List of errors in XML schema spec of AOM

Technical (archive)
1

Hi again,

As promised, I will provide a list of errors in the XML schema specification of the AOM.

  1. Element name ‘rm_type_name’ should be renamed to ‘rm_attribute_name’ for the C_ATTRIBUTE entity so that the naming follows the AOM specification.

  2. The C_ATTRIBUTE has an attribute called ‘unbounded’ which follows the AOM specification the existence attribute which is an Interval, but is it really needed in the XML spec, i.e. will there ever be a case where the existence is unbounded?

http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h-1167105029.html

  1. The element name ‘precedence_overriden’ should be corrected to ‘precedence_overridden’ here
    http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h-223793303.html

  2. There’s an entity called ‘C_CODED_TERM’ which is a DOMAIN_TYPE, but it should probably be renamed to ‘C_CODED_TEXT’. If this is a correct assumption, the ‘reference’ element should then be renamed to ‘query’ and another element called ‘subset’ should also be added. See the specification of C_DV_CODED_TEXT in the Archetype Profile specification here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/architecture/am/openehr_archetype_profile.pdf .

XML spec of ‘C_CODED_TERM’ is found here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1644545557.html

  1. In ADL archetypes the ontology section usually contains ‘primary_language’, ‘languages_available’, but neither of these are found in the ARCHETYPE_ONTOLOGY entity in the XML schema spec. When looking through the AOM specification of the ARCHETYPE_ONTOLOGY I found that these attributes didn’t exist there either - so how come the ADL archetypes contain these attributes?. I can only guess they haven’t evolved as fast as the specifications…

Furthermore, there are two attributes in the AOM specification of ARCHETYPE_ONTOLOGY that I couldn’t find in the ADL archetypes - namely ‘specialisation_depth’ and ‘parent_archetype’. These attributes are not accessible from the ArchetypeOntology class in the 1.0.1-RC1 version of the Java kernel, and the latter is not found in the XML schema spec.

See the XML schema spec of ARCHETYPE_ONTOLOGY here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1873534858.html

  1. All the sub-classes of C_PRIMITIVE has an attribute called ‘assumed_value’, but in the XML schema specs it has mistakenly been spelled ‘assummed_value’.
    http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1279968838.html

Regards,

Mattias

2

Mattias
Thanks - we were not really prepared for people to run with this right now but I include a candidate schema which is in tune with the ADL Parser and encorporates your observations

  1. Element name ‘rm_type_name’ should be renamed to ‘rm_attribute_name’ for the C_ATTRIBUTE entity so that the naming follows the AOM specification.

Andrew P pointed this out some time ago and it is corrected.

  1. The C_ATTRIBUTE has an attribute called ‘unbounded’ which follows the AOM specification the existence attribute which is an Interval, but is it really needed in the XML spec, i.e. will there ever be a case where the existence is unbounded?

I believe that it can be unbounded..Tom might want to comment.

http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h-1167105029.html

  1. The element name ‘precedence_overriden’ should be corrected to ‘precedence_overridden’ here
    http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h-223793303.html

OK

  1. There’s an entity called ‘C_CODED_TERM’ which is a DOMAIN_TYPE, but it should probably be renamed to ‘C_CODED_TEXT’. If this is a correct assumption, the ‘reference’ element should then be renamed to ‘query’ and another element called ‘subset’ should also be added. See the specification of C_DV_CODED_TEXT in the Archetype Profile specification here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/architecture/am/openehr_archetype_profile.pdf .

XML spec of ‘C_CODED_TERM’ is found here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1644545557.html

Actually the constraint required is on the CODE_PHRASE and in the new ADL Parser it is called C_CODE_PHRASE. This does not affect the ADL so noone noticed the change. The schema now has C_CODE_PHRASE instead.

  1. In ADL archetypes the ontology section usually contains ‘primary_language’, ‘languages_available’, but neither of these are found in the ARCHETYPE_ONTOLOGY entity in the XML schema spec. When looking through the AOM specification of the ARCHETYPE_ONTOLOGY I found that these attributes didn’t exist there either - so how come the ADL archetypes contain these attributes?. I can only guess they haven’t evolved as fast as the specifications…

The AOM is version 2 - so I thought we might as well go with this for the XML. Languages available is now in the translations.

Furthermore, there are two attributes in the AOM specification of ARCHETYPE_ONTOLOGY that I couldn’t find in the ADL archetypes - namely ‘specialisation_depth’ and ‘parent_archetype’. These attributes are not accessible from the ArchetypeOntology class in the 1.0.1-RC1 version of the Java kernel, and the latter is not found in the XML schema spec.

parent_archetype_id is part of archetype in the Schema.

See the XML schema spec of ARCHETYPE_ONTOLOGY here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1873534858.html

  1. All the sub-classes of C_PRIMITIVE has an attribute called ‘assumed_value’, but in the XML schema specs it has mistakenly been spelled ‘assummed_value’.
    http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1279968838.html

Andrew picked this up in the Schema also - it has been changed.

Still interested in the changing the term.text and description to attributes - any interest? The implementation is not given in the AOM.

Cheers, Sam

(attachments)

Archetype.xsd (20.4 KB)

3

2006/11/21, Sam Heard <sam.heard@oceaninformatics.biz>:

Mattias
Thanks - we were not really prepared for people to run with this right now but I include a candidate schema which is in tune with the ADL Parser and encorporates your observations

Hi Sam,

Thanks, I’ll have a look at it.

  1. There’s an entity called ‘C_CODED_TERM’ which is a DOMAIN_TYPE, but it should probably be renamed to ‘C_CODED_TEXT’. If this is a correct assumption, the ‘reference’ element should then be renamed to ‘query’ and another element called ‘subset’ should also be added. See the specification of C_DV_CODED_TEXT in the Archetype Profile specification here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/architecture/am/openehr_archetype_profile.pdf .

XML spec of ‘C_CODED_TERM’ is found here: http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/its/XML-schema/documentation/Archetype.xsd.html_h1644545557.html

Actually the constraint required is on the CODE_PHRASE and in the new ADL Parser it is called C_CODE_PHRASE. This does not affect the ADL so noone noticed the change. The schema now has C_CODE_PHRASE instead.

But the C_CODE_PHRASE is not a C_DOMAIN_TYPE according to the Archetype Profile. Where can I find this class specification? If you look at the schema you see that C_CODED_TERM inherits C_DOMAIN_TYPE…

Furthermore, there are two attributes in the AOM specification of ARCHETYPE_ONTOLOGY that I couldn’t find in the ADL archetypes - namely ‘specialisation_depth’ and ‘parent_archetype’. These attributes are not accessible from the ArchetypeOntology class in the 1.0.1-RC1 version of the Java kernel, and the latter is not found in the XML schema spec.

parent_archetype_id is part of archetype in the Schema.

Okay, I guess the XML schema isn’t completely equal to the AOM in this case because it would require to output yet another archetype in the ARCHETYPE_ONTOLOGY.

Still interested in the changing the term.text and description to attributes - any interest? The implementation is not given in the AOM.

This is found in the AOM:

items: Hash <String, String>
Hash of keys (“text”, “description” etc) and corresponding values.

at page 46
http://svn.openehr.org/specification/BRANCHES/Release-1.1-candidate/publishing/architecture/am/aom.pdf

Cheers,

Mattias

4

We need Tom’s input on this. Sam

Mattias Forss wrote:

5

Sam Heard wrote:

Mattias
Thanks - we were not really prepared for people to run with this right
now but I include a candidate schema which is in tune with the ADL
Parser and encorporates your observations

1. Element name 'rm_type_name' should be renamed to
'rm_attribute_name' for the C_ATTRIBUTE entity so that the naming
follows the AOM specification.

Andrew P pointed this out some time ago and it is corrected.

2. The C_ATTRIBUTE has an attribute called 'unbounded' which follows
the AOM specification the existence attribute which is an
Interval<Integer>, but is it really needed in the XML spec, i.e. will
there ever be a case where the existence is unbounded?

I believe that it can be unbounded..Tom might want to comment.

Mattias is correct here semantically. Unbounded is in the AM simply
because we used the generic type Interval<T>, even though in this
instance it can only ever be 0..0, 0..1, or 1..1

- thomas

6

I have removed this attribute. Sam

Thomas Beale wrote: