According to the XML spec of section 2.4, those "primitive entities" aren't really entities. They're just another way of writing the character escapes and are converted by the parser -- & is precisely equivalent to &x26;. In the DOM, these characters appear as themselves, '&' and '<' and so on, and are manipulated as characters in the string. Conversion back to the character escape form is the responsibility of whatever code serializes the DOM back out to XML syntax.
______________________________________ Joe Kesselman / IBM Research
