Fix bug in HTML parser that could cause missing text fragments when there was no head/body tag (#6713)

alecpl · alecpl · commit 03d56926d81d · 2019-04-14T09:53:02.000+02:00
diff --git a/CHANGELOG b/CHANGELOG
@@ -12,6 +12,7 @@ CHANGELOG Roundcube Webmail
 - Elastic: Fix folders list scrolling on touch devices (#6706)
 - Elastic: Fix non-working pretty selects in Chrome browser (#6705)
 - Managesieve: Fix bug where global includes were requested for vacation (#6716)
+- Fix bug in HTML parser that could cause missing text fragments when there was no head/body tag (#6713)
 - Fix bug where HTML messages with a xml:namespace tag were not rendered (#6697)
 - Fix TinyMCE download location (#6694)
 - Fix so "Open in new window" consistently displays "external window" interface (#6659)
diff --git a/program/lib/Roundcube/rcube_washtml.php b/program/lib/Roundcube/rcube_washtml.php
@@ -562,7 +562,7 @@ public function wash($html)
         if (!$this->is_xml && class_exists('Masterminds\HTML5')) {
             try {
                 $html5 = new Masterminds\HTML5();
-                $node  = $html5->loadHTML($html);
+                $node  = $html5->loadHTML($this->fix_html5($html));
             }
             catch (Exception $e) {
                 // ignore, fallback to DOMDocument
@@ -778,6 +778,28 @@ public static function fix_broken_lists(&$html)
         }
     }
 
+    /**
+     * Cleanup and workarounds on input to Masterminds/HTML5
+     */
+    protected function fix_html5($html)
+    {
+        // HTML5 requires <head> or <body> (#6713)
+        // https://github.com/Masterminds/html5-php/issues/166
+        if (!preg_match('/<(head|body)/i', $html)) {
+            $pos = stripos($html, '<html');
+
+            if ($pos === false) {
+                $html = '<html><body>' . $html;
+            }
+            else {
+                $pos  = strpos($html, '>', $pos);
+                $html = substr_replace($html, '<body>', $pos, 0);
+            }
+        }
+
+        return $html;
+    }
+
     /**
      * Explode css style value
      */
diff --git a/tests/Framework/Washtml.php b/tests/Framework/Washtml.php
@@ -444,4 +444,32 @@ function test_xml_namespace()
         $this->assertNotContains('&lt;?xml:namespace"', $washed);
         $this->assertSame($washed, '<p></p>');
     }
+
+    /**
+     * Test missing main HTML hierarchy tags (#6713)
+     */
+    function test_missing_tags()
+    {
+        $washer = new rcube_washtml();
+
+        $html   = '<head></head>First line<br />Second line';
+        $washed = $washer->wash($html);
+
+        $this->assertContains('First line', $washed);
+
+        $html   = 'First line<br />Second line';
+        $washed = $washer->wash($html);
+
+        $this->assertContains('First line', $washed);
+
+        $html   = '<html>First line<br />Second line</html>';
+        $washed = $washer->wash($html);
+
+        $this->assertContains('First line', $washed);
+
+        $html   = '<html><head></head>First line<br />Second line</html>';
+        $washed = $washer->wash($html);
+
+        $this->assertContains('First line', $washed);
+    }
 }
