ISO/IEC JTC 1/SC 34N0541
ISO/IEC JTC 1/SC 34
Information Technology --
Document Description and Processing Languages
| TITLE: | UK's comments on SC34 Ballot N524: Document Schema Definition Language (DSDL) - Part 3: Rule-base validation (Schematron) |
| SOURCE: | UK National Body |
| PROJECT: | CD 19757-3: Document Schema Definition Language (DSDL) Part 3 - Rule-based validation |
| PROJECT EDITOR: | Mr. Rick Jelliffe |
| STATUS: | Response to SC34 N524 |
| ACTION: | For information |
| DATE: | 2004-09-14 |
| DISTRIBUTION: | SC34 and Liaisons |
| REFER TO: | N0524 - 2004-06-01 - Document Schema Definition Language (DSDL) - Part 3: Rule-base
validation (Schematron) N0524b - 2004-06-01 - Ballot due 2004-09-01 - Document Schema Definition Language (DSDL) - Part 3: Rule-base validation (Schematron) |
| REPLY TO: |
Dr. James David Mason (ISO/IEC JTC 1/SC 34 Secretariat - Standards Council of Canada) Crane Softwrights Ltd. Box 266, Kars, ON K0A-2E0 CANADA Telephone: +1 613 489-0999 Facsimile: +1 613 489-0995 Network: [email protected] http://www.jtc1sc34.org |
Comments on ISO 19757-3 ISO-Schematron
Foreword
Correct list of standards to show correct numbers and titles of parts
Introduction
Add first two paragraphs from Introduction to Part 2 (also used in Part 4)
Change three occurrences of "a normative annexes" to "a normative
annex"
Delete the last paragraph
1. Scope
In second paragraph delete last "the".
3. Terms and Definitions
Order the terms alphabetically so that they are easy to refer to when reading
the text.
3.2
Delete "committee-draft"
3.3
Capitalize It.
3.14
Correct "expression" (only 2 s's)
3.19
Change "and repair" to "together with repair"
4.1
Move second sentence to the start of 5.1 (or move the first para of 5.1 to 4.1
and then adjust the title of 5.1)
4.2
Delete the third item listed (which repeats the second)
How should the first expression be displayed? Is it correct that expressions 2
to 7 are identical in the proof?
5.1
Change "grammar" to "Schematron grammar" and change URI to
form used for Part 4, based on ISO URNs.
Review last paragraph, which is ambiguous. Does "Schematron-namespace
elements" refer to any elements with an sch: prefix or only those used to
define namespaces (e.g. the RNC definition for namespace)? Is a "Schematron-namespace
attribute" one that has the sch: namespace or one that is associated with
an element in the sch: namespace?
5.2
Add sentence pointing out that the formal definitions of core elements are
provided in Annex A.
5.2.2
In note, change "information, however" to "information. Users
should note, however, that".
Some mention should be made that the flag, role and subject attributes are
defined in the following clause.
5.2.4
Change "with" to "containing"
5.2.5
Change "are used" to "are tested".
5.2.6
In Annex A a select attribute has not been defined. I suspect this is an error
in the annex, as select makes much more sense than path for the attribute name.
In second paragraph insert "element" after name.
5.2.7
Change "required prefix attribute" to "optional prefix
attribute" or change model provided in Annex A to conform with statement.
Remove Note 2 - it duplicates preceding sentence.
5.2.8
The use of the name token as the model for name attribute in Annex A. From the
definition given here it looks as if it should have been defined as xs:NAME but
at present it is impossible to work out exactly what the content of the name
attribute should really be.
5.2.9
Explain role of abstract attribute and the link between is-a and param
5.2.10
Delete the second sentence or add a name attribute to the model for phase in
Annex A.
Add definition of #ALL and #DEFAULT to Annex A or remove the second para of this
subclause.
Add description of linkable attributes and explain why they apply here.
5.2.11
In note, change "information, however" to "information. Users
should note, however, that".
5.2.12
Explain role of abstract attribute
5.2.13
Remove "empty" from end of first sentence.
A better name for the "language" attribute might be
"bound-to". (Language is ambiguous, and will be confused with xml:lang.)
5.2.14
Remove "either".
6.3
Either change subject ε schema or, in Note 7, change "the
pattern being a member of the schema" to "the subject being a member
of the schema".
6.4
Define the meaning of the phrase "query language binding" either in
this clause or in the Terms and Definitions clause.
In Note 8, add a comma after xslt. In the next paragraph change "the
specification" to "the requested specification".
Explain what a "specification language" is either in this clause or in
the Terms and Definitions clause (term is used in the last para).
6.5
Add periods to the end of the second and third paras.
Change the (to Europeans meaningless) Australian word "interected" for
one that actually appears in the Oxford English Dictionary!
In Note 9 correct the spelling of constraints.
8.1 and 8.2, first list item
After (RELAX NG) add "compact syntax"
At end of Note 12 and Note 14, add "defined in Annex D"
Annex A
Remove 7th to 18th comment lines
Check that changes noted above have been made and that schema reflects latest
version of normative text.
Should some form of limiting pattern be defined for the strings identified by
the tokens path and fpi?
Should the namespace for the datatype used to validate expr constrain the
expression to be expressed in a particular notation?
Annex B
Change "schema" at end of first line to "Schematron schema:"
Is definition of the rule for context sch:active correct? Why does it not have
the [@pattern] qualifier as is used in the next rule?
Is the definition of the rule for the rule for context sch:extends correct?
Should there not be a [@rule] qualifier for the context path?
Annex C
Add colon at end of first para.
Capitalize P in production (5 occurrences)
D.1
Remove second sentence of first para.
D.2
Replace DTD with RNC definition for consistency.
D.3
Change two occurrences of "Report Validation" to "Validation
Report" and change "SRVL" to "SVRL".
In message for the second rule change "any elements" to "any
embedded elements".
Ensure a blank line precedes each rule
Annex E
What does "line-to-one mapping" mean? Is this another Australian term
that is not known in Europe?