Closed susi-wunsch closed 8 years ago
I looked at the RELAX NG and the Schematron for SVRL in the ISO spec for Schematron, and I don't see either foreign elements or attributes being allowed in SVRL.
How do you see them as being allowed in SVRL?
Ok, you are right. SVRL does not allow either foreign or rich attributes or foreign elements. I didn't check that fact. I only looked at the Schematron schema (Relax NG) and its comments and definitions.
They are not only allowed but also promoted for better semantically defined constraints in Schematron itself. How to use these enhancements, if not pushing them to the SVRL where appropriate?
Yes, it doesn't make sense to exclude everything except the text of the assert
or report
from appearing in the SVRL.
It looks like it is necessary to also produce versions of the RELAX NG and Schematron for SVRL that allow foreign and rich content.
Of course, that would be needed. But I couldn't find svrl.rnc
in this project. Would you provide the appropriate folder therefore?
I found some other sources, that show the need for adjusting SVRL according to the Schematron opportunities.
foreign
next to other specific changes (https://stash.utahdcc.org/stash/projects/NEP/repos/nemsis_public/browse/Schematron/utilities/schema/nemsis-svrl.rnc)svrl.rnc
regarding the new draft standard admitting several errors. (http://lists.dsdl.org/dsdl-discuss/2015-01/0000.html)None of the above mentioned changed files help out for both of my needs, rich attributes and foreign attributes and elements in pattern
s.
Should this discussion be continued on an appropriate mailing list or in another project?
Thanks for finding those.
The GoogleCode project didn't have copies of the schemas. I think that the 'Schematron' organization should have a separate project for (its versions of) the schemas.
I've tried twice to subscribe to the Schematron Love-In mailing list (http://eccnet.com/mailman/listinfo/schematron) to discuss this, but I haven't received any confirmations from my requests.
I just discovered by chance the second edition of ISO Schematron ISO/IEC 19757-3:2016(E) and found, that some of my needs are covered by this new (informative) SVRL definition: rich and foreign elements and attributes are now allowed in failed-assert
and successful-report
. But active-pattern
and fired-rule
may not contain foreign elements or attributes or rich attributes. That's a pitty.
I would be interested in discussions about further development of SVRL in order to enable the full power of Schematron. Where to do this? The Schematron-Love-In mailing list seems not to work. How about the other idea of creating a separate project in the "Schematron" organization?
How did you find that out about SVRL? All that I can find without buying a copy is a preview PDF with just the informative front matter in it.
I don't know where to discuss Schematron these days. I'm surprised that there was no mention of the second edition on either Schematron-Love-In or xml-dev.
I got a copy for scientific purposes from our library.
But I have no idea, how to get in contact with people, who supported this second edition of the standard and that way are familiar with the Schematron development issues.
Thanks. What an advantage to have an academic library!
If you're subscribed to Schematron-Love-In, you can ask about the second edition there, and I'll ask on xml-dev.
I've added a 'schema' repository and invited you to join the 'Schematron' organisation. It might take a few iterations, but we should be able to come up with a core first-edition schema plus other modules that import that to make a second-edition version plus an extended version.
I'm sorry, I'm not subscribed to Schematron-Love-In, so I can't assist by asking the right people. It would be nice, if you would rise this topic at xml-dev. I hope, that most of the geeks interested in Schematron follow the xml-dev list. I would welcome some input from people who were involved in the review process of the ISO standard.
Thanks for the 'schema' repository and the invitation. I don't bother some singulary steps towards an extended and useful version of the Schematron and SVRL schema. But I would welcome, if you would create the base folder structure for the repository, leaving me the task to propose small changes.
I've posted to xml-dev already. Rick Jelliffe posts there occasionally, so we'll see if there's any response there.
I'll pre-populate the 'schema' repository, hopefully in less time that it took me to not act on this issue.
Thanks for your actions.
Never mind regarding the silence time. The same occured to me. ;-)
See https://github.com/Schematron/schema. I added the RNC schema and the Schematron from ISO Schematron 2006. I had to correct some syntax errors that were in the standard, and the use of IDREF makes a Schematron that is split into multiple files invalid, so the schemas work, but not well.
@AndrewSales from the UK national committee followed up with the national committee and the ISO committee, and the 2016 edition is now freely available. See http://standards.iso.org/ittf/PubliclyAvailableStandards/c055982_ISO_IEC_19757-3_2016.zip
The changes effect the abstact pattern expansion and the svrl compilation step. I hope, that I didn't oversee anything, that both requirements were already available before my changes. Tested both steps with rich and foreign attributes. The latter step requires 'iso:ns' declaration for a valid xxx-svrl.xsl