str_ireplace('%%PROPOSAL_LINK%%', $proposal_link, $email_content);

Be aware that str_replace and it’s brethren are not safe for use on strings that contain multi-byte characters (eg. UTF-8) by default. You can change this.

NOTE: you’ve used $link_pattern twice in your code examples. I think you mean the second one to be $url_pattern.

EDIT: it appears there’s not a mb_* replacement for str_replace or str_ireplace this comment on php.net sheds some light on the issue. str_replace should work on multibyte strings, providing the encoding of the needle and haystack are the same.

Related Posts: