Merge pull request #570 from bigbitecreative/feature/fix-encoding-issues

Fix encoding issues which affect shortcodes and hashtags

Author: philipjohn

classes/class-wpcom-liveblog-entry-extend-feature-hashtags.php

@@ -39,6 +39,32 @@ public function load() {
 		// the generated hashtag class.
 		$this->class_prefix = apply_filters( 'liveblog_hashtag_class', $this->class_prefix );
 
+		$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
 		// raw input format (e.g #hashtag).

classes/class-wpcom-liveblog-rest-api.php

@@ -651,7 +651,7 @@ public static function sanitize_numeric( $param, $request, $key ) {
 	 */
 	public static function get_json_param( $param, $json ) {
 		if ( isset( $json[ $param ] ) ) {
-			return $json[ $param ];
+			return html_entity_decode( $json[ $param ] );
 		}
 		return false;
 	}