မေႃႇၵျူး:Hani-sortkey

This module will sort text in the Han script. It is used to sort Ai-Cham, Southern Amami-Oshima, Central Bai, Northern Bai, Southern Bai, Biao-Jiao Mien, Biyo, Min Dong, Jin, Mandarin, Northern Pinghua, Chinese Pidgin English, Puxian, Macau Pidgin Portuguese, Southern Pinghua, Huizhou, Min Zhong, Dungan, Daur, Gan, Hakka, Xiang, Japanese, Kikai, Lama Bai, ၶႄႇ ပွတ်းၵၢင်, Literary Chinese, Middle Vietnamese, Caolan, Min Bei, Translingual, Miyako, Min Nan, Nung, ၶႄႇပၢၼ်ၵဝ်ႇ, Bouyei, Baekje, Tuyuhun, Tuoba, Wuhuan, Xianbei, Sui, Pai-lang, Toku-No-Shima, Kyakala, Tày, Vietnamese, Wu, Waxianghua, Classical Tibetan, Rouran, Middle Mongol, Buyeo, Cantonese, တႆးၸွင်ႈ, Zauzou, ၶႄႇ, Shaozhou Tuhua, Taishanese, Teochew, Goguryeo, Zakhring, and Khitan. The module should preferably not be called directly from templates or other modules. To use it from a template, use {{sortkey}}. Within a module, use Module:languages#Language:makeSortKey.

For testcases, see Module:Hani-sortkey/testcases.

Functions

makeSortKey(text, lang, sc)

Generates a sortkey for a given piece of text written in the script specified by the code sc, and language specified by the code lang.

When the sort fails, returns nil.

The demonstration functions that generated the content shown below are housed in Module:Hani-sortkey/templates. Modifications to the module can be tested in Module:Hani-sortkey/sandbox. Sortkeys for individual characters are retrieved from one of 178 data modules. Module:Hani-sortkey/data creates documentation for these modules.

Show sortkeys

• PS/2接口 (PS/2手08口00)
• gas爐 (gas火16)
• γ粒子 (γ米05子00)
• 命裡有時終須有，命裡無時莫強求 (口05衣07月02日06糸05頁03月02，口05衣07火08日06艸07弓08水02)
• 得個……字 (彳08人08……子03)
• 濕𣲷𣲷 (水14水05水05)
• 赛车 (貝10車00)

• 鿪 (火13)

• 九 (乙01)
• 𡆔 (口23)
• 𡎇 (土09)
• 阿坝 (阜05土04)

Ideographic description sequences

• ⿰亻革 (⿰人00革00)
• ⿰亻革家語 (⿰人00革00宀07言07)
• ⿺辶⿳穴⿲月⿱⿲幺言幺⿲長馬長刂心⿺辶⿳穴⿲月⿱⿲幺言幺⿲長馬長刂心麵 (⿺辵00⿳穴00⿲月00⿱⿲幺00言00幺00⿲長00馬00長00刀00心00⿺辵00⿳穴00⿲月00⿱⿲幺00言00幺00⿲長00馬00長00刀00心00麥09)

---

local export = {}

local cp = mw.ustring.codepoint
local concat = table.concat
local explode_utf8 = require("Module:string utilities").explode_utf8
local insert = table.insert
local namespace = mw.title.getCurrentTitle().nsText
local m_data = require("Module:Hani-sortkey/data/serialized")
local radicals_start = m_data:find("\254")
local preconvert_start = m_data:find("\255", radicals_start)

local pe = require("Module:utilities").pattern_escape

local function log(...)
	if namespace == "Module" then
		mw.log(...)
	end
end

--[[
	The number of characters or ideographic sequences that must follow each
	ideographic description character.
]]
local IDchars = {
	["⿰"] = 2,	-- left-to-right
	["⿱"] = 2,	-- above-to-below
	["⿲"] = 3,	-- left-to-middle and right
	["⿳"] = 3,	-- above-to-middle and below
	["⿴"] = 2,	-- full surround
	["⿵"] = 2,	-- surround from above
	["⿶"] = 2,	-- surround from below
	["⿷"] = 2,	-- surround from left
	["⿸"] = 2,	-- surround from upper left
	["⿹"] = 2,	-- surround from upper right
	["⿺"] = 2,	-- surround from lower left
	["⿻"] = 2,	-- overlaid
	["⿼"] = 2,	-- surround from right
	["⿽"] = 2,	-- surround from lower right
	["⿾"] = 1,	-- horizontal reflection
	["⿿"] = 1,	-- rotation
	["㇯"] = 1	-- subtraction
}

--[[
	Returns the index in the string where the ideographic description sequence
	(IDS) ends, or the index of the end of the string. Iterates whenever
	another ideographic description character (IDC) is found.
]]
local function findEndOfIDS(text, IDchar, i)
	if not (text and IDchar and i) then
		return nil
	end
	
	local j = i
	local component = 1
	
	-- Number of components expected after current IDC.
	local components = IDchars[IDchar]
	
	while component <= components do
		j = j + 1
		
		local char = text[j]
		
		if not char then
			break
		elseif IDchars[char] then
			j = findEndOfIDS(text, char, j)
		end
		
		component = component + 1
	end
	
	--[[
		If the expected number of components has been found,
		return the current index in the text.
	]]
	if component - components == 1 then
		return j
	else
		return nil
	end
end

local module_cache = {}

local function radical_lookup(b)
	local lookup = 3 * b:byte() + radicals_start
	return m_data:sub(lookup - 2, lookup)
end

-- The data is stored in [[Module:Hani-sortkey/data]]. This data is not accessed directly (due to the large amount of memory this would consume), but is instead stored in a serialized form as [[Module:Hani-sortkey/data/serialized]]. If the data is changed, the new serialized data can be generated with [[Module:Hani-sortkey/data/serializer]].
function export.getData(char, returnModule)
	if type(char) == "string" then
		char = cp(char)
	elseif type(char) ~= number then
		error("getData must operate on a single character or codepoint.")
	end
	local sections = {
		{0x3007, 0x3007},
		{0x3400, 0x4DBF},
		{0x4E00, 0x9FFF},
		{0xF900, 0xFA6D},
		{0xFA70, 0xFAD9},
		{0x20000, 0x2A6DF},
		{0x2A700, 0x2B739},
		{0x2B740, 0x2B81D},
		{0x2B820, 0x2CEA1},
		{0x2CEB0, 0x2EBE0},
		{0x2EBF0, 0x2EE5D},
		{0x2F800, 0x2FA1D},
		{0x30000, 0x3134A},
		{0x31350, 0x323AF}
	}
	local section_offset = 0
	for k, v in ipairs(sections) do
		if char > v[2] then
			section_offset = section_offset + v[2] - v[1] + 1
		elseif char >= v[1] and char <= v[2] then
			local lookup = 3 * (section_offset + char - sections[k][1] + 1)
			return m_data:sub(lookup - 2, lookup):gsub("^.", radical_lookup)
		end
	end
	return mw.ustring.char(char)
end

local unsupported_data

function export.makeSortKey(text, lang, sc)
	local scripts = {
		Hani = true,
		Hans = true,
		Hant = true,
		Jpan = true,
		Kore = true
	}
	if sc and not scripts[sc] then
		return text:uupper()
	end
	
	local sort = {}
	text = explode_utf8(text)
	local text_len = #text
	local i, char = 0
	while i < text_len do
		i = i + 1
		char = text[i]
		
		char = (m_data:match("%f[^\2\255]" .. pe(char) .. "\1([^\2]+)\2", preconvert_start)) or char
		--[=[
			If we encounter an ideographic description character (IDC),
			find out if it begins a valid ideographic description sequence (IDS).
			
			If the IDS is valid and a sortkey for it is listed in
			[[Module:Hani-sortkey/data/unsupported]], then return
			the sortkey, and move to the next character after the
			IDS.
			
			Otherwise, insert the IDC into the sortkey and move to the next
			character after the IDC.
			
			If the IDS is valid and no sortkey for it is found, track it.
		]=]
		if IDchars[char] then
			local j = findEndOfIDS(text, char, i)
			local IDS, data
			if j then
				IDS = concat(text, nil, i, j)
				unsupported_data = unsupported_data or mw.loadData("Module:Hani-sortkey/data/unsupported")
				data = unsupported_data[IDS]
			end
			
			if not data then
				if IDS then
					require("Module:debug").track("Hani-sortkey/IDS-without-sortkey")
					mw.log("ideographic description sequence without sortkey: '"
						.. IDS .. "'")
				else
					require("Module:debug").track("Hani-sortkey/invalid-IDS")
					mw.log("invalid ideographic description sequence at the beginning of '"
						.. text[i] .. "'")
				end
			end
			if IDS and data then
				insert(sort, data)
				i = j
			else
				insert(sort, char)
			end
		else
			insert(sort, export.getData(char) or char)
		end
	end
	
	return concat(sort)
end

return export