Re: [jats-list] I'm looking for examples of tagging molecular strucutres in the nlm 3.0 archiving tag set

Subject: Re: [jats-list] I'm looking for examples of tagging molecular strucutres in the nlm 3.0 archiving tag set
From: Wendell Piez <wapiez@xxxxxxxxxxxxxxx>
Date: Fri, 9 Nov 2012 09:54:07 -0500
Ian,

In the tagging you gave, there is text data appearing between two
elements (where it isn't permitted) -- that URL
"http://www.rcsb.org/pdb/explore/explore.do?structureId=1CRN"; between
inline-supplementary-material and ext-link. Correct that and you are
formally valid.

However, as Jeff indicated, formal validity is only part of the story.
You need a spec for what constitutes correct tagging -- what your next
steps in your various pipelines are obliged to be able to handle. And
ideally, you need a way to validate your data against the rules of
your spec outside your applications ... so you can catch and fix your
(inevitable) errors early.

(As you probably know, Schematron is a good way to do that although
there are also other means.)

Exactly what your spec should be depends on the details of what
form(s) your data takes and what you need to do with it. For example,
it might describe fallback behaviors for your applications to use if
certain bits of data are missing, in which case those bits might be
classed as optional in your validation rules.

I know this is very general but it's hard to be more specific without
knowing more about your system. :-)

Cheers,
Wendell

On Thu, Nov 8, 2012 at 5:14 PM, I.mulvany@xxxxxxxxxxxxxxxxx
<i.mulvany@xxxxxxxxxxxxxxxxx> wrote:
> I have three application layers in mind:
>
> PDF generation, the rules of which we can control at the typesetting level.
> Gamol interactive viewer at the publisher site HTML level, again we have control on that layer.
> Pmc deposit. I'm not too concerned on the choices that the pmc layer will go for, as long as it can accept the XML for deposit purposes.
>
>
>
> Sent from my iPad
>
> On 8 Nov 2012, at 20:44, "Beck, Jeff (NIH/NLM/NCBI) [E]" <beck@xxxxxxxxxxxxxxxx> wrote:
>
>> The one thing you will need to keep in mind when writing <alternatives>
>> anywhere is how your application will use the contents - and how anyone
>> who gets your data will use them.
>>
>> Different applications may "use" the contents of an <alternatives>
>> differently - this is part of their charm. Given, for example, a
>> disp-formula/alternatives with a graphic and mml:math inside, non-MathML
>> applications will go for the graphic. But applications that can handle
>> both types will need some rules - maybe use mml first and fall back to
>> graphic. Or fail through the contents of <alternatives> in document order.
>>
>> Either way, when you start making rules for how the contents will be used,
>> you will want to make sure you have what you need in every instance. XML
>> validation (with the schemas out of the box) will not get this for you.
>> You will want to have some application-specific validation layer.
>>
>> Jeff
>>
>> On 11/8/12 3:33 PM, "Wendell Piez" <wapiez@xxxxxxxxxxxxxxx> wrote:
>>
>>> Hi Ian,
>>>
>>> As I understand it, you are using 'alternatives' exactly correctly.
>>>
>>> I have a couple of questions about the contents as given in your
>>> example, but I suspect there may be transcription errors there. (Among
>>> other issues, it isn't valid.)
>>>
>>> In any case, the use case you describe is just what 'alternatives' is
>>> for, as documented in the Tag Library:
>>>
>>> http://jats.nlm.nih.gov/publishing/tag-library/1.0/index.html?elem=alterna
>>> tives
>>>
>>> Cheers,
>>> Wendell
>>>
>>> On Thu, Nov 8, 2012 at 10:17 AM, Ian Mulvany
>>> <i.mulvany@xxxxxxxxxxxxxxxxx> wrote:
>>>> Specifically would it be possible to use the <alternative> tag to
>>>> contain both a pointer to a local server side hosted .pdb file, and a
>>>> reference to a named pdb entity?
>>>>
>>>> I'm thinking of something like the following:
>>>>
>>>> <chem-struct-wrap>
>>>>      <caption>
>>>>      <p>This is the caption for the chemical structure. This tagging
>>>> includes reference to an
>>>>        image to be used in processing the XML. It also includes
>>>> reference to a locally hosted in-line-supplimentary data file with the
>>>> pdb data held locally. It also contains a reference to an external
>>>> link for a pdb resource. It also contains a marked up pdb named
>>>> entity.</p>
>>>>      </caption>
>>>>      <alternatives>
>>>>      <graphic xmlns:xlink="http://www.w3.org/1999/xlink";
>>>>      xlink:href="pq0209587032" specific-use="print"></graphic>
>>>>      <chem-struct>1crn</chem-struct>
>>>>      <inline-supplementary-material
>>>> xmlns:xlink="http://www.w3.org/1999/xlink";
>>>> xlink:title="local_pdb_file"
>>>> xlink:href="1crn.pdb">1crn.pdb</inline-supplementary-material>
>>>>      http://www.rcsb.org/pdb/explore/explore.do?structureId=1CRN
>>>>      <ext-link ext-link-type="pdb"
>>>>
>>>> xlink:href="http://dx.doi.org/elife/10.7554/eLife.00065.008";>1CRN</ext-li
>>>> nk>
>>>>      </alternatives>
>>>> </chem-struct-wrap>
>>>>
>>>> Has anyone any experience? What am I doing wrong here (I always start
>>>> from the assumption that I am doing something wrong)
>>>>
>>>> - Ian
>>>
>>>
>>>
>>> --
>>> Wendell Piez | http://www.wendellpiez.com
>>> XML | XSLT | electronic publishing
>>> Eat Your Vegetables
>>> _____oo_________o_o___ooooo____ooooooo_^
>



-- 
Wendell Piez | http://www.wendellpiez.com
XML | XSLT | electronic publishing
Eat Your Vegetables
_____oo_________o_o___ooooo____ooooooo_^

Current Thread