Is it possible to exclude parts of the matched string in preg_match? - php

Question

when writing a script that is supposed to download content from a specific div I was wondering if it is possible to skip some part of the pattern in such a way that it will not be included in the matching result.
examlple:
<?php
$html = '
<div class="items">
<div class="item-s-1827">
content 1
</div>
<div class="item-s-1827">
content 2
</div>
<div class="item-s-1827">
content 3
</div>
</div>
';
preg_match_all('/<div class=\"item-s-([0-9]*?)\">([^`]*?)<\/div>/', $html, $match);
print_r($match);
/*
Array
(
[0] => Array
(
[0] => <div class="item-s-1827">
content 1
</div>
[1] => <div class="item-s-1827">
content 2
</div>
[2] => <div class="item-s-1827">
content 3
</div>
)
[1] => Array
(
[0] => 1827
[1] => 1827
[2] => 1827
)
[2] => Array
(
[0] =>
content 1
[1] =>
content 2
[2] =>
content 3
) ) */
Is it possible to omit class=\"item-s-([0-9]*?)\" In such a way that the result is not displayed in the $match variable?

Answer 1

In general, you can assert strings precede or follow your search string with positive lookbehinds / positive lookaheads. In the case of a lookbehind, the pattern must be of a fixed length which stands in conflict with your requirements. But fortunately there's a powerful alternative to that: You can make use of \K (keep text out of regex), see http://php.net/manual/en/regexp.reference.escape.php:
\K can be used to reset the match start since PHP 5.2.4. For example, the patter foo\Kbar matches "foobar", but reports that it has matched "bar". The use of \K does not interfere with the setting of captured substrings. For example, when the pattern (foo)\Kbar matches "foobar", the first substring is still set to "foo".
So here's the regex (I made some additional changes to that), with \K and a positive lookahead:
preg_match_all('/<div class="item-s-[0-9]+">\s*\K[^<]*?(?=\s*<\/div>)/', $html, $match);
print_r($match);
prints
Array
(
[0] => Array
(
[0] => content 1
[1] => content 2
[2] => content 3
)
)

Answer 2

The preferred way to parse HTML in PHP is to use DomDocument to load the HTML and then DomXPath to search the result object.
Update
Modified based on comments to question so that <div> class names just have to begin with item-s-.
$html = '<div class="items">
<div class="item-s-1827">
content 1
</div>
<div class="item-s-18364">
content 2
</div>
<div class="item-s-1827">
content 3
</div>
</div>';
$doc = new DomDocument();
$doc->loadHTML($html);
$xpath = new DomXPath($doc);
$divs = $xpath->query("//div[starts-with(#class,'item-s-')]");
foreach ($divs as $div) {
$values[] = trim($div->nodeValue);
}
print_r($values);
Output:
Array (
[0] => content 1
[1] => content 2
[2] => content 3
)
Demo on 3v4l.org