Re: exporting zotxt or orgref links to HTML and ODF

[John Kitchin <johnrkitchin@gmail.com>]

Christian Moe writes:

> Richard Lawrence writes:
>
>> It looks to me like Pandoc has a quite general solution, and it also
>> looks like Org could use Pandoc's citation syntax as-is.  I would
>> suggest borrowing this syntax as a starting point for building citation
>> support into Org.
>
> It's been years since I looked at Pandoc, and I think they've added some
> functionality since then. Prefix, locator, suffix, and multiple
> references in one human-readable citation: Great! And /much/ nicer to
> look at than latex \cite commands with their frankly bizarre placement
> of locators etc.
>
>> Blah blah [see @doe99, pp. 33-35; also @smith04, ch. 1].
>
> In my current homebrewn solution for Zotero, I have tried to do
> something similarly readable using Org link syntax (sorry, Rasmus!) with
> the database entry ID as link target, and parsing the description part
> for prefix/author-date/locator/suffix, but with a slightly different
> syntax than Pandoc uses. In my solution the above would be:
>
> Blah blah [[zotero:0_A43F89;0_E25CB3][(see: Doe 1999: p.33-35; also:
> Smith 2004: ch. 1)]].
>
>> A minus sign (-) before the @ will suppress mention of the author in the
>> citation. This can be useful when the author is already mentioned in the
>> text:
>>
>> Smith says blah [-@smith04].
>
> In my current Zotero solution:
>
> Smith says blah [[zotero:0_E25CB3][(2004)]].
>
>> Does anyone have citation needs that this syntax doesn't cover?
>
> It's great, as long as your database uses mnemonic citekeys like
> doe99. Zotero doesn't, but uses keys that are meaningless to humans,
> like 0_A43F89.  Unfortunately [see @0_A43F89, p. 5] wouldn't look nearly
> as nice as [see @doe99, p.5], and it wouldn't help you remember what you
> referenced.

This doesn't have to be case. Even bibtex keys are not mnemonic enough
in my opinion (maybe it is fairer to put the limitation on what I can
remember though ;) I recently implemented a neat idea in org-ref that
will show you the full citation in the minibuffer when your cursor is
idle over a cite link. See https://www.youtube.com/watch?v=8cEb6F9AEu0
to see it in action. I played around with many variations of tool tips,
etc... and settled on the minibuffer as the lightest weight, least
disruptive flow.

On the other hand, if the concern about mnemonic is remembering what to
type in, then you should consider using a completion tool. org-ref has
completion by key (which I do not use because I do not remember keys),
but also key selection through either reftex or helm-bibtex. It is
surprisingly easy to make a helm selection buffer. All you need is a way
to get the possible candidates from zotero, e.g. by parsing the entries
to get a list of (description . zotero-key) for each entry. description
is what you see in helm to narrow, and zotero-key is what is sent to an
action function that takes care of formatting and inserting the keys.

>
> I think the typical workflow combining Zotero with Pandoc is to export a
> BibTex file from Zotero and reference the BibTex citekeys from
> there. I could live with that much of the time.
>
> But that workflow doesn't help with something I often want to do, which
> is to export to ODT and have 'live' Zotero citations that I can continue
> to work with in LibreOffice.

I don't do this often, but if you know what the zotero citation format
in odt is, I think you could get org to export it for that backend.

>
>> Using this syntax would also have the advantage that Pandoc can already
>> parse it, which would reduce friction for Org users who convert their
>> documents with Pandoc (and Pandoc users who need to deal with Org
>> inputs).  Since this seems like a significant contingent of Org users,
>> that's something to consider.
>
> That's a good point. OTOH, don't Org users convert their documents with
> Pandoc mostly because cross-backend citation support is lacking?
>
>> The bigger question is whether, in addition to a citation *syntax*, it
>> would be a lot of work to add support for the various citation database
>> formats, as well as the various output styles, and which ones to
>> support.
>
> Possibly more work if it's worth if we adopt Pandoc syntax,
> since Pandoc-citeproc seems to handle nearly everything that is based on
> plain text.
>
> To truly support citations natively, we'd essentially have to implement
> something like citeproc in elisp. Not that I haven't been thinking about
> that...
>
> Yours,
> Christian

--
Professor John Kitchin
Doherty Hall A207F
Department of Chemical Engineering
Carnegie Mellon University
Pittsburgh, PA 15213
412-268-7803
@johnkitchin
http://kitchingroup.cheme.cmu.edu