Forbidden character issue not addressed in XSLT 1.1 WD

It is possible for a result tree to contain characters that are not allowed
in a well-formed XML general parsed entity.

 - the XSLT processor may have obtained the source or stylesheet tree from a
DOM document object containing the forbidden characters

 - the stylesheet may call an extension function that returns an object
containing the forbidden characters

Since there is no provision in the spec for removing these characters, it is
possible for the XML output method to fail to meet the requirement of
serializing the result tree as a well-formed XML general parsed entity.

I suggest adding a clause to require an XSLT processor to remove the
forbidden characters during output, and to signal an error.

I believe this only applies to the XML output method; HTML allows any
character.

It might be prudent to also acknowledge this issue in an addendum or future
revision of the XPath spec.

   - Mike
____________________________________________________________________
Mike J. Brown, software engineer at            My XML/XSL resources:
webb.net in Denver, Colorado, USA              http://skew.org/xml/

Received on Friday, 5 January 2001 17:46:05 UTC