Creates a version of Module:etymology languages/data in which there is just one code per language. Codes that have the same language data are reduced to the most language-codiest one. For instance, Austrian German has three codes (Austrian German, AG., de-AT) by which it can be accessed in etymology templates like {{cog}}. de-AT is chosen as the only language code for Austrian German because it looks the most like a language code.
The following criteria are used successively to weed out candidates for language-codiest code:
This list shows those etymology languages that have multiple codes, together with the code chosen by this module and the other codes:
fr-aca): Acadian French, fra-acaen-US): AE., American Englishatn): xme-astde-AT): AG., Austrian Germanen-GB): BE., British Englishfr-CA): CF., Canadian Frenchckm): sh-chasco-osc): Early Scots, O.Sc., Old Scotsbgp): bal-easla-ecc): EL., Ecclesiastical Latinxme-gaz): gzinrf-grn): roa-grnsco-ins): Ins.Sc., Insular Scotsnrf-jer): roa-jerkjv): sh-kajbsh-kam): xvikfm): xme-xunksh): KölschLL): LL., Late Latin, la-latlng): Lombardic, goh-lnges-lun): Lunfardola-med): ML, ML., Medieval Latinira-mid): MIr.sco-smi): Mid.Sc., Middle Scotsntz): xme-natxme-nay): nyqla-new): NL., New Latinsco-nor): Nor.Sc., Northern Scotsira-old): OIr.it-oit): roa-oitfro-nor): ONF., Old Northern Frenchqfa-sub-grc): pregrcoc-pro): prvla-ren): RL., Renaissance Latinsoj): xme-soibal-sou): bccsco-sou): Borders Scots, Sou.Sc., Southern Scotsit-CH): Swiss Italianfr-CH): Swiss French, Switzerland Frenchsco-uls): Uls.Sc., Ulster Scotsxme-vaf): vafde-AT-vie): VG., Viennese Germanla-vul): VL., Vulgar Latingbz): xme-darqfa-pyg): pygmygkm): Medieval Greekgrc-koi): Koinemul-tax): Tax.local language_codes = require "Module:languages/code to canonical name"
local function determine_preferred_etymology_language_code(code1, code2)
if code2:find "^+$" then
if code1:find "^+$" then
if not code2:find "%u%l" then
if not code1:find "%u%l" then
if #code2 < #code1 then
return code2
else
-- Prefer nrf-grn and nrf-jer over roa-grn and roa-jer
-- (Guernsey and Jersey).
local first_word1, first_word2 =
code1:match "^+", code2:match "^+"
if first_word1 and first_word2
and language_codes then
return code1
else
return code2
end
end
else
return code2
end
else
return code1
end
else
return code2
end
else
return code1
end
end
local function fold(t, accum, func)
for k, v in pairs(t) do
accum = func(k, v, accum)
end
return accum
end
local function invert(t)
local inverted = {}
for k, v in pairs(t) do
inverted = k
end
return inverted
end
return invert(fold(
require "Module:etymology languages/data",
{},
function (code, data, data_to_code)
if data_to_code then
local preferred_code = determine_preferred_etymology_language_code(data_to_code, code)
data_to_code = preferred_code
table.insert(data.codes, code)
else
data_to_code = code
data.codes = { code }
end
return data_to_code
end))