Module:Hepburner - Revision history

Dinoguy1000: we can eliminate two mw.ustring.codepoint calls just by moving the kanaDelta definition below the [Lower|Upper]Bound definitions and using them to define kanaDelta

2024-06-17T04:57:18Z

we can eliminate two mw.ustring.codepoint calls just by moving the kanaDelta definition below the [Lower|Upper]Bound definitions and using them to define kanaDelta

← Previous revision		Revision as of 04:57, 17 June 2024
Line 98:		Line 98:
	local kanaFound = false		local kanaFound = false
	local convertedString = ""		local convertedString = ""
⚫	local kanaDelta = (~~mw.ustring.codepoint("ァ")~~ - ~~mw.ustring.codepoint("ぁ")~~) -- difference in the unicode table
	local hiraganaLowerBound = mw.ustring.codepoint("ぁ") -- I know magic numbers are bad but it almost seems more worth		local hiraganaLowerBound = mw.ustring.codepoint("ぁ") -- I know magic numbers are bad but it almost seems more worth
	local hiraganaUpperBound = mw.ustring.codepoint("ゖ") -- it to use them in this context		local hiraganaUpperBound = mw.ustring.codepoint("ゖ") -- it to use them in this context
	local katakanaLowerBound = mw.ustring.codepoint("ァ") -- this is really expensive		local katakanaLowerBound = mw.ustring.codepoint("ァ") -- this is really expensive
	local katakanaUpperBound = mw.ustring.codepoint("ヶ")		local katakanaUpperBound = mw.ustring.codepoint("ヶ")
		⚫	local kanaDelta = (katakanaLowerBound - hiraganaLowerBound) -- difference in the unicode table
	for c in mw.ustring.gcodepoint(data) do		for c in mw.ustring.gcodepoint(data) do
	if c<=127 then -- short circuit for ascii, which is the intended use.		if c<=127 then -- short circuit for ascii, which is the intended use.

DarmaniLink: add todo

2024-03-08T06:25:21Z

add todo

← Previous revision		Revision as of 06:25, 8 March 2024
Line 31:		Line 31:
	local function kanaToHepburn(kana)		local function kanaToHepburn(kana)
	local romanji = ""		local romanji = ""


			--TODO split map up into consonant groups and create a jump table based off the unicode value
	local kanaMap = {		local kanaMap = {
	["あ"] = "a", ["い"] = "i", ["う"] = "u", ["え"] = "e", ["お"] = "o",		["あ"] = "a", ["い"] = "i", ["う"] = "u", ["え"] = "e", ["お"] = "o",

DarmaniLink: change a rule

2024-03-08T06:15:16Z

change a rule

DarmaniLink: fix comment alignment

2024-03-08T02:48:10Z

fix comment alignment

DarmaniLink: rm todo

2024-03-08T02:37:37Z

rm todo

← Previous revision		Revision as of 02:37, 8 March 2024
Line 136:		Line 136:

	--TODO add a performant way to detect if there is kana in a string		--TODO add a performant way to detect if there is kana in a string
	--TODO differentiate accordingly
	--this could be expanded to use bopomofo too		--this could be expanded to use bopomofo too
	function p.toHepburn(frame)		function p.toHepburn(frame)

DarmaniLink: Commenting everything to kingdom come for maintenance

2024-03-08T02:35:06Z

Commenting everything to kingdom come for maintenance

← Previous revision		Revision as of 02:35, 8 March 2024
Line 53:		Line 53:
	local smallKanaMap = {		local smallKanaMap = {
	["ぁ"] = "a", ["ぃ"] = "i", ["ぅ"] = "u", ["ぇ"] = "e", ["ぉ"] = "o",		["ぁ"] = "a", ["ぃ"] = "i", ["ぅ"] = "u", ["ぇ"] = "e", ["ぉ"] = "o",
	["ヵ"] = "ka",["ヶ"] = "ke",		["ゕ"] = "ka", ["ゖ"] = "ke",
	["ゃ"] = "ya", ["ゅ"] = "yu", ["ょ"] = "yo"		["ゃ"] = "ya", ["ゅ"] = "yu", ["ょ"] = "yo"
	}		}


	-- iterate over each kana
	for character in mw.ustring.gcodepoint(kana) do		for character in mw.ustring.gcodepoint(kana) do -- iterates over each kana to convert it to romanji without diacritics
	local char = mw.ustring.char(character)		local char = mw.ustring.char(character) -- this is a really bad way of doing a foreach but
	local romanization = kanaMap[char]		local romanization = kanaMap[char] -- i'm new to mw lua, and couldn't get it to play nice
	local smallRomanization		local smallRomanization = smallKanaMap[char]

	-- short circuit eval, rare scenario
	if not romanization then
	smallRomanization = smallKanaMap[char]
	end


	if romanization then		if romanization then -- if kana was found in kanaMap
	romanji = romanji .. romanization		romanji = romanji .. romanization
	elseif smallRomanization then		elseif smallRomanization then -- if kana was not found in kanaMap, but found in smallKanaMap
	local lasttwo = romanji:sub(-3,-2)		local lasttwo = romanji:sub(-3,-2)
	if(lasttwo == "sh" or lasttwo == "ch") then		if(lasttwo == "sh" or lasttwo == "ch") then -- special case for shi and chi, because the rules
	romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1)		romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1) -- for these is しゅ=> shu rather than shyu or syu
	else		else
	romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters		romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters
			end -- this will cause issues if someone tries something like あぁぁぁぁぁ => a
	end
	else -- ~~this~~ ~~will~~ ~~cause~~ ~~issues~~ if ~~someone~~ ~~tries~~ ~~something~~ ~~like~~ ~~あぁぁぁぁぁ~~ => a		else -- special rule for vowels maybe? will make this more expensive
	romanji = romanji .. char -- ~~special rule~~ ~~for~~ ~~vowels~~ ~~maybe?~~ ~~will~~ ~~make~~ ~~this~~ ~~more~~ ~~expensive~~		romanji = romanji .. char -- character was not in either map, append it directly
	end		end
	end		end


	-- handle xtsu/っ
	-- Replace "っ" with the next consonant		-- Replace "っ" with the next consonant
	for i = 1, mw.ustring.len(romanji) do		for i = 1, mw.ustring.len(romanji) do
	local chr = mw.ustring.sub(romanji, i, i)		local chr = mw.ustring.sub(romanji, i, i) -- string[i]
	if chr == "っ" then		if chr == "っ" then
	local nextChar = mw.ustring.sub(romanji, i + 1, i + 1)		local nextChar = mw.ustring.sub(romanji, i + 1, i + 1) -- get the next letter after the small tsu
	if nextChar and not nextChar:match("[aeiou]") then		if nextChar and not nextChar:match("[aeiou]") then -- if it's a vowel, we don't replicate it.
	romanji = mw.ustring.sub(romanji, 1, i-1) .. nextChar .. mw.ustring.sub(romanji, i + 1) -- ~~surely~~ ~~there's~~ a ~~better~~ ~~way,~~ ~~right?~~		romanji = mw.ustring.sub(romanji, 1, i-1) .. nextChar .. mw.ustring.sub(romanji, i + 1) -- string before the small tsu + next character + string after that one xtsu
			end -- surely there's a better way?
	end
	end		end
	end		end
Line 109:		Line 103:
	local katakanaUpperBound = mw.ustring.codepoint("ヶ")		local katakanaUpperBound = mw.ustring.codepoint("ヶ")
	for c in mw.ustring.gcodepoint(data) do		for c in mw.ustring.gcodepoint(data) do
	if c<=127 then		if c<=127 then -- short circuit for ascii, which is the intended use.
			-- kana support was intended to be a minor feature
	--short circuit eval for ascii
	elseif (hiraganaLowerBound <= c and c <= hiraganaUpperBound) then		elseif (hiraganaLowerBound <= c and c <= hiraganaUpperBound) then
	kanaFound = true		kanaFound = true
	elseif (katakanaLowerBound <= c and c <= katakanaUpperBound) then		elseif (katakanaLowerBound <= c and c <= katakanaUpperBound) then
	kanaFound = true		kanaFound = true
	c = c - kanaDelta -- convert it to hiragana ~~codewise~~ so i dont have to remake the lookup table for katakana		c = c - kanaDelta -- convert to hiragana codepointwise so i dont have to remake the lookup table for katakana
	end		end
	convertedString = convertedString .. mw.ustring.char(c)		convertedString = convertedString .. mw.ustring.char(c) -- append the processed character to the current running string
			end -- this is bad, we're rebuilding the entire string just for katakana
	end
			-- maybe there's some string sub magic we can do?

	return {kanaFound, convertedString}		return {kanaFound, convertedString}
	end		end
Line 132:		Line 126:
	processedData = checkForKanaPresentAndConvert(data)		processedData = checkForKanaPresentAndConvert(data)

	if processedData[1] then		if processedData[1] then -- processedData[1] = kanaFound
	return kanaToHepburn(processedData[2])		return kanaToHepburn(processedData[2]) -- processedData[2] = convertedString
	else		else
	return romanjiToHepburn(data)		return romanjiToHepburn(data) -- kana not found, that should mean we were probably given romanji
			end -- and if its other unicode, they just get that string back
	end
	end		end

Line 143:		Line 137:
	--TODO add a performant way to detect if there is kana in a string		--TODO add a performant way to detect if there is kana in a string
	--TODO differentiate accordingly		--TODO differentiate accordingly
	--katakana=>hiragana? Can I just iterate over each and add a magic number like you can with ascii?
	--this could be expanded to use bopomofo too		--this could be expanded to use bopomofo too
	function p.toHepburn(frame)		function p.toHepburn(frame)
Line 149:		Line 142:
	return toHepburnKana(data)		return toHepburnKana(data)
	end		end

			-- testing function for the lua console on the module itself
	function p.toHepburnTEST(data)		function p.toHepburnTEST(data)
	return toHepburnKana(data)		return toHepburnKana(data)

	end		end

	return p		return p

DarmaniLink: oopsie

2024-03-07T19:30:18Z

oopsie

← Previous revision		Revision as of 19:30, 7 March 2024
Line 68:		Line 68:
	end		end


	if romanization then		if romanization then
	romanji = romanji .. romanization		romanji = romanji .. romanization
	elseif smallRomanization then		elseif smallRomanization then
	~~romanji~~ = romanji:sub(1, -2) ~~.. smallKanaMap[char] -- remove vowel, append small vowel or other letters~~		local lasttwo = romanji:sub(-3,-2)
			if(lasttwo == "sh" or lasttwo == "ch") then
			romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1)
			else
			romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters
			end
	else -- this will cause issues if someone tries something like あぁぁぁぁぁ => a		else -- this will cause issues if someone tries something like あぁぁぁぁぁ => a
	romanji = romanji .. char -- special rule for vowels maybe? will make this more expensive		romanji = romanji .. char -- special rule for vowels maybe? will make this more expensive

DarmaniLink: updated to add kana support

2024-03-07T19:03:38Z

updated to add kana support

@@ Line 1: / Line 1: @@
 require('strict');
 -- standard long vowel patterns
@@ Line 20: / Line 21: @@
     ["OU"] = "Ō"
+}
     end
     return romanji
 end
 return p

DarmaniLink: no long i, oops!

2024-03-03T03:37:02Z

no long i, oops!

← Previous revision		Revision as of 03:37, 3 March 2024
Line 5:		Line 5:
	local diacritics = {		local diacritics = {
	["aa"] = "ā",		["aa"] = "ā",
	["ii"] = "ī",
	["uu"] = "ū",		["uu"] = "ū",
	["ee"] = "ē",		["ee"] = "ē",
Line 11:		Line 10:
	["ou"] = "ō",		["ou"] = "ō",
	["Aa"] = "Ā",		["Aa"] = "Ā",
	["Ii"] = "Ī",
	["Uu"] = "Ū",		["Uu"] = "Ū",
	["Ee"] = "Ē",		["Ee"] = "Ē",
Line 17:		Line 15:
	["Ou"] = "Ō",		["Ou"] = "Ō",
	["AA"] = "Ā",		["AA"] = "Ā",
	["II"] = "Ī",
	["UU"] = "Ū",		["UU"] = "Ū",
	["EE"] = "Ē",		["EE"] = "Ē",

DarmaniLink: there we (actually) go :)

2024-03-02T14:54:09Z

there we (actually) go :)

← Previous revision		Revision as of 14:54, 2 March 2024
Line 1:		Line 1:
	require('strict');		require('strict');
	-- Converts romanji input to modified hepburn, I recommend subst:ing		-- Converts romanji input to modified hepburn, I recommend subst:ing

⚫	local p = {}
	-- standard long vowel patterns		-- standard long vowel patterns
	local diacritics = {		local diacritics = {
Line 24:		Line 24:
	}		}

		⚫	local p = {}
	--input: 1:1 transliterated romanji		--input: 1:1 transliterated romanji
	function p.toHepburn(frame~~, romanji~~)		function p.toHepburn(frame)
			local romanji = frame.args[1]
	for target, replacement in pairs(diacritics) do		for target, replacement in pairs(diacritics) do
	romanji = romanji:gsub(target, replacement)		romanji = romanji:gsub(target, replacement)
	end		end
	return romanji		return romanji
	end		end

← Previous revision		Revision as of 06:15, 8 March 2024
Line 46:		Line 46:
	["が"] = "ga", ["ぎ"] = "gi", ["ぐ"] = "gu", ["げ"] = "ge", ["ご"] = "go",		["が"] = "ga", ["ぎ"] = "gi", ["ぐ"] = "gu", ["げ"] = "ge", ["ご"] = "go",
	["ざ"] = "za", ["じ"] = "ji", ["ず"] = "zu", ["ぜ"] = "ze", ["ぞ"] = "zo",		["ざ"] = "za", ["じ"] = "ji", ["ず"] = "zu", ["ぜ"] = "ze", ["ぞ"] = "zo",
	["だ"] = "da", ["ぢ"] = "ji", ["づ"] = "zu", ["で"] = "de", ["ど"] = "do",		["だ"] = "da", ["ぢ"] = "ji", ["づ"] = "dzu", ["で"] = "de", ["ど"] = "do",
	["ば"] = "ba", ["び"] = "bi", ["ぶ"] = "bu", ["べ"] = "be", ["ぼ"] = "bo",		["ば"] = "ba", ["び"] = "bi", ["ぶ"] = "bu", ["べ"] = "be", ["ぼ"] = "bo",
	["ぱ"] = "pa", ["ぴ"] = "pi", ["ぷ"] = "pu", ["ぺ"] = "pe", ["ぽ"] = "po",		["ぱ"] = "pa", ["ぴ"] = "pi", ["ぷ"] = "pu", ["ぺ"] = "pe", ["ぽ"] = "po",
Line 67:		Line 67:
	elseif smallRomanization then -- if kana was not found in kanaMap, but found in smallKanaMap		elseif smallRomanization then -- if kana was not found in kanaMap, but found in smallKanaMap
	local lasttwo = romanji:sub(-3,-2)		local lasttwo = romanji:sub(-3,-2)
	if(lasttwo == "sh" or lasttwo == "ch") then -- special case for ~~shi and chi, because~~ the ~~rules~~		if(lasttwo == "sh" or lasttwo == "ch" or lasttwo == "ts" or lasttwo == "dz") then -- special case for the 3 letter romanizations
	romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1) -- ~~for these is~~ しゅ=> shu rather than shyu or syu		romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1) -- since しゅ=> shu rather than shyu or syu
	else		else
	romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters		romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters
	end -- this will cause issues if someone tries something like あぁぁぁぁぁ => a		end -- this will cause issues if someone tries something like あぁぁぁぁぁ => a
	else -- special rule for vowels maybe? will make this more expensive		else -- special rule for double little vowels maybe? will make this more expensive
	romanji = romanji .. char -- character was not in either map, append it directly		romanji = romanji .. char -- character was not in either map, append it directly
	end		end

← Previous revision		Revision as of 02:48, 8 March 2024
Line 58:		Line 58:


	for character in mw.ustring.gcodepoint(kana) do -- iterates over each kana to convert it to romanji without diacritics		for character in mw.ustring.gcodepoint(kana) do -- iterates over each kana to convert it to romanji without diacritics
	local char = mw.ustring.char(character) -- this is a really bad way of doing a foreach but		local char = mw.ustring.char(character) -- this is a really bad way of doing a foreach but
	local romanization = kanaMap[char] -- i'm new to mw lua, and couldn't get it to play nice		local romanization = kanaMap[char] -- i'm new to mw lua, and couldn't get it to play nice
	local smallRomanization = smallKanaMap[char]		local smallRomanization = smallKanaMap[char]

Line 68:		Line 68:
	local lasttwo = romanji:sub(-3,-2)		local lasttwo = romanji:sub(-3,-2)
	if(lasttwo == "sh" or lasttwo == "ch") then -- special case for shi and chi, because the rules		if(lasttwo == "sh" or lasttwo == "ch") then -- special case for shi and chi, because the rules
	romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1) -- for these is しゅ=> shu rather than shyu or syu		romanji = romanji:sub(1, -2) .. smallKanaMap[char]:sub(-1) -- for these is しゅ=> shu rather than shyu or syu
	else		else
	romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters		romanji = romanji:sub(1, -2) .. smallKanaMap[char] -- remove vowel, append small vowel or other letters
Line 88:		Line 88:
	end		end
	end		end
	return romanjiToHepburn(romanji) -- kana is converted to romanji, now change it to hepburn		return romanjiToHepburn(romanji) -- kana is converted to romanji, now change it to hepburn
	end -- TODO: add a flag to disable this, and return the normal romanji without the diacritics		end -- TODO: add a flag to disable this, and return the normal romanji without the diacritics

	-- checking for kana will need to check these bounds regardless		-- checking for kana will need to check these bounds regardless
Line 103:		Line 103:
	local katakanaUpperBound = mw.ustring.codepoint("ヶ")		local katakanaUpperBound = mw.ustring.codepoint("ヶ")
	for c in mw.ustring.gcodepoint(data) do		for c in mw.ustring.gcodepoint(data) do
	if c<=127 then -- short circuit for ascii, which is the intended use.		if c<=127 then -- short circuit for ascii, which is the intended use.
	-- kana support was intended to be a minor feature		-- kana support was intended to be a minor feature
	elseif (hiraganaLowerBound <= c and c <= hiraganaUpperBound) then		elseif (hiraganaLowerBound <= c and c <= hiraganaUpperBound) then
	kanaFound = true		kanaFound = true
Line 127:		Line 127:

	if processedData[1] then -- processedData[1] = kanaFound		if processedData[1] then -- processedData[1] = kanaFound
	return kanaToHepburn(processedData[2]) -- processedData[2] = convertedString		return kanaToHepburn(processedData[2]) -- processedData[2] = convertedString
	else		else
	return romanjiToHepburn(data) -- kana not found, that should mean we were probably given romanji		return romanjiToHepburn(data) -- kana not found, that should mean we were probably given romanji