RE: [jats-list] Element for wrapping a group of xref elements

Subject: RE: [jats-list] Element for wrapping a group of xref elements
From: "Randall, Laura (NIH/NLM/NCBI) [E]" <laura.randall@xxxxxxx>
Date: Thu, 21 Feb 2013 12:39:40 +0000
>However, downstream renderers of the XML aren't necessarily going to
>know what to do with the empty <xref> in order to display the inline
>citations.

Renderers don't know what to do with ANYTHING until you tell them.

I'm wondering if maybe your request for the new element isn't a request to
fill a need in the XML, but a request to make *your* downstream processing
easier.

This isn't to say that I don't think that's a consideration, because we all
know it is. But most of the things that exist in the Tag Suite now that fall
under processing conveniences are things that have legacy--like italic and
bold instead of just emphasis. So is this *really* a gap in the XML that needs
to be filled, or is it just something that would really make your life
easier?

Laura

________________________________________
From: Alf Eaton [eaton.alf@xxxxxxxxx]
Sent: Thursday, February 21, 2013 6:17 AM
To: jats-list@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: [jats-list] Element for wrapping a group of xref elements

Here's a concrete example of why this would be useful:
<https://peerj.com/articles/3/>, accompanied by a discussion on
Twitter today, about letting readers choose how they want inline
citations to be displayed:
<https://twitter.com/pmgjones/status/304260774459744258>.

In the introduction to the article linked above, each sentence has
multiple references to support it. From the reader's point of view,
that's mostly noise, and it would be nice to be able to collapse each
group of citations down to a single symbol, which would expand to show
all the references on click/hover.

In the absence of an <xref-group> element, there's only one way to do
this with the current JATS DTD, which is to use multiple items in the
rid attribute:

<xref xref-type="bibr" rid="ref-1 ref-3 ref-4 ref-7 ref-10 ref-19
ref-20">(Collingridge, Kehl & McLennan, 1983; Anwyl, Lee & Rowan,
1988; Malenka et al., 1988; Anwyl, Mulkeen & Rowan, 1989; Malenka,
Lancaster & Zucker, 1992; Stevens, Tonegawa & Wang, 1994; Tsien,
Huerta & Tonegawa, 1996; Volianskis & Jensen, 2003)</xref>

and in fact the <label> for each citation is already in the <ref>
element, so the <xref> can be empty:

<xref xref-type="bibr" rid="ref-1 ref-3 ref-4 ref-7 ref-10 ref-19 ref-20"/>

which conveniently leaves the opportunity to state that "the text in
this inline citation must be displayed as-is when the <xref> contains
text", e.g. "as shown in <xref ref-type="bibr" rid="ref-4">Malenka,
1988</xref>".

That (empty xref elements, with multiple, space-separate ids in the
rid attribute; text content in the xref element when it must be
rendered as-is) seems like an appropriate way to store inline
citations.

However, downstream renderers of the XML aren't necessarily going to
know what to do with the empty <xref> in order to display the inline
citations. So, for the sake of compatibility with existing renderers,
perhaps an <xref-group> container for multiple <xref> elements is
still a useful proposal.

Alf

Current Thread