Add support for dpub-aria in HTML output

[xi]

The Digital Publishing WAI-ARIA Module 1.0 has recently become a W3C recommendation. It contains semantic structures that are derived from EPUB. I think that these that are well suited for many of pandoc's usecases.

The relevant parts are:

• Use doc-endnotes, doc-endnote, doc-noteref, and doc-backlink roles for footnotes.
• Use doc-bibliography, doc-biblioentry, and doc-biblioref roles for bibliography.
• Use doc-toc role for auto-generated table of contents.

Unfortunately, I do not know enough haskell to implement this myself.

See also michelf/php-markdown#287

[mb21]

This would only require changes to https://github.com/jgm/pandoc/blob/master/src/Text/Pandoc/Writers/HTML.hs

[link2xt]

@mb21 why not also reader?

[mb21]

@labdsf as I understood, this was only about output.. but sure, input can be considered as well... just parse them to Attr?

[link2xt]

HTML reader does not parse footnotes/endnotes now, but given these attributes, it can.

[jgm]

+++ Alexander [Dec 30 17 12:27 ]:

HTML reader does not parse footnotes/endnotes now, but given these attributes, it can.

Yes, this would be worth doing, though a bit complex. Probably the easiest way would be to store the note references as Spans a special attribute, and store the notes themselves as Divs with a special attribute. Then, after parsing, apply an AST transformation that matches these and converts the Spans into regular notes.

[Porges]

I just came here looking for this. Might tackle a few things myself...

[jgm]

• doc-bibliography, doc-biblioentry, and doc-biblioref will need to be handled in pandoc-citeproc.
• Reader changes to parse footnotes are still TBD.