@@ Line 3: / Line 3: @@
 local U = mw.ustring
 local gsub = U.gsub
-local find = U.find
+local match = U.match
 local sub = U.sub
-local match = U.match
 local len = U.len
@@ Line 12: / Line 11: @@
 	['ข'] = 'kh', ['ฃ'] = 'kh', ['ค'] = 'kh', ['ฅ'] = 'kh', ['ฆ'] = 'kh',
 	['ง'] = 'ng',
-	['จ'] = 'ch', ['ฉ'] = 'ch', ['ช'] = 'ch',
+	['จ'] = 'ch',
+	['ฉ'] = 'ch', ['ช'] = 'ch',
 	['ซ'] = 's',
 	['ฌ'] = 'ch',
@@ Line 46: / Line 46: @@
 	['ข'] = 'k', ['ฃ'] = 'k', ['ค'] = 'k', ['ฅ'] = 'k', ['ฆ'] = 'k',
 	['ง'] = 'ng',
-	['จ'] = 't', ['ฉ'] = 't', ['ช'] = 't', ['ฌ'] = '',
+	['จ'] = 't',
-	['ซ'] = 't',
+	['ฉ'] = 't', ['ช'] = 't', ['ซ'] = 't',
+	['ฌ'] = '',
 	['ญ'] = 'n',
 	['ฎ'] = 't', ['ฏ'] = 't',
@@ Line 71: / Line 72: @@
 }
--- Vowel patterns are tested longest-first.
+local CONSONANTS = 'กขฃคฅฆงจฉชซฌญฎฏฐฑฒณดตถทธนบปผฝพฟภมยรลวศษสหฬฮ'
-local VOWELS = {
-	{ 'เ' , 'ะ', 'e' },
-	{ 'เ' , '็', 'e' },
-	{ 'เ' , '',  'e' },
-	{ 'แ' , 'ะ', 'ae' },
-	{ 'แ' , '',  'ae' },
-	{ 'โ' , 'ะ', 'o' },
-	{ 'โ' , '',  'o' },
-	{ 'เ' , 'าะ', 'o' },
-	{ ''  , 'อ', 'o' },
-	{ 'เ' , 'อะ', 'oe' },
-	{ 'เ' , 'ิ', 'oe' },
-	{ 'เ' , 'อ', 'oe' },
-	{ 'เ' , 'ียะ', 'ia' },
-	{ 'เ' , 'ีย', 'ia' },
-	{ 'เ' , 'ือะ', 'uea' },
-	{ 'เ' , 'ือ', 'uea' },
-	{ ''  , 'ัวะ', 'ua' },
-	{ ''  , 'ัว', 'ua' },
-	{ ''  , 'ว', 'ua' },
-	{ 'ใ', '', 'ai' },
-	{ 'ไ', '', 'ai' },
-	{ '' , 'ัย', 'ai' },
-	{ 'ไ', 'ย', 'ai' },
-	{ '' , 'าย', 'ai' },
-	{ 'เ', 'า', 'ao' },
-	{ '' , 'าว', 'ao' },
-	{ '' , 'ุย', 'ui' },
-	{ 'โ', 'ย', 'oi' },
-	{ '' , 'อย', 'oi' },
-	{ 'เ', 'ย', 'oei' },
-	{ 'เ', 'ือย', 'ueai' },
-	{ '' , 'วย', 'uai' },
-	{ '' , 'ิว', 'io' },
-	{ 'เ', '็ว', 'eo' },
-	{ 'เ', 'ว', 'eo' },
-	{ 'แ', '็ว', 'aeo' },
-	{ 'แ', 'ว', 'aeo' },
-	{ 'เ', 'ียว', 'iao' },
-	{ '' , 'ะ', 'a' },
-	{ '' , 'ั', 'a' },
-	{ '' , 'า', 'a' },
-	{ '' , 'ำ', 'am' },
-	{ '' , 'ิ', 'i' },
-	{ '' , 'ี', 'i' },
-	{ '' , 'ึ', 'ue' },
-	{ '' , 'ื', 'ue' },
-	{ '' , 'ือ', 'ue' },
-	{ '' , 'ุ', 'u' },
-	{ '' , 'ู', 'u' },
-	{ '' , 'รร', 'a' }
-}
 local function isConsonant(c)
-	return c and INITIAL[c] ~= nil
+	return c and match(c, '[' .. CONSONANTS .. ']') ~= nil
 end
@@ Line 153: / Line 82: @@
 end
-local function isToneMark(c)
+local function isVowelMark(c)
-	return c and match(c, '[่้๊๋]') ~= nil
+	return c and match(c, '[ะาิีึืุูัำเแโใไอ]') ~= nil
-end
-local function isSilentMark(c)
-	return c == '์'
 end
@@ Line 168: / Line 93: @@
 end
-local function applyOrthographicExceptions(str)
+local function removeMarks(str)
-	-- Silent letters marked with ์.
+	-- Tone marks are not represented in RTGS.
-	-- Remove the marked letter and its ์ marker.
-	str = gsub(str, '([ก-ฮ])์', '')
-	-- Tone marks have no RTGS representation.
 	str = gsub(str, '[่้๊๋]', '')
-	-- ฤ
+	-- Mai thanthakhat makes the marked consonant silent.
-	str = gsub(str, 'ฤกษ์', 'roe')
+	str = gsub(str, '([ก-ฮ])์', '')
-	str = gsub(str, 'ฤิ', 'ri')
-	str = gsub(str, 'ฤ', 'rue')
-	str = gsub(str, 'ฤๅ', 'rue')
-	-- ฦ / ฦๅ
-	str = gsub(str, 'ฦๅ', 'lue')
-	str = gsub(str, 'ฦ', 'lue')
 	return str
@@ Line 190: / Line 104: @@
 local function getVowel(str, pos)
-	local remaining = sub(str, pos, -1)
+	local s = sub(str, pos, -1)
-	-- Explicit special vowel patterns.
+	-- Longest / most specific patterns first.
 	local patterns = {
 		{ 'เ', 'ื', 'อ', 'ย', 'ueai' },
 		{ 'เ', 'ี', 'ย', 'ว', 'iao' },
 		{ 'เ', 'ี', 'ย', 'ะ', 'ia' },
-		{ 'เ', 'ี', 'ย', '',  'ia' },
+		{ 'เ', 'ี', 'ย', '', 'ia' },
 		{ 'เ', 'ื', 'อ', 'ะ', 'uea' },
-		{ 'เ', 'ื', 'อ', '',  'uea' },
+		{ 'เ', 'ื', 'อ', '', 'uea' },
+		{ '', 'ั', 'ว', 'ะ', 'ua' },
+		{ '', 'ั', 'ว', '', 'ua' },
+		{ 'แ', '็', 'ว', '', 'aeo' },
+		{ 'แ', 'ว', '', '', 'aeo' },
 		{ 'เ', '็', 'ว', '', 'eo' },
 		{ 'เ', 'ว', '', '', 'eo' },
-		{ 'แ', '็', 'ว', '', 'aeo' },
-		{ 'แ', 'ว', '', '', 'aeo' },
-		{ 'โ', 'ย', '', '', 'oi' },
+		{ '', 'ิ', 'ว', '', 'io' },
-		{ '', 'อ', 'ย', '', 'oi' },
+		{ '', 'ุ', 'ย', '', 'ui' },
-		{ 'เ', 'ย', '', '', 'oei' },
+		{ '', 'ว', 'ย', '', 'uai' },
-		{ '', 'ว', 'ย', '', 'uai' },
+		{ 'เ', 'ื', 'อ', '', 'ue' },
-		{ '', 'ุ', 'ย', '', 'ui' },
+		{ 'เ', 'ื', '', '', 'ue' },
-		{ '', 'ิ', 'ว', '', 'io' },
 		{ 'เ', 'า', '', '', 'ao' },
 		{ '', 'า', 'ว', '', 'ao' },
-		{ '', 'ั', 'ว', 'ะ', 'ua' },
+		{ 'โ', 'ย', '', '', 'oi' },
-		{ '', 'ั', 'ว', '', 'ua' },
+		{ '', 'อ', 'ย', '', 'oi' },
-		{ '', 'ว', '', '', 'ua' },
+		{ 'เ', 'ย', '', '', 'oei' },
 		{ 'ใ', '', '', '', 'ai' },
 		{ 'ไ', '', '', '', 'ai' },
 		{ '', 'ั', 'ย', '', 'ai' },
-		{ 'ไ', 'ย', '', '', 'ai' },
 		{ '', 'า', 'ย', '', 'ai' },
-		{ 'เ', 'า', 'ะ', '', 'ao' },
+		{ 'เ', 'าะ', '', '', 'o' },
 		{ 'โ', 'ะ', '', '', 'o' },
-		{ 'เ', 'า', '', '', 'ao' },
+		{ 'โ', '', '', '', 'o' },
-		{ 'เ', 'าะ', '', '', 'o' },
+		{ '', 'อ', '', '', 'o' },
+		{ 'เ', 'อะ', '', '', 'oe' },
+		{ 'เ', 'ิ', '', '', 'oe' },
 		{ 'เ', 'อ', '', '', 'oe' },
-		{ 'เ', 'ิ', '', '', 'oe' },
-		{ 'เ', 'อะ', '', '', 'oe' },
 		{ 'แ', 'ะ', '', '', 'ae' },
 		{ 'แ', '', '', '', 'ae' },
 		{ 'เ', 'ะ', '', '', 'e' },
 		{ 'เ', '็', '', '', 'e' },
 		{ 'เ', '', '', '', 'e' },
-		{ 'โ', '', '', '', 'o' },
 		{ '', 'ำ', '', '', 'am' },
 		{ '', 'ะ', '', '', 'a' },
 		{ '', 'ั', '', '', 'a' },
 		{ '', 'า', '', '', 'a' },
 		{ '', 'ิ', '', '', 'i' },
 		{ '', 'ี', '', '', 'i' },
 		{ '', 'ึ', '', '', 'ue' },
-		{ '', 'ือ', '', '', 'ue' },
 		{ '', 'ื', '', '', 'ue' },
 		{ '', 'ุ', '', '', 'u' },
-		{ '', 'ู', '', '', 'u' },
+		{ '', 'ู', '', '', 'u' }
-		{ '', 'อ', '', '', 'o' }
 	}
 	for _, v in ipairs(patterns) do
-		local a, b, c, d, roman = v
+		local a = v[1] or ''
+		local b = v[2] or ''
+		local c = v[3] or ''
+		local d = v[4] or ''
+		local roman = v[5] or ''
 		local pattern = a .. b .. c .. d
-		if pattern ~= '' and sub(remaining, 1, len(pattern)) == pattern then
+		if pattern ~= '' and sub(s, 1, len(pattern)) == pattern then
 			return roman, len(pattern)
 		end
@@ Line 269: / Line 193: @@
 end
-local function nextVowelPosition(str, pos)
+local function romanizeSyllable(str, start)
-	local i = pos
+	local pos = start
-	while i <= len(str) do
+	local n = len(str)
-		local c = sub(str, i, i)
-		if c == 'เ' or c == 'แ' or c == 'โ' or c == 'ใ' or c == 'ไ' then
+	local initial = ''
-			return i
+	local initialRom = ''
-		end
-		if match(c, '[ะาิีึืุูัำอ]') then
+	-- Preposed vowels are written before the initial consonant.
-			return i
+	local pre = sub(str, pos, pos)
-		end
-		i = i + 1
+	if pre == 'เ' or pre == 'แ' or pre == 'โ' or pre == 'ใ' or pre == 'ไ' then
-	end
-	return nil
-end
-local function romanizeSyllable(str, startPos)
-	local pos = startPos
-	local n = len(str)
-	-- Preposed vowel.
-	local preVowel = ''
-	local preLen = 0
-	local c = sub(str, pos, pos)
-	if c == 'เ' or c == 'แ' or c == 'โ' or c == 'ใ' or c == 'ไ' then
-		preVowel = c
-		preLen = 1
 		pos = pos + 1
 	end
-	-- Initial consonant.
+	local first = sub(str, pos, pos)
-	local initial = sub(str, pos, pos)
-	if not isConsonant(initial) then
+	if not isConsonant(first) then
-		-- Non-Thai character.
+		return first, 1
-		return initial, 1
 	end
+	initial = first
+	initialRom = INITIAL[first] or ''
 	pos = pos + 1
-	local initialRom = INITIAL[initial] or ''
-	-- ห- exception.
+	-- ห + sonorant consonant: ห is not romanized.
-	if initial == 'ห' then
+	local second = sub(str, pos, pos)
-		local second = sub(str, pos, pos)
-		if second and match(second, '[งนมยรลวญณฬ]') then
+	if initial == 'ห' and second and match(second, '[งนมยรลวญณฬ]') then
-			initialRom = INITIAL[second]
+		initial = second
-			pos = pos + 1
+		initialRom = INITIAL[second]
-			initial = second
+		pos = pos + 1
-		end
 	end
-	-- Consonant clusters where ร is not represented.
+	-- จร-, ซร-, ศร-, สร-: ร is not romanized.
-	if (initial == 'จ' or initial == 'ซ' or initial == 'ศ' or initial == 'ส')
+	second = sub(str, pos, pos)
-		and sub(str, pos, pos) == 'ร' then
-		initialRom = INITIAL[initial]
+	if second == 'ร'
+		and (initial == 'จ' or initial == 'ซ' or initial == 'ศ' or initial == 'ส') then
 		pos = pos + 1
 	end
-	-- Find the vowel from the current position, including preposed vowels.
+	-- Read the vowel.
-	local vowelStart = startPos + preLen + 1
+	local vowelPos
-	local vowelRoman
-	local vowelLen
-	if preLen > 0 then
+	if pre == 'เ' or pre == 'แ' or pre == 'โ' or pre == 'ใ' or pre == 'ไ' then
-		vowelRoman, vowelLen = getVowel(str, startPos + 1)
+		vowelPos = start
 	else
-		vowelRoman, vowelLen = getVowel(str, pos)
+		vowelPos = pos
 	end
-	-- If the vowel was explicitly consumed by getVowel, advance.
+	local vowel, vowelLen = getVowel(str, vowelPos)
+	-- Advance over the vowel.
 	if vowelLen > 0 then
-		if preLen > 0 then
+		pos = vowelPos + vowelLen
-			pos = startPos + preLen + 1 + vowelLen
+	end
-		else
-			pos = pos + vowelLen
+	-- If there was a preposed vowel, getVowel consumed from its position,
-		end
+	-- so the resulting position already includes it.
+	if vowelLen == 0 and (pre == 'เ' or pre == 'แ' or pre == 'โ' or pre == 'ใ' or pre == 'ไ') then
+		pos = start + 1
 	end
-	-- Collect possible final consonant.
+	-- Final consonant.
 	local final = sub(str, pos, pos)
 	local finalRom = ''
@@ Line 358: / Line 263: @@
 		local after = sub(str, pos + 1, pos + 1)
-		-- A consonant followed by a vowel-starting syllable is normally
+		-- If another vowel begins immediately afterward, this consonant
-		-- treated as the next syllable's initial rather than this syllable's final.
+		-- belongs to the following syllable.
-		local nextIsVowel = after and (
+		if not isVowelMark(after) and not (after == 'เ' or after == 'แ' or after == 'โ' or after == 'ใ' or after == 'ไ') then
-			match(after, '[ะาิีึืุูัำอ]') or
-			match(after, '[เแโใไ]')
-		)
-		if not nextIsVowel then
 			finalRom = FINAL[final] or ''
 			pos = pos + 1
-			-- -ร following another final consonant is silent in RTGS.
+			-- Final -ร after another final consonant.
 			if final == 'ร' then
 				finalRom = 'n'
@@ Line 382: / Line 282: @@
 	end
-	return initialRom .. vowelRoman .. finalRom, math.max(1, pos - startPos)
+	-- -รร-
+	-- With a final consonant, รร = a.
+	-- Without a final consonant, รร = an.
+	if sub(str, vowelPos, vowelPos + 1) == 'รร' then
+		if finalRom ~= '' then
+			vowel = 'a'
+		else
+			vowel = 'an'
+		end
+	end
+	return initialRom .. vowel .. finalRom, math.max(1, pos - start)
 end
 local function romanizeWord(word)
-	if word == '' then
+	local result = {}
-		return ''
-	end
-	-- Manual markers:
-	-- < > = force space
-	-- <-> = force hyphen
-	word = gsub(word, '< >', ' ')
-	word = gsub(word, '<->', '-')
-	local output = {}
 	local pos = 1
 	local n = len(word)
@@ Line 404: / Line 305: @@
 		if c == '<' then
-			local close = find(word, '>', pos + 1, true)
+			local close = mw.ustring.find(word, '>', pos + 1, true)
 			if close then
+				local marker = sub(word, pos, close)
+				if marker == '< >' then
+					table.insert(result, ' ')
+				elseif marker == '<->' then
+					table.insert(result, '-')
+				else
+					table.insert(result, marker)
+				end
 				pos = close + 1
 			else
-				table.insert(output, c)
+				table.insert(result, c)
 				pos = pos + 1
 			end
-		elseif not isThai(c) then
+		elseif isThai(c) then
-			table.insert(output, c)
-			pos = pos + 1
-		else
 			local roman, consumed = romanizeSyllable(word, pos)
-			table.insert(output, roman)
+			table.insert(result, roman)
 			pos = pos + consumed
+		else
+			table.insert(result, c)
+			pos = pos + 1
 		end
 	end
-	return table.concat(output)
+	return table.concat(result)
 end
-local function capitalize(str, sentence)
+local function capitalizeWords(str)
 	local chars = mw.text.split(str, '')
-	local shouldCap = true
+	local capitalize = true
 	for i, c in ipairs(chars) do
-		if match(c, '%a') then
+		if match(c, '[A-Za-z]') then
-			if shouldCap then
+			if capitalize then
 				chars[i] = U.upper(c)
-				shouldCap = false
+				capitalize = false
-			elseif sentence then
-				shouldCap = false
 			end
-		elseif c == ' ' then
+		elseif c == ' ' or c == '-' then
-			if sentence then
+			capitalize = true
-				shouldCap = true
+		end
-			else
+	end
-				shouldCap = true
+	return table.concat(chars)
+end
+local function sentenceCase(str)
+	local chars = mw.text.split(str, '')
+	local capitalize = true
+	for i, c in ipairs(chars) do
+		if match(c, '[A-Za-z]') then
+			if capitalize then
+				chars[i] = U.upper(c)
+				capitalize = false
 			end
-		elseif match(c, '[%.!?]') then
+		elseif match(c, '[.!?]') then
-			shouldCap = true
+			capitalize = true
 		end
 	end
@@ Line 462: / Line 384: @@
 	local str = clean(args[1])
-	-- Don't romanize text which contains no Thai script.
+	-- Nothing to romanize.
 	if not match(str, '[ก-๛]') then
 		return ''
 	end
-	str = applyOrthographicExceptions(str)
+	str = removeMarks(str)
 	local result = romanizeWord(str)
 	if args.capi then
-		result = capitalize(result, false)
+		result = capitalizeWords(result)
 	elseif args.sent then
-		result = capitalize(result, true)
+		result = sentenceCase(result)
 	end

Module:Th-rm: Difference between revisions

ReisuDesign (talk | contribs)