Feature request: <hi> for metrical quantity indicated by scribe

[samosafuz]

It happens in poetic papyri that the scribe indicates the quantity of vowel, using marks we might most easily term 'brevis' and 'longum'. At present, DCLP has been rendering the latter using <hi rend="supraline"> (since the longum is a macron resembling the supraline) but it is preferable in the long run to distinguish the metrical indications. Like other cases of <hi>, the new <hi rend="brevis"> and <hi rend="longum"> will ideally trigger an apparatus entry rather than printing the metrical notation in the text itself.

See, for example, https://papyri.info/dclp/60527, where one finds <hi rend="supraline"> at xi/xii.1 and xi/xii.17.

[samosafuz]

Further to this request, the appropriate Unicode points for these are U+0304 (combining macron) and U+0306 (combining breve). I think I've located where changes to Stylesheets and Navigator .xsl are required, but would appreciate confirmation before I modify and submit a PR for review.

1. https://github.com/EpiDoc/Stylesheets/blob/e034bf0001782e798e8cd702f9c32548c0f8a3e6/tpl-apparatus.xsl#L51
2. https://github.com/EpiDoc/Stylesheets/blob/e034bf0001782e798e8cd702f9c32548c0f8a3e6/tpl-apparatus.xsl#L1376-L1382

3. navigator/pn-xslt/MakeHTML.xsl

   Line 897 in c95ac6a

   .//t:hi[@rend = ('diaeresis','grave','acute','asper','lenis','circumflex')] |

4. navigator/pn-xslt/MakeHTML.xsl

   Lines 921 to 927 in c95ac6a

   	$pass3//t:w[t:g[@type=('apostrophe','high-punctus','middot','low-punctus','diastole','hypodiastole')]] |
   	$pass3//t:w[t:hi[@rend=('diaeresis','grave','acute','asper','lenis','circumflex')]][not(ancestor::t:*[local-name()=('orig','reg','sic','corr','lem','rdg')
   	or self::t:del[@rend='corrected']
   	or self::t:add[@place='inline']][1][local-name()=('reg','corr','rdg')
   	or self::t:del[@rend='corrected']]
   	or ancestor::t:hi)] |
   	$pass3//t:del[@rend='slashes' or @rend='cross-strokes'] | $pass3//t:milestone[@rend = 'box']">

Have I missed any? I'm happy to muck around on my own in search of improvements but am reluctant to break something.

[samosafuz]

I realize as well that this proposal will require new Xsugar, which is also something I'm happy to tinker with. The principal question on this front is what the appropriate Leiden+ for <hi rend="brevis"/> and <hi rend="longum"/> should be, respectively.

I wonder whether (-) and (+) would work for short and long, respectively. Other characters that are both easily accessible on a standard keyboard and that are still unused are ~, *, =, and _. I list them in the hope of avoiding the problem we have with, for example, the question mark character in Xsugar. Would any cause a headache?