Hi,<br />I think that rule [20] (and other similar) are wrong:<br />CData ::= (Char* - (Char* ']]&gt;' Char*))<br />The purpose of the rule is to match (reduce) any Char sequence not containing ']]&gt;'.<br />But this result is not achieved since the Char definition includes ']' and '&gt;' so the exception part of the rule:<br />-(Char* ']]&gt;' Char*)<br />is ambiguous. Most parsers solve the ambiguity by applying the rule "reduce as soon, as much as possible"&nbsp;<br />thus the rule will always mismatch because the first Char* reduces also the sequence ']]&gt;' and the next terminal ']]&gt;' will never match.<br />I think the rule (and other similar) should be written:<br />Cdata ::= ( Char - ']]&gt;' )*<br /><br />Best regards<br />Bacchi Raffaele