[
https://issues.apache.org/jira/browse/DAFFODIL-828?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16660931#comment-16660931
]
Michael Beckerle commented on DAFFODIL-828:
-------------------------------------------
This is a subject of discussion (Oct 2018) in the DFDL Workgroup. Heretofore,
the assumption was that model groups are "always required" and hence, would
always have a separator in the enclosing sequence, if that sequence has
separators specified.
There are exceptions to this for (a) computed elements (input Value calc), and
(b) hidden groups. But there is a question about whether a group that has no
syntax, and exists only to express a statement such as a dfdl:assert or a
dfdl:setVariable, should get a separator or not.
Further discussion is whether model groups should in fact EVER have separators,
or if the presence of separators is a function of whether there is anything to
separate (something with syntax), and the separatorSuppressionPolicy.
Ultimately this is closely related to issues surrounding trailing separator
suppression, so assigning to myself as I am dealing with those currently.
> IBM Compatibility - Test out HL7 schemas
> ----------------------------------------
>
> Key: DAFFODIL-828
> URL: https://issues.apache.org/jira/browse/DAFFODIL-828
> Project: Daffodil
> Issue Type: New Feature
> Components: Compatibility, DFDL Schemas, Infrastructure, QA
> Reporter: Jessie Chab
> Priority: Major
> Labels: ForInteroperabilityTest
> Fix For: 2.3.0
>
>
> IBM has been creating public DFDL schemas for various data types. When these
> are released, we try to parse data against the schema using Daffodil for
> compatibility's sake. A set of DFDL schemas that model HL7 v2.7 messages are
> now available in a DFDLSchemas GitHub repsitory at
> http://dfdlschemas.github.io/HL7-v2.7/
> ([View on github|https://github.com/DFDLSchemas/HL7-v2.7])
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
