Workaround more invalid HTML cases parsed incorrectly by Mastermind/HTML5 (#6713)

5 years ago · 55cca61134
parent 07d4996825
commit 55cca61134
2 changed files with 27 additions and 4 deletions
--- a/program/lib/Roundcube/rcube_washtml.php
+++ b/program/lib/Roundcube/rcube_washtml.php
@ -782,17 +782,28 @@ class rcube_washtml
     */
    protected function fix_html5($html)
    {
+        // There might be content before html/body tag, we'll move it to the body
+        // We'll wrap it by a div container, it's an invalid HTML anyway
+        if (strpos($html, '<')) {
+            $pos     = stripos($html, '<!DOCTYPE') ?: stripos($html, '<html') ?: stripos($html, '<body');
+            $prefix  = '<div>' . substr($html, 0, $pos) . '</div>';
+            $html    = substr($html, $pos);
+        }
+
        // 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 (isset($prefix) || !preg_match('/<(head|body)/i', $html)) {
+            $body_pos = stripos($html, '<body');
+            $pos      = $body_pos !== false ? $body_pos : stripos($html, '<html');

+            // No HTML and no BODY tag
            if ($pos === false) {
-                $html = '<html><body>' . $html;
+                $html = '<html><body>' . $prefix . $html;
            }
+            // Either HTML or BODY tag found
            else {
                $pos  = strpos($html, '>', $pos);
-                $html = substr_replace($html, '<body>', $pos + 1, 0);
+                $html = substr_replace($html, ($body_pos === false ? '<body>' : '') . $prefix, $pos + 1, 0);
            }
        }

--- a/tests/Framework/Washtml.php
+++ b/tests/Framework/Washtml.php
@ -471,5 +471,17 @@ class Framework_Washtml extends PHPUnit_Framework_TestCase
        $washed = $washer->wash($html);

        $this->assertContains('First line', $washed);
+
+        // Not really valid HTML, but because its common in email world
+        // and because it works with DOMDocument, we make sure its supported
+        $html   = 'First line<br /><html><body>Second line';
+        $washed = $washer->wash($html);
+
+        $this->assertContains('First line', $washed);
+
+        $html   = 'First line<br /><html>Second line';
+        $washed = $washer->wash($html);
+
+        $this->assertContains('First line', $washed);
    }
 }