ข้ามไปเนื้อหา

มอดูล:en-utilities

จาก วิกิพจนานุกรม พจนานุกรมเสรี

local export = {}

local add_suffix -- Defined below.
local find = string.find
local match = string.match
local reverse = string.reverse
local sub = string.sub
local toNFD = mw.ustring.toNFD
local ugsub = mw.ustring.gsub
local ulower = mw.ustring.lower
local umatch = mw.ustring.match
local usub = mw.ustring.sub

local vowels = "aæᴀᴁɐɑɒ@eᴇǝⱻəɛɘɜɞɤiıɪɨᵻoøœᴏɶɔᴐɵuᴜʉᵾɯꟺʊʋʌyʏ"
local hyphens = "%-‐‑‒–—"
local diacritics

local function get_diacritics()
	diacritics = mw.loadData("Module:headword/data").page.comb_chars.diacritics_all .. "+"
	return diacritics
end

-- Normalize a string, so that case and diacritics are ignored. By default,
-- "gu" and "qu" are normalized to "g" and "q", because they behave like
-- consonants under certain conditions (e.g. final "y" does not usually have
-- the plural "ies" after a vowel, but it's regular for "quy" to become "quies".
-- The option `not_gu` prevents this happening to "gu", and is needed because
-- terms ending "-guy" are almost always compounds of "guy" (→ "guys").
local function normalize(str, followed_by, not_gu)
	if not followed_by then
		followed_by = ""
	end
	str = ugsub(toNFD(str) .. followed_by, "([" .. (not_gu and "" or "Gg") .. "Qq])u([".. vowels .. "])", "%1%2")
	return ulower(ugsub(sub(str, 1, #str - #followed_by), diacritics or get_diacritics(), ""))
end

local function epenthetic_e_default(stem)
	return sub(stem, -1) ~= "e"
end

local function epenthetic_e_for_s(stem, term)
	-- If the stem is different, it must be from "y" → "i".
	if stem ~= term then
		return true
	end
	local final
	if match(stem, "^[^\128-\255]*$") then
		final = sub(stem, -1)
	else
		stem = ugsub(toNFD(stem), diacritics or get_diacritics(), "")
		final = usub(stem, -1)
	end
	-- Epenthetic "e" is added after a sibilant or sibilant-affricate. The vast
	-- majority of these are spelled "s", "x", "z", "ch" and "sh", but "dg"
	-- (→ "dge") and "ß" (→ "ss") can be found in obsolete spellings, "shh" in
	-- onomatopoeia, and "zh", "dj", "jj" (and more) in loanwords.
	return (
		final == "g" and sub(stem, -2, -2) == "d" or
		final == "h" and match(stem, "[csz]h+$") or
		final == "j" and umatch(stem, "[^" .. vowels .. "]j$") or
		final == "s" or
		final == "x" or
		final == "z" or
		final == "ß"
	)
end

local suffixes = {}

-- FIXME, a subrule of rule #1 below says the -ies ending doesn't
-- apply to proper nouns, hence "the Gettys", "the public Ivys".
-- We should maybe consider applying this rule here; but it may not
-- be important as this function is almost always called on common nouns
-- (e.g. parts of speech, place types).
suffixes["s.plural"] = {
	final_y_is_i = true,
	epenthetic_e = epenthetic_e_for_s
}

suffixes["s.verb"] = {
	final_y_is_i = true,
	final_consonant_is_doubled = true,
	epenthetic_e = epenthetic_e_for_s
}

suffixes["ing"] = {
	final_consonant_is_doubled = true,
	remove_silent_e = true,
}

suffixes["d"] = {
	final_y_is_i = true,
	final_consonant_is_doubled = true,
	epenthetic_e = epenthetic_e_default,
}

suffixes["dst"] = suffixes["d"]
suffixes["st.verb"] = suffixes["d"]
suffixes["th"] = suffixes["d"]

suffixes["n"] = {
	final_y_is_i = true,
	final_consonant_is_doubled = true,
	epenthetic_e = function(stem)
		local final = sub(stem, -1)
		return not (
			final == "e" or
			final == "w" or
			final == "r" and umatch(usub(normalize(stem), -2, -2), "[" .. vowels .. "]")
		)
	end,
}

suffixes["r"] = {
	final_y_is_i = true,
	final_ey_is_i = true,
	final_guy_is_gui = true,
	final_consonant_is_doubled = true,
	epenthetic_e = epenthetic_e_default
}

suffixes["st.superlative"] = suffixes["r"]

-- Returns the stem used for suffixes that sometimes convert final "y"
-- into "i", such as "-es" ("-ies"), e.g. "penny" → "penni" ("pennies").
-- If `final_ey_is_i` is true, final "ey" may also be converted, e.g.
-- "plaguey" → "plagui"; this is needed for "-er" ("-ier") and "-est"
-- ("-iest"). If `not_gu` is true, then normalize() will be called with the
-- `not_gu` flag (see there for more info); this is true in most cases.
local function convert_final_y_to_i(str, not_gu, final_ey_is_i)
	local final2 = usub(str, -2)
	-- If `final_ey_is_i` is true, treat final "-ey" can also be reduced.
	if final_ey_is_i and final2 == "ey" then
		local final3 = usub(str, -3)
		-- Special case: treat "eey" as "ee" + "y" (e.g. "treey" →
		-- "treeiest"). This doesn't apply if `final_ey_is_i` is not true, as
		-- `final_ey_is_i` assumes final "(e)y" is likely to be syllabic.
		-- This sometimes applies to "oey" and "uey" etc., but those are more
		-- likely to be vowel + "ey"; however, both types exist (compare "go" →
		-- "goey" → "goier" and "doe" → "doey" → "doeier"; "flu" → "fluey" →
		-- "fluiest" and "flue" → "fluey" → "flueiest" form an unlikely-but-
		-- possible minimal pair).
		if final3 == "eey" then -- FIXME: should this apply if final_ey_is_i is false?
			-- Replace final "y" with "i".
			return sub(str, 1, -2) .. "i"
		end
		-- Remove "ey" to get the base stem.
		local base_stem = sub(str, 1, -3)
		-- Special case: allow final "-ey" ("potato-ey" → "potato-iest").
		if umatch(final3, "[" .. hyphens .. "]ey") then
			return base_stem .. "i"
		end
		-- Final "ey" becomes "i" iff the term is polysyllabic (e.g. not
		-- "grey"). "ey" is common if the base stem ends in a vowel
		-- ("echo" → "echoey"), so the presence of a vowel anywhere in the
		-- base stem is sufficient to deem it polysyllabic. ("echoey" →
		-- "echo" → "echoiest", "beigey" → "beig" → "beigiest", but
		-- "grey" → "gr" → "greyest"). The first "y" in "-yey" can be
		-- treated as a vowel as long as it's preceded by something
		-- ("clayey" → "clay" → "clayiest", "cryey" → "cry" → "cryiest"),
		-- so it needs to be treated as a special case.
		local normalized = normalize(base_stem, "ey")
		if sub(normalized, -1) == "y" then
			if umatch(normalized, "[%w@][yY]$") then
				return base_stem .. "i"
			end
		elseif umatch(normalized, "[" .. vowels .. "%d]%w*$") then
			return base_stem .. "i"
		end
	-- Special cases: match final "quy" ("soliloquy" → "soliloquies"), and "-y"
	-- ("bro-y" → "bro-iest"), accounting for hyphen variation.
	elseif umatch(final2, "[" .. hyphens .. "]y") then
		-- Replace final "y" with "i".
		return sub(str, 1, -2) .. "i"
	-- Otherwise, final "y" becomes "i" iff it's not preceded by a vowel
	-- ("shy" → "shiest", "horsy" → "horsies", but "day" → "days", "coy" →
	-- "coyest").
	else
		-- Remove "y" to get the base stem.
		local base_stem = sub(str, 1, -2)
		if umatch(normalize(base_stem, "y", not_gu), "[^%s%p" .. vowels .. "]$") then
			return base_stem .. "i"
		end
	end
	return str
end

local function double_final_consonant(str, final)
	local initial = umatch(normalize(sub(str, 1, -2), final), "^.*%f[^%z%s" .. hyphens .. "…]([%l%p]*)[" .. vowels .. "]$")
	return initial and (
		initial == "" or
		initial == "y" or
		match(initial, "^.[\128-\191]*$") and umatch(initial, "[^" .. vowels .. "]") or
		umatch(initial, "^[^" .. vowels .. "]*%f[^%l]$")
	) and (str .. final) or str
end

local function remove_silent_e(str)
	local final2 = sub(str, -2)
	if final2 == "ie" then
		-- Replace "ie" with "y".
		return sub(str, 1, -3) .. "y"
	end
	local base_stem = sub(str, 1, -2)
	-- Silent "e" occurs after "u" or a consonant (cluster) preceded by a vowel.
	return (
		final2 == "ue" or
		umatch(normalize(base_stem, "e"), "[" .. vowels .. "][^" .. vowels .. "]+$")
	) and base_stem or str
end

function export.add_suffix(term, suffix)
	local data = suffixes[suffix]
	suffix = match(suffix, "^([^.]*)")
	local final, stem = sub(term, -1)
	if data.final_y_is_i and final == "y" then
		stem = convert_final_y_to_i(term, not data.final_guy_is_gui, data.final_ey_is_i)
	elseif data.remove_silent_e and final == "e" then
		stem = remove_silent_e(term)
	else
		stem = term
	end
	local epenthetic_e = data.epenthetic_e
	if epenthetic_e and epenthetic_e(stem, term) then
		suffix = "e" .. suffix
	end
	if (
		data.final_consonant_is_doubled and
		match(final, "^[bcdfgjklmnpqrstvz]$") and -- Only double regular consonants.
		umatch(suffix, "^[" .. vowels .. "]")
	) then
		stem = double_final_consonant(term, final)
	end
	return stem .. suffix
end
add_suffix = export.add_suffix

--[==[
Pluralize a word in a smart fashion, according to normal English rules.
# If the word ends in a consonant or "qu" + "-y", replace "-y" with "-ies".
# If the word ends in "s", "x", "z", "ch", "sh" or "zh", add "-es".
# Otherwise, add "-s".

This handles links correctly:
# If a piped link, change the second part appropriately.
# If a non-piped link and rule #1 above applies, convert to a piped link with the second part containing the plural.
# If a non-piped link and rules #2 or #3 above apply, add the plural outside the link.
]==]
function export.pluralize(str)
	-- Treat as a link if a "[[" is present and the string ends with "]]".
	if not (find(str, "[[", 1, true) and sub(str, -2) == "]]") then
		return add_suffix(str, "s.plural")
	end
	-- Find the last "[[" (in case there is more than one) by reversing
	-- the string.
	local str_rev = reverse(str)
	local open = find(str_rev, "[[", 3, true)
	-- If the last "[[" is followed by a "]]" which isn't at the end,
	-- then the final "]]" is just plaintext (e.g. "[[foo]]bar]]").
	local bad_close = find(str_rev, "]]", 3, true)
	-- Note: the bad "]]" will have a lower index than the last "[[" in
	-- the reversed string.
	if bad_close and bad_close < open then
		return add_suffix(str, "s.plural")
	end
	open = #str - open + 2
	-- Get the target and display text by searching from just after "[[".
	local target, display = match(str, "([^|]*)|?(.*)%]%]$", open)
	display = add_suffix(display ~= "" and display or target, "s.plural")
	-- If the link target is a substring of the display text, then
	-- use a trail (e.g. "[[foo]]" → "[[foo]]s", since "foo" is a substring
	-- of "foos").
	local index, trail = find(display, target, 1, true)
	if index == 1 then
		return sub(str, 1, open - 1) .. target .. "]]" .. sub(display, trail + 1)
	end
	-- Otherwise, return a piped link.
	return sub(str, 1, open - 1) .. target .. "|" .. display .. "]]"
end

-- Returns true if `plural` is an expected, regular plural of `term`.
function export.is_regular_plural(plural, term)
	-- Ignore any final punctuation that occurs in both forms, which is common
	-- in abbreviations (e.g. "abbr." → "abbrs.").
	local final_punc = umatch(term, "%p*$")
	local final_punc_len = #final_punc
	if sub(plural, -final_punc_len) == final_punc then
		term = sub(term, 1, -final_punc_len - 1)
		plural = sub(plural, 1, -final_punc_len - 1)
	end
	if plural == term .. "s" or plural == add_suffix(term, "s.plural") then
		return true
	end
	local final = sub(term, -1)
	return (
		-- Doubled final consonants in "s" and "z".
		final == "s" and plural == term .. "ses" or -- e.g. "busses"
		final == "z" and plural == term .. "zes" or -- e.g. "quizzes"
		-- convert_final_y_to_i() without the `not_gu` flag set, to catch
		-- "-guy" → "-guies", but not "day" → "daies".
		final == "y" and plural == convert_final_y_to_i(term) .. "es" or
		-- Capitalized terms like "$DEITY" → "$DEITIES (should we treat this as regular?)
		final == "Y" and ulower(plural) == convert_final_y_to_i(ulower(term)) .. "es"
	)
end

return export