JATS standard non-alignment case: named-content typographic handling

[castedo]

The typographic handling specified in the NISO JATS standard does not appear to reflect actual adoption by publisher MDPI for actual PMC JATS XML archived in the PMC Open Access Subset.

The HTML rendering by PubMed Central could be interpreted as following the NISO standard, but this results in meaningful loss of scientific information for readers of the HTML version in PMC.

The specification by the NISO standard is:
https://jats.nlm.nih.gov/archiving/tag-library/1.4/element/named-content.html

The NISO standard states that named-content is "... likely to be treated differently, for example, given a stand-out typographic style such as italics or background shading." but does not document how actual JATS XML in PMC specifies italics or background shading or other possible "stand-out" typographic styles.

The NISO standard also states that the difference between <named-content> and <styled-content> is that <styled-content> would be used to, as an example, 'merely tell you that it was always in green shaded background @style="green-shading"'.

But actual XML deposited with in the Open Access Subset makes use of <named-content content-type=> to specify, specifically background color, specifically font color, in addition to many other values that are not specific.

In the case below, there is scientifically relevant information in the publisher's PDF that is lost to readers of the HTML version in PMC. The publisher's PDF is clearly using <named-content content-type=> to set specific colors as background and the author text references those background colors.

For example, in section 2.10 of PMC11762822 is the author text:

"Green, 1× MS2 stem-loop; sequence pink, complementary to the target; yellow, mismatched base."

which describes adjacent nucleotide sequences in the following JATS:

<p><named-content content-type="background:#00ff00">ACATGAGGATCACCCATGTC</named-content><named-content content-type="background:#ff00ff">ATTGCTGCGG</named-content><named-content content-type="background:#FFF44F">A</named-content><named-content content-type="background:#ff00ff">TCCCATCCAG</named-content><named-content content-type="background:#00ff00">AACATGAGGATCACCCATGTC</named-content></p>

It is clear that the intent of the author is to use a specific background color. The PDF reflects the author's intent and provides readers scientifically relevant information by NOT following the NISO standard.

In contrast, the HTML version on PMC follows the ambiguity of the NISO standard to take liberty to style the text in any stand-out typographic style and styles all of the text in italic with no background color. For the reader of the HTML version, the author text now makes no sense and the reader can not see which nucleotide sub-sequenced are being described by the author.

[vincentml]

Hi @castedo

I just saw your comment at https://groups.niso.org/higherlogic/ws/public/view_comment?comment_id=1364 linking to this issue.

The <named-content> element in JATS is intended to be used to tag a "Word or phrase whose content/subject matter has distinct semantics or content-related significance." The <named-content> element is also often used as a mechanism to tag things that do not have a specific element in JATS but need to be tagged for semantic meaning or to support some functionality or display styling. The way in which the <named-content> element is styled for display is the responsibility of the display system, which might use the value of the content-type attribute to apply styling (like a class attribute in HTML/CSS).

By contrast, the <styled-content> element can be used to tag details of the intended display style, as you have pointed out.

The example of adjacent nucleotide sequences uses the <named-content> element in a way that is not intended by the standard. The example could be retagged in a better way, for example:

<p><named-content content-type="nucleotide"><styled-content style="background:#00ff00">ACATGAGGATCACCCATGTC</styled-content><styled-content style="background:#ff00ff">ATTGCTGCGG</styled-content><styled-content style="background:#FFF44F">A</styled-content><styled-content style="background:#ff00ff">TCCCATCCAG</styled-content><styled-content style="background:#00ff00">AACATGAGGATCACCCATGTC</styled-content></named-content></p>

[castedo]

I agree on the intention of the NISO standard writers that <styled-content> should be used in this situation. I see the logic in doing that and I can see the appeal. How appealing logic relates to reality is a difficult predicament.

In case anybody is wondering, I'm not advocating any particular solution to this case of non-alignment. That's up to NISO, PMC, and publishers (like MDPI in this case) to work out ... or not. I'm not party to the non-alignment.

I'm just documenting the reality of a case of non-alignment since I've been researching JATS XML recently.