More recent of UTR MathClass specs (and a question on diacritics)

[Omikhleia]

Greetings,

If am not mistaken the "mathclass" (on "unicodedata" elements) in the xml-entities XML is derived from Unicode Technical Reports - and as of yet, using MathClass-9, MathClass-11 and MathClass-13 (from the repository files, I haven't check which one if not all were used in the assembling scripts).

These has since been revised:

• MathClass-14 in 2014
• MathClass-15 in 2017

Should it need to be update to ensure being up-to-date?

I have an additional question, which I am unsure pertains to this xml-entities project as-is, but I suppose there are certainly contacts between the various committees involved in math specifications (W3C MathML, Unicode Math, OpenType etc.) so it's probably worth mentioning it:

• U+20E1 COMBINING LEFT RIGHT ARROW ABOVE has mathclass "D" (consistent with UTR MathClass-xx).
• U+034D COMBINING LEFT RIGHT ARROW BELOW doesn't have a mathclass at all, and indeed isn't even listed in the 2017 MathClass-15 iteration. It does sound weird to have such asymmetry, esp. when many TeX -- and TeX math implementations outside TeX, incl. MathJax, TeMML -- support a \underleftrightarrow command, which seems to argue in favor of covering such a character as a (bottom) diacritic in math context.

If the above question is out-of-scope here, please don't hesitate telling me where it should be best addressed.

[davidcarlisle]

It should be up to date with mathclass 15 I certainly have that in the sources.

See

https://github.com/w3c/xml-entities/blob/gh-pages/unicodedata.xsl#L35

which is the script that merges in the data to make the <unicodedata element mathclass attributes.

If the above question is out-of-scope here, please don't hesitate telling me where it should be best addressed.

well it's formally out of scope for the entities spec itself but this is also of course the main repository for unicode.xml so happy to take the issue here. As you surmise it's a small world and the there is a large overlap in personnel in the various activities.

This MathML issue is also relevant w3c/mathml#247

[Omikhleia]

Thanks for the clarification @davidcarlisle
I opened a somewhat related question w3c/mathml-core#137 (comment) which seems to also refer so the more general MathML issue you're referring. 'Feeling a bit lost navigating through the all the resources ;-)

[davidcarlisle]

issues can easily be moved between any of the W3C repos so it doesn't make a lot of difference, often we move issues around if they are first handled in mathml-core or here then need something in the full mathml spec to synchronise, the issue can be moved there. So any more or less sensible repo is fine, and you don't normally need to raise the same issue in two places.