The value of @ident of ﹤valItem﹥, which is for “here is a value that belongs in an enumerated list”, is defined as xsd:NCName (via data.name). The datatype data.enumerated, which is for “put a value from an enumerated list here”, is defined as rng:string (via data.text).
It’s not clear to me what they should be (I’m leaning towards data.word), but they should match.
Diff:
Doesn't data.enumerated have to allow multiple tokens, and therefore have to permit whitespace?
no. you use data.enumerated inside a oneOrMore to allow a sequence of non-whitespace-containing tokens.
SFAICR
You're right of course, In which case, I vote for data.name, then; I'd prefer that enumerated values be NCNames, so that it's safer to convert them to other X(HT)ML things such as tag and attribute names and class names.
EM, FC, and HC agree this should be data.name.
That issue seems to be connected to [bugs:#681]. The discussion there tends to data.word …
Related
Bugs:
#681F2F at Raleigh agrees that this needs clarification, but more discussion.
One issue is that
<valItem/>
can also be used within<content>
thus defining the content model of an element (e.g.<term>
). In this case it might be rational to allow for whitespace in fixed terms.Just came across http://tei-l.970651.n3.nabble.com/do-you-use-valList-in-content-in-your-TEI-ODD-tp2345217.html
This seems somehow related …