Add Latin hyphenation

[jperon]

This PR is also an issue: some months ago (I don't record exactly when), it was possible to choose hyphenation from a big list, an latin hyphenation thanks to this file, generated from this project was possible.

Since the only option depends on "typographical rules":

• I tried to add the modification proposed in this PR;

• I edited data/hyph/languages.json, adding following lines:

    {
        "name": "Latin",
        "filename": "Latin.pattern",
        "language": "la",
        "left_hyphen_min": 2,
        "right_hyphen_min": 2,
        "aliases": ["lat"]
    },

• I added Latin.pattern into data/hyph.

I'd have expected Latin.pattern to be used as hyphen dictionary; but the only option I have is English-US, although typographical rules are correctly set to la.

---

This change is 

[NiLuJe]

This is because things now require a bit more metadata on CRe's side (c.f., koreader/crengine#383, koreader/crengine#378, koreader/crengine#376, koreader/crengine#372) ;).

[poire-z]

Yes, since a few months ago, some stuff is now hardcoded into crengine.
What you did is just the frontend part, which just makes stuff available in the menu.
You'll need to do a little PR in crengine (see https://github.com/koreader/crengine/commits/master/cr3gui/data/hyph for the little it takes to add each language, some can have a little bit more tweaks if needed).

[jperon]

Just submitted as koreader/crengine#393

[poire-z]

koreader-base updated with the crengine bits, you'll have to update koreader-base in here (you'll be bringing 5 other little commits with it, no worry).

[Frenzie]

Looks like I'm too late, but that should of course be Latin_liturgical.pattern. ;-)

[poire-z]

Going to fix that (no real issue, just a display one, we'll show the hyph filename instead of this pretty language name).
While I'm at it, should these be re-ordered, so the shorter/plain name is first, before the next longer/specific ones?

     { "ko",                    {},   "    ",   _("Korean") },
-    { "la-lit",       {"lat-lit"},   "H   ",   _("Latin (liturgical)"),
     { "la",               {"lat"},   "H   ",   _("Latin"),
+    { "la-lit",       {"lat-lit"},   "H   ",   _("Latin (liturgical)"),
     { "lv",               {"lav"},   "H   ",   _("Latvian"),
@@ -77,9 +77,9 @@ local LANGUAGES = {
     { "pms",                   {},   "H   ",   _("Piedmontese"),
-    { "pt-BR",                 {},   "HB  ",   _("Portuguese (BR)"),
     { "pt",               {"por"},   "HB  ",   _("Portuguese"),
+    { "pt-BR",                 {},   "HB  ",   _("Portuguese (BR)"),
     { "rm",               {"roh"},   "H   ",   _("Romansh"),
     { "ro",               {"ron"},   "H   ",   _("Romanian"),
+    { "ru",               {"rus"},   "Hb  ",   _("Russian"),
     { "ru-GB",                 {},   "Hb  ",   _("Russian + English (UK)"),
     { "ru-US",                 {},   "Hb  ",   _("Russian + English (US)"),
-    { "ru",               {"rus"},   "Hb  ",   _("Russian"),
     { "sr",               {"srp"},   "HB  ",   _("Serbian"),

[poire-z]

(It needs to be in the current order in crengine to match longer ones first - but in frontend, this order does not matter.)

[Frenzie]

Definitely makes more sense to me.

[poire-z]

Done with #6913.