Update regex property extractor to avoid matching og:image tags multiple times and overwriting it's value (like og:image:width overwriting og:image)

1 files changed, 2 insertions, 2 deletions

diff --git a/src/Readability.php b/src/Readability.php
index 7b7eed6..8c55e69 100644
--- a/src/Readability.php
+++ b/src/Readability.php
@@ -287,10 +287,10 @@ class Readability
 
         $values = [];
         // property is a space-separated list of values
-        $propertyPattern = '/\s*(dc|dcterm|og|twitter)\s*:\s*(author|creator|description|title|image)\s*/i';
+        $propertyPattern = '/\s*(dc|dcterm|og|twitter)\s*:\s*(author|creator|description|title|image)(?!:)\s*/i';
 
         // name is a single value
-        $namePattern = '/^\s*(?:(dc|dcterm|og|twitter|weibo:(article|webpage))\s*[\.:]\s*)?(author|creator|description|title|image)\s*$/i';
+        $namePattern = '/^\s*(?:(dc|dcterm|og|twitter|weibo:(article|webpage))\s*[\.:]\s*)?(author|creator|description|title|image)(?!:)\s*$/i';
 
         // Find description tags.
         foreach ($this->dom->getElementsByTagName('meta') as $meta) {