Journals without page numbers (e.g., Article 7 or Paper 122)

jameshowison

I'm working with Zotero 4.04 stand-alone and the MIS Quarterly style, but I think this applies more broadly.

Some journals or proceedings in my field don't use page numbers, but supply an Article or Paper number. An example is JAIS:

In Zotero for Journal Article I don't see a field for Number or similar. So I've followed convention and put pages="Article 7".

In the MIS Quarterly style (and many others) that gets rendered as pp. Article 7. What is desired is

Crowston, K., and Wade, M. 2010. “Introduction to JAIS Special Issue on Empirical Research on Free/Libre Open Source Software,” Journal of the Association for Information Systems (11:11), Article 7.

One thought was to use is-numeric on page and not include the pages label if false. i.e.:

<macro name="page-locator">
<group prefix=", ">
<choose>
<if is-numeric="page">
<group delimiter=" ">
<label variable="page" form="short"/>
<text variable="page"/>
</group>
</if>
<else>
<text variable="page"/>
</else>
</choose>
</group>
</macro>

This doesn't work because is-numeric="page" always returns false. My reading of the spec is that 10-23 should return true for is-numeric. But it may be that is-numeric doesn't work with pages (which is special due to the page-range formatting?)

I don't particular like putting "Article 7" in the pages field anyway, seems semantically wrong. But that would seem to require adding another field to Journal Article in Zotero?

Any advice?

--J

adamsmith

Bumping this - I believe this is a processor bug.

This doesn't work because is-numeric="page" always returns false. My reading of the spec is that 10-23 should return true for is-numeric.

I can replicate this and the specifications don't have anything about an exception for page.

I also think we may need to think about adding another option for "position" or so that can cover locations in electronic resources, be it Kindle numbers or positions on a webpage, but that's a different story. Right now I'd like to sort out if we agree this is a processor bug (and if so, ask Frank to fix it).

fbennett

Now fixed in the latest processor patch plugin.

Thanks for bringing this up, and for your patience. IIRC, the early CSL spec (the comments in the schema) provided that is-numeric should only test true for numeric variables. Page is not in that list, but you're right that the current is-numeric description doesn't say anything about requiring that it be there. I do seem to remember that Rintze mentioned the change to broader behaviour to me at some point, but I didn't get around to making the change.

(Some other variables may be similarly restricted, contrary to the specification. I'll be cautious about extending the set of testable variables for the present, to be sure the processor is happy with it.)

adamsmith

thanks!

jameshowison

Good stuff. (hopefully) quick question:

That link is to an .xpi; does that mean I can update the CSL processor in Zotero by installing it? Any documentation on that? Cool stuff :)

For future googlers, this is the change set (to citeproc-js)

https://bitbucket.org/fbennett/citeproc-js/commits/8069632ff0f39f03d02c2a84bce049e0241415ab

adamsmith

That link is to an .xpi; does that mean I can update the CSL processor in Zotero by installing it? Any documentation on that? Cool stuff :)

yes, that's what it does, and no, I don't think there is any documentation. It's a tool that Frank developed for faster troubleshooting/fixes that does nothing but switch out the version of citeproc in Zotero.

fbennett

Yep, it's just a simple little thing that overlays a different copy of the processor. The code for it is here.