is-numeric in Zotero 4 and 5

tillheilmann · October 17, 2017

I've only noticed this since switching from Zotero 4 to 5: Zotero 4 treated variables like "6.1" as numeric (in an is-numeric conditional). At least that is what my "test suite" PDFs generated with Zotero 4.x show. But Zotero 5 does not do this anymore. It this intended?

bwiernik · October 17, 2017

I don’t think that 6.1 should test true for numeric — there isn’t a meaningful way to transform 6.1 into ordinal or Roman form (which is the main point of testing for this).

@adamsmith Can you confirm that?

tillheilmann · October 17, 2017

Then why does something like U1 test true for numeric?

adamsmith · October 17, 2017

That's correct:

Numbers may have prefixes and suffixes (“D2”, “2b”, “L2d”), and may be separated by a comma, hyphen, or ampersand, with or without spaces (“2, 3”, “2-4”, “2 & 4”). For example, “2nd” tests “true” whereas “second” and “2nd edition” test “false”.

i.e. only , - & can stand between numbers and still have them be numeric (I'd expect en-dash, too, though haven't tested).
(This also explains U1)

Also agree with bwiernik that this isn't just according to specs but also desirable behavior. In the U1 example, U is treated as a prefix, which can at least kind of be handled.

tillheilmann · October 18, 2017

Okay, the thing is that for German publications, at least, it is quite common to have collections such as "Gesamtausgabe", "Gesammelte Schriften" etc. comprising of volumes ("Band") and subvolumes ("Teilband") like

80.1 in Heidegger Gesamtausgabe:

https://www.klostermann.de/epages/63574303.sf/de_DE/?ObjectPath=/Shops/63574303/Categories/Buecher/Philosophie/"Martin Heidegger"/Editionsplan

V.2 in Benjamin Gesammelte Schriften:

http://www.suhrkamp.de/werkausgabe/gesammelte_schriften_sieben_baende_in_teilbaenden_broschur_17.html

IV.5 in Aristoteles Lehrschriften:

https://www.schoeningh.de/katalog/reihe/aristoteles.html

26,1 in Hegel Gesammelte Werke:

https://meiner.de/vorlesungen-uber-die-philosophie-des-rechts-i.html

All of these test false for is-numeric although they are clearly meant to follow a numeric scheme.

Best
Till

adamsmith · October 18, 2017

For one -- do you really need to test is-numeric on these, though? Presumably you won't convert them to roman numerals or ordinals.

Also, in systematic terms, as you point out, these are two different numbers, e.g. Band und Teilband in the Heidegger case. I'm not sure if we'll be able to capture that in Zotero, but ideally that would seem to be the right approach.

tillheilmann · October 22, 2017

In my German reference style, I would like to test is-numeric to decide how to join together, for example, collection-title and collection-number. So, if collection-number is-numeric then no prefix (e.g. "Gesamtausgabe 9"), but if not then a comma should be put in between (e.g. "Gesamtausgabe, Bd. 9: Wegmarken"). I guess I will have to adapt the style to is-numeric behavior.