I'm now looking through the existing bib API documentation to see what
we might be able to retain for the bib project, and what would need to
change.
Some quick thoughts:
1) Reference types
One thing we could do is propose to change this to use my class list:
<http://api.openoffice.org/docs/common/ref/com/sun/star/text/
BibliographyDataType.html>
That would go a long way, and would be simple. They're just values.
2) Data model
A more tricky one is this:
<http://api.openoffice.org/docs/common/ref/com/sun/star/text/
BibliographyDataField.html>
The data model is totally flat; everything is just a string. My
approach would treat authors, for example, as arrays of objects.
Likewise containers like edited books, journals, etc.
We'll need to think about how much change we need here. I think the
flat model is highly problematic myself.
Also, the current metadata discussions at the TC could impact on this.
E.g., ideally OOo could get a new general relational metadata API that
we could easily plug into that might replace this.
3) Bibliograhy/Reference list
At least I think this is what this refers to:
<http://api.openoffice.org/docs/common/ref/com/sun/star/text/
FieldMaster/Bibliography.html>
I think we might be able to keep this fairly intact, and just enhance
it with additional properties; for example, an array for et al rules,
configuration to replace duplicate author entries with em-dashes, etc.
I'm a little confused about what is the current API to handle the
citations, but my suspicion is that this part needs to be totally
rewritten. This is what CPH was working on.
Ideally, again, we'd have a base class to handle metadata-enhanced text
fields like this.
Bruce
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
