feat: add directive for removing blank nodes (#404)

library/HTMLPurifier/ConfigSchema/schema/Core.RemoveBlanks.txt

@@ -0,0 +1,10 @@
+Core.RemoveBlanks
+TYPE: bool
+DEFAULT: false
+VERSION: 4.18
+--DESCRIPTION--
+<p>
+    If set to true, blank nodes will be removed. This can be useful for maintaining
+    backwards compatibility when upgrading from previous versions of PHP.
+</p>
+--# vim: et sw=4 sts=4

library/HTMLPurifier/Lexer/DOMLex.php

@@ -72,6 +72,9 @@ public function tokenizeHTML($html, $config, $context)
         if ($config->get('Core.AllowParseManyTags') && defined('LIBXML_PARSEHUGE')) {
             $options |= LIBXML_PARSEHUGE;
         }
+        if ($config->get('Core.RemoveBlanks') && defined('LIBXML_NOBLANKS')) {
+            $options |= LIBXML_NOBLANKS;
+        }
 
         set_error_handler(array($this, 'muteErrorHandler'));
         // loadHTML() fails on PHP 5.3 when second parameter is given

plugins/phorum/config.default.php

@@ -54,5 +54,6 @@
   $config->set('Core.EscapeNonASCIICharacters', true);
 }
 $config->set('Core.AllowParseManyTags', false);
+$config->set('Core.RemoveBlanks', false);
 
 // vim: et sw=4 sts=4

tests/HTMLPurifier/FixtureData/RemoveBlankTestCaseInput.html

@@ -0,0 +1,10 @@
+<table>
+  <caption>
+    Cool Table
+  </caption>
+  <tr>
+    <td>
+      Element 1
+    </td>
+  </tr>
+</table>

tests/HTMLPurifier/FixtureData/RemoveBlankTestCaseOutput.html

@@ -0,0 +1,7 @@
+<table><caption>
+    Cool Table
+  </caption>
+  <tr><td>
+      Element 1
+    </td>
+  </tr></table>

tests/HTMLPurifier/HTMLDefinitionTest.php

@@ -399,6 +399,19 @@ public function test_manyNestedTags()
         $this->assertIdentical($input, $output);
     }
 
+    public function test_removeBlanks()
+    {
+        $config = HTMLPurifier_Config::createDefault();
+        $config->set('Core.RemoveBlanks', true);
+
+        $input = file_get_contents(__DIR__ . '/FixtureData/RemoveBlankTestCaseInput.html');
+        $expected = file_get_contents(__DIR__ . '/FixtureData/RemoveBlankTestCaseOutput.html');
+
+        $purifier = new HTMLPurifier($config);
+        $actual = $purifier->purify($input);
+        $this->assertIdentical($expected, $actual);
+    }
+
 }
 
 // vim: et sw=4 sts=4