Closed gimsieke closed 4 years ago
That seems reasonable. I considered it briefly the other day when I was going through the steps for something else, but apparently neglected to actually do it.
Does anyone have a preference for one of the wording alternatives given above?
Nope!
I think the latter wording is a little better. But shouldn't the p:directory-list
step use the directory as the base-uri? I guess it doesn't really matter since (as I recall!) it will have an xml:base
attribute on the document element.
Yes, the p:directory-list
is supposed to have @xml:base
on /c:directory
.
I think it is convenient to stipulate that, in addition, the document will have the same directory URI as its base-uri
property, instead of having no base-uri
property and therefore no base URI.
Without it, base-uri(/*)
will report the directory’s base URI while base-uri(/)
will give an empty sequence.
Precisely!
The following steps state that “no document properties are preserved” despite the fact that there is no input port:
p:www-form-urldecode
p:www-form-urlencode
p:directory-list
(also: add remark thatbase-uri
is the URI of the directory)p:file-copy
p:file-delete
p:file-info
p:file-mkdir
p:file-move
p:file-file-create-tempfile
p:file-touch
(it is not specified what is contained in the result document if the step succeeds; also “one of the following errors” precedes a list of exactly one possible error)p:os-info
I’d say that the concept of property preservation only applies to steps with a (primary) input port.
Apart from the parenthesized remarks in the list above that might require special wording, a proposal for a general change of the document property sentence for these steps is something like this:
or