\cite: expose optional parameter

[pkra]

Right now something like

\cite[Lemma 4.1]{Noz08}

is turned into

<x>[</x><xref ref-type="bibr" rid="bibr-Noz08" specific-use="cite">7<x>, </x>Lemma 4.1</xref><x>]</x>

This makes it difficult to specifically process/style the optional part Lemma 4.1 downstream.

Since PDF output separates these (e.g., the optional part is not part of the link), could texml add some kind of structure in its output?

[pkra]

We should also consider moving [, ] inside the xref (or add a wrapper).

[pkra]

To follow up from today's conversation and #198

Going back to the example

<x>[</x><xref ref-type="bibr" rid="bibr-Noz08" specific-use="cite">7<x>, </x>Lemma 4.1</xref><x>]</x>

Since we may have nested xref's in the optional parameter, we need to move the optional parameter outside the xref.

This reminded me of #56 and the "temporary" cite-group element. Perhaps cite is a good candidate? (which we can pass through to HTML)? E.g.,

<cite><x>[</x><xref ref-type="bibr" rid="bibr-Noz08" specific-use="cite">7</xref><x>, </x>Lemma 4.1<x>]</x></cite>

(I don't care too much if the [ ] are inside or outside.)

[davidmjones]

@pkra How about this?

<cite-group><x>[</x><xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08</xref><x>, </x>Section 5<x>; </x><xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08</xref><x>, </x>theorem 4<x>]</x></cite-group>

I'd prefer to stick with <cite-group/> because (a) then it's mostly just a matter of not stripping those and (b) in the general case there can be multiple individual citations grouped together. Neither <cite/> nor <cite-group/> are defined in JATS/BITS of course.

See #201

[pkra]

Thanks, David. I like cite-group but the code example above wouldn't (naturally) help with the problem downstream - handling the optional parameter. I tend to prefer the optional parameter to be part of the link (in particular to increase link target size) but I could move it into the link downstream -- I'd just really like a wrapper to identify it.

[davidmjones]

@pkra I'm confused. Above you wrote

<cite><x>[</x><xref ref-type="bibr" rid="bibr-Noz08" specific-use="cite">7</xref><x>, </x>Lemma 4.1<x>]</x></cite>

with the optional text outside of the <xref/> and I thought that was the markup you were proposing. Did you leave out the tags around "Lemma 4.1"? Can you give an example of the markup you'd like to see?

[pkra]

I was trying 😄

My example was a single citation, your example was multiple ones grouped together.

Let me use your example and combine the two.

With a wrapper on the outside:

<cite-group>
  <x>[</x>
   <cite><xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08</xref><x>, </x>Section 5</cite><x>; </x>
   <cite><xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08</xref><x>, </x>theorem 4</cite>
<x>]</x>
</cite-group>

With a wrapper on the inside:

<cite-group>
  <x>[</x>
   <xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08<cite-detail><x>, </x>Section 5</cite-detail></xref><x>; </x>
   <xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08<cite-detail><x>, </x>theorem 4</cite-detail></xref><x>]</x>
</cite-group>

Both have their advantages and drawbacks downstream so I'm not perfectly sure which I'd prefer. I started with the first option in mind but the second one is growing on me.

[EDIT: fixed second example]

[davidmjones]

@pkra Thanks for the clarification. Let me know what you think about the new revision.

[davidmjones]

Implemented following markup:

<cite-group>
    <x>[</x>
    <xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08<cite-detail><x>, </x>Section 5</cite-detail></xref>
    <x>; </x>
    <xref ref-type="bibr" rid="bibr-AEG0" specific-use="cite">AEG08</xref>
    <x>]</x>
</cite-group>

[pkra]

@davidmjones is the following statement correct: xref@specific-use=cite will no longer have x elements as children (instead x will only occur as descendants inside cite-detail)?

[davidmjones]

@davidmjones is the following statement correct: xref@specific-use=cite will no longer have x elements as children (instead x will only occur as descendants inside cite-detail)?

Yes. And I've made a note in the source code that this is the expected behaviour downstream.

[pkra]

Thanks, David.