Module:User:Surjection/fi-pron-inflect

-- fi_pron_inflect: Generates fi-pronunciation inputs for inflected forms.
-- 
-- Inputs:
--     lemma (string): The spelling of the lemma.
--     lemma_pron (string): The fi-pronunciation input for the lemma. May be an empty string or nil for no input, or shorthand * or (*).
--     infl_class (number): The inflection class: 1-49 for nominals (see ]), 52-78 for verbs (see ]).
--     grad_class (string | nil): The gradation class: nil for no gradation, or a letter A-M (see ]).
--     form_tag (string): The tag for the form; see the table below.
--     form (string): The spelling of the inflected form.
-- 
-- Outputs:
--     The fi-p input for the inflected form. May be shorthand, e.g. empty string
--     for no input, * for an input with only *, etc.
-- 
--     Note that the output may change based only on the form_tag, so the
--     pronunciation for each value must be checked separately and then merged
--     as needed.
-- 
--     Returns nil only if no input could be determined.

--[[

Allowed values for form_tag:
    Nominals:
        nom_sg:         nominative singular
        nom_pl:         nominative plural
        gen_sg:         genitive singular
        gen_pl:         genitive plural
        par_sg:         partitive singular
        par_pl:         partitive plural
        ine_sg:         inessive singular
        ine_pl:         inessive plural
        ela_sg:         elative singular
        ela_pl:         elative plural
        ill_sg:         illative singular
        ill_pl:         illative plural
        ade_sg:         adessive singular
        ade_pl:         adessive plural
        abl_sg:         ablative singular
        abl_pl:         ablative plural
        all_sg:         allative singular
        all_pl:         allative plural
        ess_sg:         essive singular
        ess_pl:         essive plural
        tra_sg:         translative singular
        tra_pl:         translative plural
        abe_sg:         abessive singular
        abe_pl:         abessive plural
        ins_sg:         instructive singular
        ins_pl:         instructive plural
        com_pl:         comitative plural

        Any of these can be suffixed with /1s, /2s, /1p, /2p, /3,
        for possessive forms (first/second/third-person, singular/plural).

    Verbs:
        pres_1sg:       indicative present, first-person singular
        pres_2sg:       indicative present, second-person singular
        pres_3sg:       indicative present, third-person singular
        pres_1pl:       indicative present, first-person plural
        pres_2pl:       indicative present, second-person plural
        pres_3pl:       indicative present, third-person plural
        pres_conn:      indicative present, non-passive connegative
        pres_pasv:      indicative present, passive
        pres_pasv_conn: indicative present, passive connegative
        past_1sg:       indicative past, first-person singular
        past_2sg:       indicative past, second-person singular
        past_3sg:       indicative past, third-person singular
        past_1pl:       indicative past, first-person plural
        past_2pl:       indicative past, second-person plural
        past_3pl:       indicative past, third-person plural
        past_pasv:      indicative past, passive
        cond_1sg:       conditional, first-person singular
        cond_2sg:       conditional, second-person singular
        cond_3sg:       conditional, third-person singular
        cond_1pl:       conditional, first-person plural
        cond_2pl:       conditional, second-person plural
        cond_3pl:       conditional, third-person plural
        cond_conn:      conditional, non-passive connegative
        cond_pasv:      conditional, passive
        cond_pasv_conn: conditional, passive connegative
        impr_2sg:       imperative, second-person singular
        impr_3sg:       imperative, third-person singular
        impr_1pl:       imperative, first-person plural
        impr_2pl:       imperative, second-person plural
        impr_3pl:       imperative, third-person plural
        impr_conn:      imperative, non-passive connegative
        impr_pasv:      imperative, passive
        impr_pasv_conn: imperative, passive connegative
        potn_1sg:       potential, first-person singular
        potn_2sg:       potential, second-person singular
        potn_3sg:       potential, third-person singular
        potn_1pl:       potential, first-person plural
        potn_2pl:       potential, second-person plural
        potn_3pl:       potential, third-person plural
        potn_conn:      potential, non-passive connegative
        potn_pasv:      potential, passive
        potn_pasv_conn: potential, passive connegative
        inf1:           first infinitive (lemma for verbs)
        inf1_long/1s:   long first infinitive, first-person singular possessive
        inf1_long/2s:   long first infinitive, second-person singular possessive
        inf1_long/1p:   long first infinitive, first-person plural possessive
        inf1_long/2p:   long first infinitive, second-person plural possessive
        inf1_long/3:    long first infinitive, third-person singular
        inf2_ine:       second active infinitive, inessive
        inf2_ine/1s:    second active infinitive, first-person singular possessive
        inf2_ine/2s:    second active infinitive, second-person singular possessive
        inf2_ine/1p:    second active infinitive, first-person plural possessive
        inf2_ine/2p:    second active infinitive, second-person plural possessive
        inf2_ine/3:     second active infinitive, third-person singular
        inf2_ins:       second active infinitive, instructive
        inf2_pasv_ine:  second passive infinitive, inessive
        inf3_ine:       third active infinitive, inessive
        inf3_ela:       third active infinitive, elative
        inf3_ill:       third active infinitive, illative
        inf3_ade:       third active infinitive, adessive
        inf3_abe:       third active infinitive, abessive
        inf3_ins:       third active infinitive, instructive
        inf3_pasv_ins:  third passive infinitive, instructive
        inf4:           -minen verbal noun
        inf5/1s:        fifth infinitive, first-person singular possessive
        inf5/2s:        fifth infinitive, second-person singular possessive
        inf5/1p:        fifth infinitive, first-person plural possessive
        inf5/2p:        fifth infinitive, second-person plural possessive
        inf5/3:         fifth infinitive, third-person singular
	    pres_part:      present active participle
	    pres_pasv_part: present passive participle
	    past_part:      past active participle
	    past_part_pl:   past active participle, plural form
	    past_pasv_part: past passive participle
	    agnt_part:      agent participle
	    nega_part:      negative participle

]]--

local export = {}
local consonants = "bdfghjklmnprsštvxzž"
local vowels = "aeiouyäö"

-- inflected form ends in * (1) or (*) (2)
local infl_finalgem = {
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 1,
	 = 2,
}

-- nominal forms with weak grade
local infl_grad_weak = {
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
}

-- infl classes with reverse gradation
local infl_grad_reverse = {
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
}

-- nominal forms have weak grade

-- infl class 69: ts is weak
local infl69_ts_weak = {
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
	 = true,
}

-- Split form_tag into primary form and possessive suffix.
local function split_tag_decl(form_tag)
	if form_tag:find("/") then
		return mw.ustring.match(form_tag, "(*)/(.*)")
	else
		return form_tag, nil
	end
end

-- Wye: for the lemma and the inflected form, returns the common prefix
-- and the two suffixes.
local function wye(lemma, form)
	local n = 1
	while mw.ustring.sub(lemma, n, n) == mw.ustring.sub(form, n, n) do
		n = n + 1
	end
	return mw.ustring.sub(lemma, 1, n - 1), mw.ustring.sub(lemma, n), mw.ustring.sub(form, n)
end

local unravel_spelling_ignore = {
	 = true,  = true,  = true,  = true
}

-- Unravel fi-p to remove a suffix. Returns nil in case of error.
local function pron_unravel(pron, suf)
	local n = mw.ustring.len(suf)
	while true do
		-- always remove these before
		if mw.ustring.match(pron, "$") then
			pron = mw.ustring.gsub(pron, "%(%*%)$", "")
			pron = mw.ustring.gsub(pron, "%*$", "")
		end

		if n == 0 then
			break
		end

		local c = mw.ustring.sub(suf, n, n)

		if unravel_spelling_ignore then
			-- do nothing more
		elseif mw.ustring.match(c, "") then
			if c == "d" then
				-- either D gets removed
				mask = ""
			else
				mask = c
			end
			if mw.ustring.match(pron, mask .. "$") then
				pron = mw.ustring.sub(pron, 1, -2)
			else
				return nil -- unexpected character
			end
			if c == "s" then
				-- remove ts specs
				pron = mw.ustring.gsub(pron, "$", "")
			end
		else
			return nil -- unsupported character
		end
		
		n = n - 1

		-- always remove these after
		if mw.ustring.match(pron, "$") then
			pron = mw.ustring.gsub(pron, "%(%.%)$", "")
			pron = mw.ustring.gsub(pron, "%.$", "")
		end
	end
	return pron
end

-- Wye fi-p: given the fi-p for the lemma, the lemma and the inflected form,
-- return the common fi-p prefix, lemma suffix and inflected form suffix.
-- The common prefix is nil in case of error.
local function pron_wye(lemma_pron, lemma, form)
	local common, lemma_suf, form_suf = wye(lemma, form)
	local prefix = pron_unravel(lemma_pron, lemma_suf)

	-- add an <i> if the spelling ends in a vowel and the pronunciation does not (unadapted loanwords)
	if prefix and lemma_suf == "" and mw.ustring.match(prefix, "$") and mw.ustring.match(common, "$") then
		prefix = prefix .. "i"
	end

	return prefix, lemma_suf, form_suf
end

local diphthongs_pure = {
	 = true,  = true,                 = true,
	 = true,  = true,  = true,  = true,
	 = true,  = true,  = true,  = true,
	 = true,  = true,  = true,  = true,
	                = true,  = true,  = true,
}

local function ts_fix(pron, weak_grade)
	if weak_grade then
		-- final ts > t!s
		return mw.ustring.gsub(pron, "ts(*)$", "t!s%1")
	else
		-- final t!s > ts
		return mw.ustring.gsub(pron, "t%!s(*)$", "ts%1")
	end
end

-- adds/removes syllable breaks
local function dotter(pron, grad_class, lemma_suf, form_suf)
	if not mw.ustring.match(form_suf, "^") then
		-- remove trailing dot if the suffix does not start with a vowel
		pron = mw.ustring.gsub(pron, "%(%.%)$", "")
	end
	if grad_class == "D" and mw.ustring.match(lemma_suf, "^k") and not mw.ustring.match(form_suf, "^k") then
		-- may need to add trailing dot if we have a diphthong
		local m_fi_p = require("Module:fi-pronunciation")
		local partial = pron .. mw.ustring.sub(form_suf, 1, 1)
		local syl = m_fi_p.syllabify(partial)
		local syl_last = syl

		if diphthongs_pure then
			pron = pron .. "(.)"
		end
	end
	return pron
end

local function reapply(prefix, suffix, ts_weak)
	suffix = mw.ustring.gsub(suffix, "d", "D")
	if ts_weak then suffix = mw.ustring.gsub(suffix, "ts", "t!s") end
	return prefix .. suffix
end

function export.fi_pron_inflect(lemma, lemma_pron, infl_class, grad_class, form_tag, form)
	local infl, poss = split_tag_decl(form_tag)

	-- undo shorthand
	lemma_pron = lemma_pron or ""
	if mw.ustring.match(lemma_pron, "^%(?%*?%)?$") then lemma_pron = lemma .. lemma_pron end

	-- wye split
	local pron, lemma_suf, form_suf = pron_wye(lemma_pron, lemma, form)
	if pron == nil then return nil end

	-- nom_sg and nom_pl are gen_sg with possessive
	if poss and infl:match("^nom_") then
		infl = "gen_sg"
	end
	
	if lemma_pron:match("%]$") then
		error("syllabification hints must currently be dealt with manually")
	end
	
	-- ts fixes for nominals
	if 1 <= infl_class and infl_class <= 49 and mw.ustring.find(lemma, "ts") then
		local reverse_gradation = not not infl_grad_reverse
		local grade_weak = not not infl_grad_weak
		-- strong grade for nom/gen_sg with possessive
		if poss and infl == "gen_sg" then grade_weak = false end
		grade_weak = grade_weak ~= reverse_gradation
		pron = ts_fix(pron, grade_weak)
	end

	-- apply form_suf to pron
	local final = reapply(dotter(pron, grad_class, lemma_suf, form_suf), form_suf, infl_class == 69 and infl69_ts_weak)
	if final == form then
		-- shorthand
		final = ""
	end

	if poss == "3" and mw.ustring.match(form, "ns$") then
		final = final .. "*"
	else
		local finalgem = infl_finalgem
		if finalgem == 1 then
			final = final .. "*"
		elseif finalgem == 2 then
			final = final .. "(*)"
		end
	end

	return final
end

return export