- From: Norman Walsh <ndw@nwalsh.com>
- Date: Fri, 15 Apr 2016 14:08:07 -0500
- To: public-xml-processing-model-wg@w3.org
- Message-ID: <87twj2667c.fsf@nwalsh.com>
Alex Miłowski <alex@milowski.com> writes:
> Can steps query the readable ports by type and other metadata?
I can see the appeal, but I think it’s going to lead very quickly to a
lot of complexity. Given
[source=$in, stylesheet=$style] -> xslt()
the by-name mapping is easy to explain. Given
[$in, $style] -> xslt()
the ordinal mapping is also easy to explain.
If $in happens to have a content type of “application/xslt+xml” and
$style happens to have a content type of “application/xml”, then it’s
logically reasonable to suggest that $in should be used as the
stylesheet and $style should be used as the document, but I think that
would be very difficult to explain as an alternative to the preceding
two cases.
It’s also unclear what the content type of a sequence of documents
means.
I think we might provide a step that can filter based on content type,
then a pipeline author could do it by hand, in the very rare case that
it wasn’t statically knowable.
[source=($in,$style)] -> content-type(match="application/xml+xslt") >> $ss
[source=($in,$style)] -> content-type(not-match="application/xml+xslt") >> $doc
[$doc, $ss] -> xslt() …
Be seeing you,
norm
--
Norman Walsh
Lead Engineer
MarkLogic Corporation
Phone: +1 512 761 6676
www.marklogic.com
Received on Friday, 15 April 2016 19:08:38 UTC