- From: Innovimax SARL <innovimax@gmail.com>
- Date: Tue, 5 Jun 2007 00:06:34 +0200
- To: "Norman Walsh" <ndw@nwalsh.com>
- Cc: public-xml-processing-model-wg@w3.org
On 5/29/07, Norman Walsh <ndw@nwalsh.com> wrote: > / "Innovimax SARL" <innovimax@gmail.com> was heard to say: > |> For that matter, what would constitute consistent? If I pass a DocBook > |> document to an XSLT step and I get an HTML document out, how is that > |> consistent? > | > | It is consistent because it is the same content in another form (or format) > > Only if that's what the stylesheet does. Stylesheets are free to > produce any output at all, possibly not related to the input in the > slightest. (Should stylesheets that accept no input document at all > use a different port?) > > Steps read their inputs and produce their outputs. The store step > reads its input and, as its output, produces a document that > identifies the location where the document was stored. That seems > reasonable to me. Ok fair enough > > |> This would require the invention of some entirely new mechanism and > |> I'm strongly opposed. > | > | It is not really new > | It is strictly equivalent as today's > | > | <p:count name="count"/> > | <p:option name="$count" select="@value"> > | <p:pipe step="count" port="result" /> > | </p:option> > > Except that you can't write that today. An option can't occur after > a step. I suppose you could do that with an intervening group, but > I still don't think it's "strictly equivalent". Hum... the definition of pipeline is <p:pipeline name? = NCName p:ignore-prefixes? = prefix list> ((p:input | p:output | p:option | p:parameter | p:import | p:declare-step | p:journal)*, subpipeline) </p:pipeline> so you can have p:option after p:step !! > > At present, we have no mechanism for a step to manufacture a variable > in the XPath context of the pipeline and I have no desire to introduce > such a mechanism. Ok I get your point > > | and it has the advantage of pushing simple types data into options > | instead of XML files > > Except for the simple types that I want to read from other steps, like > XSLT. Are you just saying that we can get p:option out of file, but not the opposite ? > > |> But having the result of p:store be the same as its input seems > |> unnecessary. > | > | That's not mandatory in my proposal. I could leave with p:store having > | NO output at all > > I prefer the status quo. > Mohamed -- Innovimax SARL Consulting, Training & XML Development 9, impasse des Orteaux 75020 Paris Tel : +33 8 72 475787 Fax : +33 1 4356 1746 http://www.innovimax.fr RCS Paris 488.018.631 SARL au capital de 10.000 €
Received on Monday, 4 June 2007 22:06:37 UTC