From e1052e2eb37c2899d9876690d08e7ac094f6fe35 Mon Sep 17 00:00:00 2001
From: Jon McPartland <jon@bigbitecreative.com>
Date: Thu, 25 Oct 2018 14:30:47 +0100
Subject: [PATCH] Add documentation to hashtag regex

---
 ...liveblog-entry-extend-feature-hashtags.php | 27 +++++++++++++++++--
 1 file changed, 25 insertions(+), 2 deletions(-)

diff --git a/classes/class-wpcom-liveblog-entry-extend-feature-hashtags.php b/classes/class-wpcom-liveblog-entry-extend-feature-hashtags.php
index b1baffccb..f9a051207 100644
--- a/classes/class-wpcom-liveblog-entry-extend-feature-hashtags.php
+++ b/classes/class-wpcom-liveblog-entry-extend-feature-hashtags.php
@@ -39,8 +39,31 @@ public function load() {
 		// the generated hashtag class.
 		$this->class_prefix = apply_filters( 'liveblog_hashtag_class', $this->class_prefix );
 
-		// Set a better regex for hashtags to allow for hex values in content.
-		$this->set_regex( '~(?:(?<!\S)|>?)(?:(?<!&|&amp;))((?:' . implode( '|', $this->get_prefixes() ) . ')([0-9_\-\p{L}]*))~um' );
+		$prefixes = implode( '|', $this->get_prefixes() );
+
+		// Set a better regex for hashtags to allow for hex values in content -- see https://regex101.com/r/CLWsCo/
+		// phpcs:disable Squiz.Strings.ConcatenationSpacing.PaddingFound -- ignore indentation
+		$this->set_regex(
+			'~'
+			. '(?:'
+			.     '(?<!\S)'          // any visible character
+			. '|'
+			.     '>?'               // possible right angle bracket(s)
+			. ')'
+			. '(?:'
+			.     '(?<!'
+			.         '&'            // literal ampersand
+			.     '|'
+			.         '&amp;'        // encoded ampersand
+			.     ')'
+			. ')'
+			. '('
+			.     "(?:{$prefixes})"  // hashtag prefixes
+			.     '([0-9_\-\p{L}]*)' // 1: numerals, underscores, dashes, and any letter in any language
+			. ')'
+			. '~um'
+		);
+		// phpcs:enable
 
 		// This is the regex used to revert the
 		// generated hashtag html back to the