From a56b43814bbe140e6aa94311fc0308f0fa9b220f Mon Sep 17 00:00:00 2001
From: Christophe Coevoet <stof@notk.org>
Date: Sat, 24 Nov 2018 10:41:49 +0100
Subject: Optimize the handling of references when consuming data

---
 src/HTML5/Parser/Tokenizer.php | 30 ++++++++----------------------
 1 file changed, 8 insertions(+), 22 deletions(-)

(limited to 'src')

diff --git a/src/HTML5/Parser/Tokenizer.php b/src/HTML5/Parser/Tokenizer.php
index 7b0b3f3..4aeeb50 100644
--- a/src/HTML5/Parser/Tokenizer.php
+++ b/src/HTML5/Parser/Tokenizer.php
@@ -121,11 +121,16 @@ class Tokenizer
      */
     protected function consumeData()
     {
-        // Character reference
-        $this->characterReference();
-
         $tok = $this->scanner->current();
 
+        if ($tok === '&') {
+            // Character reference
+            $ref = $this->decodeCharacterReference();
+            $this->buffer($ref);
+
+            $tok = $this->scanner->current();
+        }
+
         // Parse tag
         if ($tok === '<') {
             // Any buffered text data can go out now.
@@ -303,25 +308,6 @@ class Tokenizer
         return false;
     }
 
-    /**
-     * Handle character references (aka entities).
-     *
-     * This version is specific to PCDATA, as it buffers data into the
-     * text buffer. For a generic version, see decodeCharacterReference().
-     *
-     * HTML5 8.2.4.2
-     */
-    protected function characterReference()
-    {
-        if ($this->scanner->current() !== '&') {
-            return false;
-        }
-
-        $ref = $this->decodeCharacterReference();
-        $this->buffer($ref);
-        return true;
-    }
-
     /**
      * Look for markup.
      */
-- 
cgit v1.2.3