On 23/11/11 09:14, Stefan Majewski wrote:
> of TEI that is going to be produced. When reading TEI, nevertheless, you
> want to be able to read and process as much as possible. That means
> <biblStruct> as well as plain (sometimes rather unstructured)<bibl>s.
> Then it is sometimes not that straightforward to guesstimate the
> appropriate mapping to Zotero fields.
My suggestion would be to set some minimal standards for import.
(And if you wanted to you could develop a TEI ODD to generate a
schema to validate these.)
For example, I would have your import refuse to process any
<bibl> unless there was
a) at least one <author> or <editor> and
b) at least one <title>
or similar rules you might feel more appropriate.
I don't think there is anything wrong with an import script
refusing to process certain unstructured <bibl> (or other)
elements that don't meet its criteria. However, I'd add that it
needs to clearly identify (by @xml:id? by XPath?, by spewing out
a copy of it?) which <bibl> elements it has ignored and indicate why.
Dr James Cummings, InfoDev,
Computing Services, University of Oxford