Wow, I got my moneys worth on that one. I sort of had my head
in the sand cranking out the unit test code for the SQL Extensions
and I am a little confused with the outcome.
Do I need the code Gary outlined in bugzilla or should it work as
before after Scott's check in ??
I just tested it, and with Gary's workaround it works fine, but it still
does not work as it did before.
Regards
JohnG
[EMAIL PROTECTED] wrote:
> Gary L Peskin <[EMAIL PROTECTED]> wrote:
> > I kind of like just
> > documenting: "Here's how you pass in a node-set, here's how you pass in
> > an RTF, here's how you pass in a Widget ..." with perhaps maybe some
> > small static methods somewhere to free up some of the tediousness of
> > creating the appropriate XObject.
>
> Remember that setParameter is part of the JAXP 1.1 API. I think they
> should be able to pass in a node without calling any special Xalan methods.
> They should be able to switch from XSLTC to Xalan-interpretive (or other
> transformation technology) without doing anything special.
>
> The DOM API is special because it is the defined contract for the source
> tree nodes to the outside world. Therefore, we have to do special
> conversion from our internal tree to the DOM API and visa-versa whenever we
> pass these objects or accept these objects.
>
> BTW, I guess select="java:getNodeName($obj)" is an interesting general
> issue in some ways. What happens if you:
>
> <xsl:variable name="nodeSet" select="foo"/>
> <xsl:value-of select="java:getNodeName($nodeSet)"/>
>
> Would you expect this to work? Dunno. In some ways, it would be nice if
> it would. In other ways, there's no reason that it should work.
>
> -scott
