[
https://issues.apache.org/jira/browse/DAFFODIL-2029?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18022499#comment-18022499
]
Mike Beckerle commented on DAFFODIL-2029:
-----------------------------------------
Doing anything with `lengthUnits='characters'` is easy for a fixed width
charset. The challenge is for variable-width charsets like UTF-8.
Disallowing lengthUnits 'characters' for prefixed length still seems entirely
reasonable to me.
This particular combination of lengthKind and lengthUnits is only theoretical
AFAIK. I've certainly never seen data with this combination.
Did this get selected for 4.1.0 due to wanting to get all the 'prefixed'
issues? If so this one is really orthogonal to all the others and remains very
low priority, as well as quite hard to fix.
> Support unparsing lengthKind="prefixed" with lengthUnits="characters"
> ---------------------------------------------------------------------
>
> Key: DAFFODIL-2029
> URL: https://issues.apache.org/jira/browse/DAFFODIL-2029
> Project: Daffodil
> Issue Type: New Feature
> Components: Unparsing
> Affects Versions: 2.3.0
> Reporter: Steve Lawrence
> Priority: Minor
> Fix For: 4.1.0
>
>
> DAFFODIL-114 added support for lengthKind="prefixed", but does not yet
> support unparsing when lengthUnits="characters". However this is implemented
> can likely be shared with the implementation for dfdl:content-length and
> dfdl:value-length functions where the second paramter is "characters", which
> is also unimplemented.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
