Re: [xml] bug report

[Daniel Veillard <veillard redhat com>]

On Mon, Jun 11, 2001 at 05:46:57PM +1000, Robert Collins wrote:
> -->tests/multiple is wrong though :]. The HTML files have no DOCTYPE,
> and last I checked HTML 4 requires a doctype to be a valid document.

  Valid yes, i.e. allowing SGML validation. However looking at the current
rules for HTML generation:

    http://www.w3.org/TR/xslt#section-HTML-Output-Method

------------------
If the doctype-public or doctype-system attributes are specified,
then the html output method should output a document type declaration
immediately before the first element.
------------------

  So by default, XSLT should not output the doctype if not specified it
seems. Actually some XSLT based processing do generate the doctype outside
of the transformation itself.

  Considering multiple output, it's an extension to the spec. The most
valid definition is the one from the XSLT-1.1 draft which is based
on xsl:output and hence allow doctype-public or doctype-system attributes.
That's the version I suggest to use, it allows to select the doctype level.
If this doesn't work well, I take patches :-)

  Considering xt:document, well in this case libxslt follows what
XT seems to do, Liam Quin provided this test including the results,
and libxslt simply copy it. The namespace is XT's one and the result
is conform to XT's one.

> -->What other things was it breaking? The first version I sent in broke
> just about everything that piped the output, the one I attached in my
> last mail was based on your suggestion of creating a HTML document, not
> a XML document, and shouldn't have altered any output other than that
> with multiple documents, where the document type was HTML.

  yes that was your first patch,

Daniel

-- 
Daniel Veillard      | Red Hat Network http://redhat.com/products/network/
veillard redhat com  | libxml Gnome XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/