- From: Jeni Tennison <jeni@jenitennison.com>
- Date: Mon, 23 Jul 2007 16:35:54 +0100
- To: public-xml-processing-model-wg@w3.org
Shouldn't this section also talk about <p:input>s that *aren't* declarations (ie are used within an atomic step)? We need to say somewhere that in these cases, the 'sequence', 'primary' and 'kind' attributes aren't allowed. In 5.1.1 Document Inputs, third paragraph says: If sequence is specified with the value “yes”, then a sequence is allowed. If sequence is not specified, or has the value “no”, then it is a dynamic error (err:XD0006) for a sequence of more than one document to appear on the declared port. I think this should say that if sequence is not specified or has the value "no" then the port must be provided with exactly one document, or the dynamic error occurs. (At the moment it sounds as if zero documents would be acceptable.) Also in 5.1.1, it talks about the select expression "matching" nodes. But if it's a select expression then it must be *selecting* nodes. It would also be good to explicitly say that the elements are selected at every level (in other words, the <div>s that are selected in the example could occur within another <div>) rather than only at the top level (if, indeed, this is the case -- I can't remember). The syntax summaries imply that it's not legal to put kind="document" on a document input port. Is that really the case? Should it be? Spelling mistake in paragraph four of 5.1.2: s/paramters/parameters In 5.1.2.1: I don't think it should be an error if the name attribute of c:parameter contains a colon and there's a namespace attribute specified: it can just override the normal namespace binding. Section 5.1.2 is very dense, and could do with an example to help illustrate the concepts, particularly to show the defaulting. Jeni -- Jeni Tennison http://www.jenitennison.com
Received on Monday, 23 July 2007 15:36:20 UTC