[xsl] Re: Identity transformation for XSLT 2.0

Subject: [xsl] Re: Identity transformation for XSLT 2.0
From: "Jesper Tverskov" <jesper@xxxxxxxxxxx>
Date: Wed, 19 Dec 2007 12:04:17 +0100

I have now updated my better identity transformation for XSLT 2.0,

The tutorial will also be updated in a day or two.

I have added one more replacement character to protect all forms of
entities simply by replacing the ampersands. This works nicely for
named entities declared in a DTD and for character entities. Compare
that to the traditional identity template replacing e.g. &#160; with a

I now use restricted words as parameters instead of UNICODE characters
from the Private Use Area. Words are clumsy but less confusing, and it
is easier to detect if they are already used in input. Such testing is
also now in place using xsl:message.

The REGEX for DOCTYPE now also gets internal DTD subsets right, I hope.

I have tested it with my poor man's test suite, including the XSLT 2.0
spec and a heavy Chinese homepage well-formed but far from valid.

I both cases the identity transformation is 100%. I would call them
photographic. But there must still be a lot of issues?

Jesper Tverskov

Current Thread