Quibbles with COins

garey

Hello -

I have written COinS into the access pages for some technical reports that I archive, but I am having two sort of problems. The first is that the local identifier for the report is not showing up in the metadata that Zotero pulls from the page. I am hoping that Zotero will recognize and grab the identifier if I put it into an 'rft_id' keyword value, but I have not seen any discussion of which fields Zotero gets from the COinS. Is there one? The second is that I am using the 'book' metadata since the reports are monographs. Zotero is not picking up the 'rft.series' value from the COinS. Because they are reports the series title is important, so this is pretty bad. Is there any way to get Zotero to recognize that tag?

Thanks for any help;

Garey Mills

noksagt

See:
http://forums.zotero.org/discussion/2240/

You might put your local id in rft.description or rft.subject, so that it gets placed in extra or tags (respectively).

I don't know why rft.series isn't working for you. AFAIK, it should. Do you have an example?

garey

You can see the access pages that contain the COinS here

garey

By the way, where are you finding rft.subject and rft.description. They are not in KEV book representation that I can see.

Another thought, is key/value pair order important?

noksagt

By the way, where are you finding rft.subject and rft.description. They are not in KEV book representation that I can see.

sorry--you're right. these are dublin core entities.

By the way, where are you finding rft.subject and rft.description. They are not in KEV book representation that I can see.

It shouldn't be.

noksagt

You can see the access pages that contain the COinS here

Ah--you type those as "reports." Currently, series isn't mapped for that type (it is only mapped for "books" and "book sections").

garey

So, how can I get series mapped for reports?

garey

I would like to revive this discussion. noksagt points me to http://forums.zotero.org/discussion/2240/, where a bunch of terms are listed, but it is unclear which of these can be used when (and the link there requires a username and password to follow it). When I set rft.type = report, which of the terms will be picked up? It is also suggested that I use rft.description to encode the identifier of each report, but then admitted that rft.description comes from the Dublin Core encoding. Apart from that being a rather weird use of the DC description field, can I actually mix DC and book encodings? Finally, how do I get series mapped for reports? I am willing to work.

dstillman

and the link there requires a username and password to follow it

Try the branch instead.

noksagt

It is also suggested that I use rft.description to encode the identifier of each report, but then admitted that rft.description comes from the Dublin Core encoding. Apart from that being a rather weird use of the DC description field, can I actually mix DC and book encodings?

No, you should not mix the encodings--it won't work & it is dirty.

Finally, how do I get series mapped for reports? I am willing to work.

I think that a trivial patch to the COinS parser would work, but I do question the difference in zotero between (depending on type) "series" and "series title."

In general, though: COinS is not able to represent data that is as rich as formats actually intended for exchange of bibliographic information. If you stick with COinS, you will have to hope that some issues are fixed (and can submit patches to this end) and also to put up with some deficiencies that probably can't be fixed easily. Alternatively, you can use unAPI to point to a file to be imported that can be much richer. Or you can look at the bibliographic ontology that Bruce and others are building & see if it is rich enough for you, make suggestions if it is lacking, and wait for it to be finalized/adopted & use it.