Hidden tokens are replaced by a whitespace

Question

Hi,
The attached folder contains 3 projects generated with Xtext that allows to have a reproducible example.
I have a short grammar that allows to write paragraphs delimited by begin and end. Comments should be accepted everywhere, that's why they are hidden.
grammar org.xtext.example.commentws.CommentWs
hidden(ML_COMMENT)
CwFile:
{CwFile}
elems+=Elem*;
Elem:
'begin' SEPARATOR content=ID SEPARATOR 'end' SEPARATOR;
SEPARATOR:
LINE_RETURN;
ID:
ALPHABETIC+ (WS+ ALPHABETIC+)*;
terminal ML_COMMENT:
'/*'->'*/';
terminal ALPHABETIC:
'a'..'z' | 'A'..'Z';
terminal WS:
' ' | '\t';
terminal LINE_RETURN:
'\r'? '\n';
Reproductible example
The `tests` module contain a unit test that should success.
The following element content field (Elem.content) should be "some text" but I got "so me text".
begin
so/* a comment that should be ignored */me text
end
Installation details
* Eclipse IDE for Java developers 4.4.2
* Xtext Complete SDK 2.8.1
Is there a way to change this behaviour? Should I make a bug report?

Answer 1

The value converter I should use is the result of a method call.
Is there a way to simply inject that or should I create a new class that implements IValueConverter and that uses an XtextValueConverters?