Skip to content

Instantly share code, notes, and snippets.

@micahjon
Last active December 16, 2020 01:46
Show Gist options
  • Save micahjon/e47aff47bfa093091563 to your computer and use it in GitHub Desktop.
Save micahjon/e47aff47bfa093091563 to your computer and use it in GitHub Desktop.
<?php
/**
* Move image inside <p> tag above the <p> tag while preserving any link around image.
* Can be prevented by adding any attribute or whitespace to <p> tag, e.g. <p class="yolo"> or even <p >
*/
function gc_remove_p_tags_around_images($content)
{
$contentWithFixedPTags = preg_replace_callback('/<p>((?:.(?!p>))*?)(<a[^>]*>)?\s*(<img[^>]+>)(<\/a>)?(.*?)<\/p>/is', function($matches)
{
/*
Groups Regex Description
<p> starting <p> tag
1 ((?:.(?!p>))*?) match 0 or more of anything not followed by p>
.(?!p>) anything that's not followed by p>
?: non-capturing group.
*? match the ". modified by p> condition" expression non-greedily
2 (<a[^>]*>)? starting <a> tag (optional)
\s* white space (optional)
3 (<img[^>]+>) <img> tag
\s* white space (optional)
4 (<\/a>)? ending </a> tag (optional)
5 (.*?)<\/p> everything up to the final </p>
i modifier case insensitive
s modifier allows . to match multiple lines (important for 1st and 5th group)
*/
// image and (optional) link: <a ...><img ...></a>
$image = $matches[2] . $matches[3] . $matches[4];
// content before and after image. wrap in <p> unless it's empty
$content = trim( $matches[1] . $matches[5] );
if ($content) {
$content = '<p>'. $content .'</p>';
}
return $image . $content;
}, $content);
// On large strings, this regular expression fails to execute, returning NULL
return is_null($contentWithFixedPTags) ? $content : $contentWithFixedPTags;
}
add_filter('the_content', 'gc_remove_p_tags_around_images');
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment