Wiktionary

Module:User:Theknightwho/ru-translit

< Module:User:Theknightwho
This module lacks a documentation subpage. You may create it.
Useful links: root page • root page’s subpages • links • transclusions • testcases • user page • user talk page • userspace

This is a private module sandbox of Theknightwho, for their own experimentation. Items in this module may be added and removed at Theknightwho's discretion; do not rely on this module's stability. 

local export = {}

--[=[

FIXME:

1. (DONE) If you write '''Б'''ез, it transliterates to '''B'''jez instead of
   '''B'''ez, as it should.
2. (DONE) Convert ъ to nothing before comma or other non-letter particle, e.g.
   in Однимъ словомъ, идешь на чтеніе.
3. (DONE) Make special-casing for adjectives in -го and for что (and friends)
    be the default, and implement transformations in Cyrillic rather than after
    translit so that we can display the transformed Cyrillic in the
    "phonetic respelling" notation of {{ru-IPA}}.
4. (DONE) Convert apostrophe to ъ before transliteration when after a consonant
   and before a vowel (requested by Atitarev).
]=]

local u = mw.ustring.char
local explode = require("Module:string utilities").explode_utf8
local concat = table.concat
local insert = table.insert
local rfind = mw.ustring.find
local rsub = mw.ustring.gsub
local rsplit = mw.text.split
local toNFC = mw.ustring.toNFC
local toNFD = mw.ustring.toNFD
local decompose = require("Module:ru-common").decompose

local AC = u(0x301) -- acute = ́
local GR = u(0x0300) -- grave = ̀
local DI = u(0x0308) -- diaeresis = ̈
local TEMP_G = u(0xFFF1) -- substitute to preserve g from changing to v

local function ine(x) -- if not empty
	if x == "" then return nil else return x end
end

-- In this table, we now map Cyrillic е and э to je and e, and handle the
-- post-consonant version (plain e and ɛ) specially.
local tab = {
	["А"]="A", ["Б"]="B", ["В"]="V", ["Г"]="G", ["Д"]="D", ["Е"]="Je", ["Ж"]="Ž", ["З"]="Z", ["И"]="I", ["Й"]="J",
	["К"]="K", ["Л"]="L", ["М"]="M", ["Н"]="N", ["О"]="O", ["П"]="P", ["Р"]="R", ["С"]="S", ["Т"]="T", ["У"]="U", ["Ф"]="F",
	["Х"]="X", ["Ц"]="C", ["Ч"]="Č", ["Ш"]="Š", ["Щ"]="Šč", ["Ъ"]="ʺ", ["Ы"]="Y", ["Ь"]="ʹ", ["Э"]="E", ["Ю"]="Ju", ["Я"]="Ja",
	['а']='a', ['б']='b', ['в']='v', ['г']='g', ['д']='d', ['е']='je', ['ж']='ž', ['з']='z', ['и']='i', ['й']='j',
	['к']='k', ['л']='l', ['м']='m', ['н']='n', ['о']='o', ['п']='p', ['р']='r', ['с']='s', ['т']='t', ['у']='u', ['ф']='f',
	['х']='x', ['ц']='c', ['ч']='č', ['ш']='š', ['щ']='šč', ['ъ']='ʺ', ['ы']='y', ['ь']='ʹ', ['э']='e', ['ю']='ju', ['я']='ja',
	-- Russian style quotes
	['«']='“', ['»']='”',
	-- archaic, pre-1918 letters
	['І']='I', ['і']='i', ['Ѳ']='F', ['ѳ']='f',
	['Ѣ']='Jě', ['ѣ']='jě', ['Ѵ']='I', ['ѵ']='i',
	-- archaic, pre-1700 letters
	['Ѕ']='Z', ['ѕ']='z', ['Ꙃ']='Z', ['ꙃ']='z', ['Ꙁ']='Z', ['ꙁ']='z',
	['Ѡ']='O', ['ѡ']='o', ['Ѿ']='Ot', ['ѿ']='ot', ['Ꙋ']='U', ['ꙋ']='u',
	['Ꙑ']='Y', ['ꙑ']='y', ['Ꙗ']='Ja', ['ꙗ']='ja', ['Ѥ']='Je', ['ѥ']='je',
	['Ѧ']='Ja', ['ѧ']='ja', ['Ѩ']='Ja', ['ѩ']='ja', ['Ѫ']='U', ['ѫ']='u', ['Ѭ']='Ju', ['ѭ']='ju',
	['Ѯ']='Ks', ['ѯ']='ks', ['Ѱ']='Ps', ['ѱ']='ps', ['Є']='E', ['є']='e', ['Ї']='I', ['ї']='i',
}

local tab_jo = {
	["Ё"] = "Jo", ["Ѣ̈"] = "Jǒ", ["Я̈"] = "Jǫ", ["Ѧ̈"] = "Jǫ", ["Ѩ̈"] = "Jǫ",
	["ё"] = "jo", ["ѣ̈"] = "jǒ", ["я̈"] = "jǫ", ["ѧ̈"] = "jǫ", ["ѩ̈"] = "jǫ"
}

local plain_e = {
	["Е"] = "E", ["Ѣ"] = "Ě", ["Э"] = "Ɛ",
	["е"] = "e", ["ѣ"] = "ě", ["э"] = "ɛ"
}

local vowels = "аеєиіоуүꙋѡѿꙑыѣэюꙗяѥѧѫѩѭѵaæɐeəɛiɪɨoɵuyʊʉ"
local hushing_jo = "жчшщ"
local izhitsa_v = "аеєиіѣэꙗяѥѧѩaæɐeəɛiɪɨ"

-- Apply transformations to the Cyrillic to more closely match pronunciation.
-- Return two arguments: the "original" text (after decomposing composed
-- grave characters), and the transformed text. If the two are different,
-- {{ru-IPA}} should display a "phonetic respelling" notation. 
-- NOADJ disables special-casing for adjectives in -го, while FORCEADJ forces
-- special-casing for adjectives, including those in -аго (pre-reform spelling)
-- and disables checking for exceptions (e.g. много, ого). NOSHTO disables
-- special-casing for что and related words.
function export.apply_tr_fixes(text, noadj, noshto, forceadj)
	-- decompose stress accents without decomposing letters we want to treat
	-- as units (e.g. й or ё)
	text = decompose(text)

	local origtext = text
	-- the second half of the if-statement below is an optimization; see above.
	if not noadj and text:find("го") then
		local v = {["г"] = "в", ["Г"] = "В"}
		local repl = function(e, g, o, sja) return e .. v[g] .. o .. (sja or "") end
		-- Handle какого-нибудь/-либо/-то; must be done first because of an exception
		-- made for бого-, снего-, etc.
		text = rsub(text, "([кКтТ][аА][кК][оеОЕ" .. (forceadj and "аА" or "") .. "][" .. AC .. GR .. "]?)([гГ])([оО]%-)", repl)
		if not forceadj then
			local function go(text, case)
				local pattern = rsub(case, "^(.)(.*)(го[" .. AC .. GR .. "]?)(%-?)$", function(m1, m2, m3, m4)
					m1 = "%f[%a" .. AC .. GR .. "]([" .. mw.ustring.upper(m1) .. m1 .. "]"
					m2 = m2:gsub("\204[\128\129]", "[" .. AC .. GR .. "]?") .. ")"
					m3 = m3:gsub("\204[\128\129]", "[" .. AC .. GR .. "]?")
						:gsub("^г(.*)", "г(%1")
					m4 = m4 == "-" and "%-)" or ")%f[^%a" .. AC .. GR .. "]"
					return m1 .. m2 .. m3 .. m4
				end)
				return rsub(text, pattern, "%1" .. TEMP_G .. "%2")
			end
			for _, case in ipairs{"мно́го", "н[еа]мно́го", "до́рого", "недо́рого", "стро́го", "нестро́го", "на́строго", "убо́го", "пол[ао]́го"} do
				text = go(text, case)
			end
			-- check for neuter short forms of compound adjectives in -но́гий
			if rfind(text, "но[" .. AC .. GR .. "]?го%f[^%a" .. AC .. GR .. "]") then
				for _, case in ipairs{"безно́го", "босоно́го", "веслоно́го", "длинноно́го", "двуно́го", "коротконо́го", "кривоно́го", "одноно́го", "пятино́го", "трёхно́го", "трехно́го", "хромоно́го", "четвероно́го", "шестино́го"} do
					text = go(text, case)
				end
			end
			for _, case in ipairs{"ого́", "го́го", "ваго́го", "ло́го", "п[ео]́го", "со́го", "То́го", "ле́го", "игого́", "огого́", "альбиньязего", "д[иі]е́го", "бо́лого", "гр[иі]е́го", "манче́го", "пичис[иі]е́го", "тенкодого", "хио́го", "аго-", "его-", "ого-"} do
				text = go(text, case)
			end
		end
		--handle genitive/accusative endings, which are spelled -ого/-его/-аго
		-- (-ogo/-ego/-ago) but transliterated -ovo/-evo/-avo; only for adjectives
		-- and pronouns, excluding words like много, ого (-аго occurs in
		-- pre-reform spelling); \204\129 is an acute accent, \204\128 is a grave accent
		local pattern = "([оеОЕ" .. (forceadj and "аА" or "") .. "][" .. AC .. GR .. "]?)([гГ])([оО][" .. AC .. GR .. "]?)"
		local reflexive = "([сС][яЯ][" .. AC .. GR .. "]?)"
		text = rsub(text, pattern .. "%f[^%a" .. AC .. GR .. TEMP_G .. "]", repl)
		text = rsub(text, pattern .. reflexive .. "%f[^%a" .. AC .. GR .. TEMP_G .. "]", repl)
		-- handle сегодня
		text = rsub(text, "%f[%a" .. AC .. GR .. "]([Сс]е)г(о[" .. AC .. GR .. "]?дня)%f[^%a" .. AC .. GR .. "]", "%1в%2")
		-- handle сегодняшн-
		text = rsub(text, "%f[%a" .. AC .. GR .. "]([Сс]е)г(о[" .. AC .. GR .. "]?дняшн)", "%1в%2")
		-- replace TEMP_G with g; must be done after the -go -> -vo changes
		text = rsub(text, TEMP_G, "г")
	end

	-- the second half of the if-statement below is an optimization; see above.
	if not noshto and text:find("то") then
		local ch2sh = {["ч"] = "ш", ["Ч"] = "Ш"}
		-- Handle что
		text = rsub(text, "%f[%a" .. AC .. GR .. "]([Чч])(то[" .. AC .. GR .. "]?)%f[^%a" .. AC .. GR .. "]",
			function(ch, to) return ch2sh[ch] .. to end)
		-- Handle чтобы, чтоб
		text = rsub(text, "%f[%a" .. AC .. GR .. "]([Чч])(то[" .. AC .. GR .. "]?бы?)%f[^%a" .. AC .. GR .. "]",
			function(ch, to) return ch2sh[ch] .. to end)
		-- Handle ничто
		text = rsub(text, "%f[%a" .. AC .. GR .. "]([Нн]и)ч(то[" .. AC .. GR .. "]?)%f[^%a" .. AC .. GR .. "]", "%1ш%2")
	end

	-- Handle мягкий, лёгкий, легчать, etc.
	text = rsub(text, "([МмЛл][яеё][" .. AC .. GR .. "]?)г([кч])", "%1х%2")

	return origtext, text
end

do
	local function get_char(text, pattern, d, dir)
		local i, ch = 0
		repeat
			i = i + 1
			ch = text[d.i + (i * dir)]
		until not (ch and (pattern):find(ch, 1, true))
		return ch
	end
	
	local function do_iteration(output, text, d)
		-- Get current, previous and next characters, skipping over brackets, and
		-- ignoring diacritics for the previous character (which simplifies checks).
		local this = text[d.i]
		local this_lower = this:ulower()
		local prev = get_char(text, AC .. GR .. DI .. "%(", d, -1)
		local nxt = get_char(text, "%)", d, 1)
		-- A word is monosyllabic if it has only one vowel.
		if vowels:find(this_lower, 1, true) then
			d.vowels = d.vowels + 1
		end
		-- Convert ё (and archaic friends) here, since they need to be dealt with
		-- as a special-case. In general terms, ё becomes jo, ѣ̈ becomes jǒ and я̈,
		-- ѧ̈ & ѩ̈ become jǫ. When determining stress, they are treated
		-- as interchangeable.
		-- Jos do not implicitly take stress accents if an explicit primary stress is
		-- given. Otherwise, the final jo which doesn't have secondary stress takes
		-- primary stress.
		-- Prefixes do not take implicit primary stress.
		-- Primary stress will be shown on monosyllables if either they are a suffix
		-- or include_monosyllabic_jo_accent is true.
		if nxt == DI then
			d.i = d.i + 1
			this = toNFC(this .. DI)
			local t = tab_jo[this]
			if t then
				-- An initial lowercase j is removed from a jo transliteration
				-- if preceded by a hushing consonant (ж ч ш щ) (but not for a
				-- capital, on the assumption that a medial J is part of an
				-- initialism).
				if (
					t:sub(1, 1) == "j" and
					prev and hushing_jo:find(prev:ulower(), 1, true)
				) then
					t = t:sub(2)
				end
				insert(output, t)
				if text[d.i + 1] ~= GR then
					d.final_jo = #output
				end
				return
			end
		elseif this == AC then
			d.primary = true
		-- е after a consonant or a dash at the beginning of a word becomes e,
		-- and э becomes ɛ.
		elseif (this_lower == "е" or this_lower == "ѣ" or this_lower == "э") and (
			(prev and not (vowels .. "ъьʹʺ"):find(prev:ulower(), 1, true)) or
			(not prev and d.dash_before)
		) then
			insert(output, plain_e[this])
			return
		elseif this_lower == "ю" then
			local prev_lower = prev and prev:ulower()
			if prev_lower == "ж" or prev_lower == "ш" then
				insert(output, "u")
				return
			end
		-- Make izhitsa display as -v- after /a/, /e/ and /i/ (matching the equivalent
		-- Greek digraphs αυ, ευ and ηυ).
		elseif (
			this_lower == "ѵ" and
			prev and izhitsa_v:find(prev:ulower(), 1, true)
		) then
			this = this == "Ѵ" and "В" or "в"
			text[d.i] = this
		-- Convert apostrophe to the hard sign between consonant and vowel (i.e.
		-- in the places where the hard sign normally occurs in modern text).
		-- Apostrophe is sometimes used to indicate the hard sign; this may have
		-- originated from the forcible removal of the hard sign from printing
		-- offices in the 1920's, after the implementation of the Russian
		-- orthography reform.
		elseif (
			this == "'" and
			prev and not vowels:find(prev:ulower(), 1, true) and
			nxt and vowels:find(nxt, 1, true)
		) then
			text[d.i] = nxt and nxt:uupper() == nxt and "Ъ" or "ъ"
			this = text[d.i]
		-- Ignore word-final hard signs.
		elseif this_lower == "ъ" and d.i == #text then
			return
		end
		insert(output, tab[this] or this)
	end

	-- Transliterate after the pronunciation-related transformations of
	-- export.apply_tr_fixes() have been applied. Called from {{ru-IPA}}.
	-- INCLUDE_MONOSYLLABIC_JO_ACCENT is as in export.tr().
	function export.tr_after_fixes(text, include_monosyllabic_jo_accent)
		local word_chars = "%a'%(%)%[%]" .. AC .. GR .. DI
		local output = {}
		
		text = rsub(toNFC(text), "([^" .. word_chars .. "]*)([" .. word_chars .. "]*)", function(before, text)
			for _, ch in ipairs(explode(before)) do
				insert(output, ch)
			end
			text = explode(rsub(text, "[^Йй]", toNFD))
			local d = {
				i = 0,
				vowels = 0
			}
			if output[#output] == "-" then
				local prev = output[#output - 1]
				if not prev or rfind(prev, "%s") then
					d.dash_before = true
				end
			end
			while d.i < #text do
				d.i = d.i + 1
				do_iteration(output, text, d)
			end
			if (
				d.final_jo and
				(not (d.primary or text[#text] == "-")) and
				(include_monosyllabic_jo_accent or d.vowels > 1 or d.dash_before)
			) then
				output[d.final_jo] = output[d.final_jo] .. AC
			end
		end)
		
		return toNFC(concat(output))
	end
end

-- Transliterates text, which should be a single word or phrase. It should
-- include stress marks, which are then preserved in the transliteration.
-- ё is a special case: it is rendered (j)ó in multisyllabic words and
-- monosyllabic words in multi-word phrases, but rendered (j)o without an
-- accent in isolated monosyllabic words, unless INCLUDE_MONOSYLLABIC_JO_ACCENT
-- is specified. (This is used in conjugation and declension tables.)
-- NOADJ disables special-casing for adjectives in -го, while FORCEADJ forces
-- special-casing for adjectives and disables checking for exceptions
-- (e.g. много). NOSHTO disables special-casing for что and related words.
function export.tr(text, lang, sc, include_monosyllabic_jo_accent, noadj, noshto, forceadj)
	local origtext, subbed_text = export.apply_tr_fixes(text, noadj, noshto, forceadj)
	return export.tr_after_fixes(subbed_text, include_monosyllabic_jo_accent)
end

-- translit with various special-case substitutions; NOADJ disables
-- special-casing for adjectives in -го, while FORCEADJ forces special-casing
-- for adjectives and disables checking for expections (e.g. много).
-- NOSHTO disables special-casing for что and related words. SUB is used
-- to implement arbitrary substitutions in the Cyrillic text before other
-- transformations are applied and before translit. It is of the form
-- FROM/TO,FROM/TO,...
function export.tr_sub(text, include_monosyllabic_jo_accent, noadj, noshto, sub,
	forceadj)
	if type(text) == 'table' then -- called directly from a template
		include_monosyllabic_jo_accent = ine(text.args.include_monosyllabic_jo_accent)
		noadj = ine(text.args.noadj)
		noshto = ine(text.args.noshto)
		sub = ine(text.args.sub)
		text = text.args[1]
	end

	if sub then
		local subs = rsplit(sub, ",")
		for _, subpair in ipairs(subs) do
			local subsplit = rsplit(subpair, "/")
			text = rsub(text, subsplit[1], subsplit[2])
		end
	end

	return export.tr(text, nil, nil, include_monosyllabic_jo_accent, noadj, noshto, forceadj)
end

--for adjectives, pronouns
function export.tr_adj(text, include_monosyllabic_jo_accent)
	if type(text) == 'table' then -- called directly from a template
		include_monosyllabic_jo_accent = ine(text.args.include_monosyllabic_jo_accent)
		text = text.args[1]
	end

	-- we have to include "forceadj" because typically when tr_adj() is called
	-- from the noun or adjective modules, it's called with suffix ого, which
	-- would otherwise trigger the exceptional case and be transliterated as ogo
	return export.tr(text, nil, nil, include_monosyllabic_jo_accent, false,
		"noshto", "forceadj")
end

return export

-- For Vim, so we get 4-space tabs
-- vim: set ts=4 sw=4 noet:
Retrieved from "https://en.wiktionary.org/w/index.php?title=Module:User:Theknightwho/ru-translit&oldid=77608771"