Closed GoogleCodeExporter closed 9 years ago
As mentioned in the call, it may be that some practical difficulties Keith
mentioned may
be resolved by subcontracting part of the validation to XSD 1.1 instead of to
Schematron.
The assertions appearing currently can be expressed in XSD 1.1 (and, of course,
not in
1.0).
Original comment by rsperb...@gmail.com
on 10 Mar 2010 at 4:52
Original comment by ed.summers
on 10 Mar 2010 at 5:55
Original comment by abdela...@gmail.com
on 25 May 2010 at 5:42
Original comment by abdela...@gmail.com
on 25 May 2010 at 5:44
I have made several revisions to the schema. The biggest change was to remove
the Schematron assertion and instead use Relax NG syntax to state that an
opds:price element is required if an atom:link's rel attribute has a value of
"http://opds-spec.org/acquisition/buy"; that opds:price is optional if it has a
value of "http://opds-spec.org/acquisition/borrow", ".../acquisition/sample" or
".../acquisition/subscribe"; and that no opds:price is permitted with any other
value of rel.
I also changed commas (,) to ampersands (&) in element and attribute lists
where no specific order is required and added explanatory comments for
non-obvious definitions.
Original comment by rsperb...@gmail.com
on 15 Jul 2010 at 9:23
In numerous places, the Atom spec requires an IRI to be used. However, the
atom.rnc schema defines the atomUri pattern only as text.
Since our schema is non-normative, I suggest we use the XSD 1.1 (candidate)
recommendation xsd:anyURI data type to augment/replace the atom.rnc atomUri
pattern. Even the XSD 1.0 anyURI data type will at least permit validation of
URI's (if not IRI's).
Original comment by rsperb...@gmail.com
on 15 Jul 2010 at 9:30
Original comment by abdela...@gmail.com
on 16 Aug 2010 at 7:08
Original issue reported on code.google.com by
ed.summers
on 10 Mar 2010 at 4:31