Using stripslashes within preg_replace with the “e” modifier

First published on June 11, 2011

Recently, I had to do some formatting on existing HTML code, to transform nested paragraphs within blockquotes into individual blockquotes, for XML processing. For example, turning this:


... lots of different HTML before this
<blockquote><p>This is a quote with a <a href="http://www.theblog.ca" title="Peter's Useful Crap">link to a fun site</a></p><p>You should come back often</p></blockquote>
... lots of different HTML after this


Into this:

... lots of different HTML before this
<blockquote>This is a quote with a <a href="http://www.theblog.ca" title="Peter's Useful Crap">link to a fun site</a></blockquote><blockquote>You should come back often</blockquote>
... lots of different HTML after this


This is not a particularly exciting or common task, but it comes with a simple tip. As you will see, below, my solution was to use the “e” directive on your preg_replace function so that you can run code on the match only. In my case, I had a simple str_replace:


<?php

function stripParagraphsFromBlockquotes( $text )
{
    $find = '/<blockquote><p>(.*?)<\/p><\/blockquote>/ise';
    $replace = "'<blockquote>' . stripslashes( str_replace( '</p><p>', '</blockquote>\n<blockquote>', '\\1' ) ) . \"</blockquote>\n\"";
    $text = preg_replace( $find, $replace, $text );
    return $text;
?>


At first, however, I did not have the outer stripslashes call, and was ending up with this output, which breaks XML parsers:


<blockquote>This is a quote with a <a href=\"http://www.theblog.ca\" title=\"Peter's Useful Crap\">link to a great fun</a></blockquote><blockquote>You should come back often</blockquote>


The tip is to use stripslashes on your inner “replace” code, since the “e” modifier automatically adds slashes on the preg_replace match ‘\\1′.

Facebook Twitter Email this

This entry is filed under Web development tutorials and written by Peter Keung. You can follow any responses to this entry through the RSS 2.0 feed.