IProcessConfiguration.getCharacterData() removes linebreaks?

I'm using my custom process operation participant, which parses its process config via by com.ibm.team.process.common.IProcessConfigurationElement...

However, it seems that loading text nodes (via element.getCharacterData()) remove the linebreaks at load time.

Here's the code to load the character data:

IProcessConfigurationElement[] children = element.getChildren();

for (int i = 0; i < children.length; i++) {

IProcessConfigurationElement child = children[i];

String elementName = child.getName();

if (elementName.equals(IStateChangeHookDefinitions.TAG_SCRIPT)) {

			parsedConfig.fHookCodeScript =child.getCharacterData();

  }

}

The process configuration source looks like this:

<action>

  <script>for (int i = 0; i < 10: i++) {&#13 ;

		&#13 ;

  do something else

 }&#13 ;

  </script>

</action>

Note that there are linebreaks (entity-encoded as &#13 ;)... This gets correctly saved/loaded by the UI part, e.g. IMemento ...
However, loading that config element inside the serverside participant from the config element, results in a single-line String - without linebreaks.

Am I missing something, or is actually not working correctly?
Thanks for any hints...