Fix bug involving whitespace-only nodes. Thanks Eric Wald for reporting.

Signed-off-by: Edward Z. Yang <edwardzyang@thewritingpot.com>

NEWS

@@ -15,6 +15,8 @@ NEWS ( CHANGELOG and HISTORY )                                     HTMLPurifier
   does not do the "right thing" with characters not supported in the output
   set.
 - Spellcheck UTF-8: The Secret To Character Encoding
+- Fix improper removal of the contents of elements with only whitespace. Thanks
+  Eric Wald for reporting.
 . Add verbose mode to command line test runner, use (--verbose)
 . Turn on unit tests for UnitConverter
 . Fix missing version number in configuration %Attr.DefaultImageAlt (added 3.2.0)

TODO

@@ -15,7 +15,6 @@ afraid to cast your vote for the next feature to be implemented!
   prevent structures from being parsed and serialized multiple times.
 - Built-in support for target="_blank" on all external links
 - Allow <a id="asdf" name="asdf">
-- Implement overflow CSS property (as per jlp09550)
 - Convert configuration to allow an arbitrary number of namespaces;
   then rename as appropriate.
 

library/HTMLPurifier/ChildDef/Optional.php

@@ -13,8 +13,10 @@ class HTMLPurifier_ChildDef_Optional extends HTMLPurifier_ChildDef_Required
     public $type = 'optional';
     public function validateChildren($tokens_of_children, $config, $context) {
         $result = parent::validateChildren($tokens_of_children, $config, $context);
+        // we assume that $tokens_of_children is not modified
         if ($result === false) {
             if (empty($tokens_of_children)) return true;
+            elseif ($this->whitespace) return $tokens_of_children;
             else return array();
         }
         return $result;

library/HTMLPurifier/ChildDef/Required.php

@@ -10,6 +10,10 @@ class HTMLPurifier_ChildDef_Required extends HTMLPurifier_ChildDef
      * @public
      */
     public $elements = array();
+    /**
+     * Whether or not the last passed node was all whitespace.
+     */
+    protected $whitespace = false;
     /**
      * @param $elements List of allowed element names (lowercase).
      */
@@ -31,6 +35,9 @@ class HTMLPurifier_ChildDef_Required extends HTMLPurifier_ChildDef
     public $allow_empty = false;
     public $type = 'required';
     public function validateChildren($tokens_of_children, $config, $context) {
+        // Flag for subclasses
+        $this->whitespace = false;
+
         // if there are no tokens, delete parent node
         if (empty($tokens_of_children)) return false;
         
@@ -98,7 +105,10 @@ class HTMLPurifier_ChildDef_Required extends HTMLPurifier_ChildDef
             }
         }
         if (empty($result)) return false;
-        if ($all_whitespace) return false;
+        if ($all_whitespace) {
+            $this->whitespace = true;
+            return false;
+        }
         if ($tokens_of_children == $result) return true;
         return $result;
     }

tests/HTMLPurifier/ChildDef/OptionalTest.php

@@ -20,5 +20,12 @@ class HTMLPurifier_ChildDef_OptionalTest extends HTMLPurifier_ChildDefHarness
         $this->assertResult('');
     }
     
+    function testWhitespace() {
+        $this->assertResult(' ');
     }
     
+    function testMultipleWhitespace() {
+        $this->assertResult('    ');
+    }
+    
+}

tests/HTMLPurifier/HTMLT/whitespace-preserve.htmlt

@@ -0,0 +1,3 @@
+--HTML--
+Foo<b> </b>bar
+