Open main menu

Wiktionary β

Module:zh-han

This is used by {{Han char}} and {{Han simp}}. {{Han char}} is used for headwords in Translingual entries on Han script characters, while {{Han simp}} is used in Etymology or Glyph origin sections of Translingual or Chinese entries.


local export = {}

local PAGENAME = mw.title.getCurrentTitle().text

local cangjie = {
	['A']='日',['B']='月',['C']='金',['D']='木',['E']='水',['F']='火',['G']='土',
	['H']='竹',['I']='戈',['J']='十',['K']='大',['L']='中',['M']='一',['N']='弓',
	['O']='人',['P']='心',['Q']='手',['R']='口',
	['S']='尸',['T']='廿',['U']='山',['V']='女',['W']='田',['X']='X',['Y']='卜',['Z']='Z'
}

local radicals = {
	"一","丨","丶","丿","乙","亅",
	"二","亠","人","儿","入","八","冂","冖","冫","几","凵","刀","力","勹","匕","匚","匸","十","卜","卩","厂","厶","又",
	"口","囗","土","士","夂","夊","夕","大","女","子","宀","寸","小","尢","尸","屮","山","巛","工","己","巾","干","幺","广","廴","廾","弋","弓","彐","彡","彳",
	"心","戈","戶","手","支","攴","文","斗","斤","方","无","日","曰","月","木","欠","止","歹","殳","毋","比","毛","氏","气","水","火","爪","父","爻","爿","片","牙","牛","犬",
	"玄","玉","瓜","瓦","甘","生","用","田","疋","疒","癶","白","皮","皿","目","矛","矢","石","示","禸","禾","穴","立",
	"竹","米","糸","缶","网","羊","羽","老","而","耒","耳","聿","肉","臣","自","至","臼","舌","舛","舟","艮","色","艸","虍","虫","血","行","衣","西",
	"見","角","言","谷","豆","豕","豸","貝","赤","走","足","身","車","辛","辰","辵","邑","酉","釆","里",
	"金","長","門","阜","隶","隹","雨","靑","非",
	"面","革","韋","韭","音","頁","風","飛","食","首","香",
	"馬","骨","高","髟","鬥","鬯","鬲","鬼",
	"魚","鳥","鹵","鹿","麥","麻",
	"黃","黍","黑","黹",
	"黽","鼎","鼓","鼠",
	"鼻","齊",
	"齒",
	"龍","龜",
	"龠"
}

local simplified_radical = {
	['讠']='言',['门']='門',['饣']='食',['飞']='飛',['马']='馬',['见']='見',['贝']='貝',['纟']='糸',['车']='車',['长']='長',['韦']='韋',['风']='風',['钅']='金',['鸟']='鳥',['竜']='龍',['龙']='龍',['页']='頁',['斉']='齊',['齐']='齊',['麦']='麥',['亀']='龜',['龟']='龜',['鱼']='魚',['黾']='黽',['鼡']='鼠',['歯']='齒',['齿']='齒',['卤']='鹵'
}

local function mul_link(term)
	return require('Module:links').full_link({lang = require('Module:languages').getByCode('mul'), term = term})
end

local function radical(name,number,variant)
	number = number or '1'
	local len = mw.ustring.len(number)
	local padleft = (len == 2 and '0' .. number or (len == 1 and '00' .. number or number))
	local text = '<div class="noprint" style="border:solid #aaa 1px;margin:0;font-size:74%;background:#f9f3f6;width:120px;padding:3px;spacing:0;text-align:center;float:right"><div style="float:left">[[Image:Commons-logo.svg|24px|none|link=Commons:Category:Radical ' .. padleft .. ']]</div>[[Commons:Category:Radical ' .. padleft .. "-0|See images of <br>'''Radical " .. number .. " <big>" .. ((name and name ~= '') and PAGENAME or '※') .. "</big>''']]</div>"
	if name == PAGENAME then
		text = text .. '[[Category:Han character radicals| ]]'
	else
		if variant and variant ~= '' then
			text = text .. '[[Category:Han character radicals| ]]'
		else
			text = text .. '[[Category:Han char without var| ]]'
		end
	end
	return text
end

function export.simp(frame)
	local args = frame:getParent().args
	
	local trad = args[1] or ''
	if trad == '' then
		trad = mw.loadData("Module:zh/data/st").st[PAGENAME] or PAGENAME
	end
	
	local nocap = args['nocap'] or ''
	local alt = args['a'] or ''
	local from = {args['f'] or '',args['f2'] or ''}
	local to = {args['t'] or '',args['t2'] or ''}
	if from[1] == '' and to[1] ~= '' then from[1] = trad end
	local text = {}
	local result = ''
	result = (nocap == '' and 'S' or 's') .. 'implified from ' .. mul_link(trad)
	if alt ~= '' then
		result = result .. ' and ' .. mul_link(alt)
	end
	for i=1,2 do
		if from[i] ~= '' then
			if to[i] == '' then
				table.insert(text,mul_link(from[i]) .. " → a component which cannot be displayed independently[[Category:Kenny's testing category 3]]")
			elseif to[i] == ' ' or to[i] == '&nbsp;' then
				table.insert(text,'elimination of ' .. mul_link(from[i]))
			else
				table.insert(text,mul_link(from[i]) .. ' → ' .. mul_link(to[i]))
			end
		end
	end
	if #text > 0 then
		result = result .. ' (' .. table.concat(text,' and '):gsub('and elimination of','and') .. ')'
	end
	return result
end

function export.character(frame)
	local args = frame:getParent().args
	if radicals[tonumber(args['rn'])] ~= (simplified_radical[args['rad']] or args['rad']) then
		error('Radical number (rn) and radical (rad) do not match')
	end
	local text = {}
	local categories = {}
	local as = tostring(tonumber(args['as'] or '-1'))
	if as == '0' then
		table.insert(text, radical(args['rad'], args['rn'], (args['var'] or '')))
	end
	
	-- Creating headword with [[Module:headword]] ensures that page title is script-tagged with Hani (generic Han).
	-- This assumes that {{Han char}} is only used in Translingual sections.
	local mul = require("Module:languages").getByCode("mul")
	local Hani = require("Module:scripts").getByCode("Hani")
	local sort_key = args['rad'] .. args['as'] .. PAGENAME
	table.insert(
		text, 
		require("Module:headword").full_headword{
			lang = mul,
			sc = Hani,
			heads = { args['head'] or PAGENAME },
			categories = {},
			pos_category = "symbols",
			sort_key = sort_key,
		}
	)
	table.insert(text, " (''radical " .. (args['rn'] or '') .. "'' [[Index:Chinese radical/" .. (args['rad'] or '') .. '|' .. (args['rad'] or ' ') .. ']]+' .. as)
	if args['asj'] and args['asj'] ~= '' then
		table.insert(text, " ''in Chinese'', " .. (args['rad'] or '') .. '+' .. tonumber(args['asj']) .. " ''in Japanese''")
	end
	table.insert(text, ", ")
	if args['snj'] and args['snj'] ~= '' then
		table.insert(text, "''in Chinese'' ")
	end
	table.insert(text, (args['sn'] or '') .. " ''stroke")
	if not (args['sn'] and args['sn'] == '1') then
		table.insert(text, 's')
	end
	table.insert(text, "''")
	if args['snj'] and args['snj'] ~= '' then
		table.insert(text, ", ''in Japanese'' " .. args['snj'] .. " ''stroke")
		if not (args['snj'] and args['snj'] == '1') then
			table.insert(text, 's')
		end
		table.insert(text, "''")
	end
	if args['canj'] and args['canj'] ~= '' then
		local canj = args['canj']
		local canj_split = mw.text.split(canj,',')
		for i = 1, #canj_split do
			if mw.ustring.match(canj_split[i],'^[A-Z]+$') then --confirmed UTF-8, use string library instead of ustring
				canj_split[i] = canj_split[i]:gsub('[%(%)]','')
				canj_split[i] = canj_split[i]:gsub('[A-Z]',cangjie) .. ' ([[Index:Chinese Cangjie/' .. canj_split[i]:sub(1,1):lower() .. '|' .. canj_split[i] .. ']])'
			else
				table.insert(categories, "[[Category:Han char unsimplified usage]]")
			end
		end
		table.insert(text, ", ''[[Index:Chinese Cangjie|cangjie input]]'' " .. table.concat(canj_split,' or '))
		if #canj_split > 1 then
			table.insert(categories, "[[Category:Han char with multiple canj]]")
		end
	end
	if args['four'] and args['four'] ~= '' then
		local four = args['four']
		local four_split = mw.text.split(four,',')
		for i=1,#four_split do
			if not four_split[i]:match('[<>]') then
				four_split[i] = four_split[i]:gsub('(%d)(%d%d%d)%.?(%d?)','[[Index:Chinese four corner/%1|%1%2<sub>%3</sub>]]')
			else
				table.insert(categories, "[[Category:Han char unsimplified usage]]")
			end
		end
		table.insert(text, ", ''[[Index:Chinese four corner|four-corner]]'' " .. table.concat(four_split,' or '))
		if #four_split > 1 then
			table.insert(categories, "[[Category:Han char with multiple four]]")
		end
	end
	if args['ids'] and args['ids'] ~= '' then
		local ids = args['ids']
		local ids_split = mw.text.split(ids,',')
		for i=1, #ids_split do
			if not mw.ustring.match(ids_split[i],'[%[%]]') then
				local before = ""
				ids_split[i] = mw.ustring.gsub(ids_split[i], "([⺀-⻳一-鿪㐀-䶵 -〿𠀀-𮯠!-○])",
					function (character)
						if not mw.ustring.find(before, character) then
							before = before .. character
							return "[[" .. character .. "]]"
						else return character end
					end)
			end
		end
		table.insert(text, ", ''[[w:Chinese character description languages#Ideographic Description Sequences|composition]]'' " .. table.concat(ids_split,' or '))
		if #ids_split > 1 then
			table.insert(categories, "[[Category:Han char with multiple ids]]")
		end
	end
	table.insert(text, ')')
	if args['rad'] and args['rad'] ~= '' then
		if mw.title.getCurrentTitle().nsText == '' then
			table.insert(categories, "[[Category:Han script characters|" .. sort_key .. "]]")
		end
	else
		if mw.title.getCurrentTitle().nsText == '' then
			table.insert(categories, "[[Category:Han char without rad]]")
		end
	end
	local params_to_check = { "as", "rn", "sn", "ids", "four", "canj" }
	local sortkey = args['rad'] and "|" .. args['rad'] or ""
	for i, param in pairs(params_to_check) do
		if not args[param] or args[param] == '' then
			table.insert(categories, "[[Category:Han char without " .. param .. sortkey .. "]]")
		end
	end
	
	if mw.ustring.sub(args['sn'],1,1) == '0' then
		table.insert(categories, "[[Category:Invalid parameter in Han char]]")
	end
	return table.concat(text) .. table.concat(categories)
end

return export