local p = {}
local m_utils = require('Module:Ko-utils')
local gsub = mw.ustring.gsub
local gsub_iterate = m_utils.gsub_iterate
--[[
IMPORTANT NOTE before editing this module:
1. Make sure that you use a font that displays the following characters differently, and that you know the differences of them:
ᄀ (U+1100)
ᆨ (U+11A8)
ㄱ (U+3131)
2. When dealing with decomposed Hangul,
a. [ᄀ-ᄒ] should not be directly followed by [ᅡ-ᅵ] because MediaWiki applies Unicode Normalization Form C (NFC), which converts any sequence of [ᄀ-ᄒ][ᅡ-ᅵ] into a precomposed character. Use square brackets or parentheses to avoid this (e.g. ᄀ[ᅡ], (ᄀ)ᅡ, etc.)
b. ᄀ[ᅡ] or (ᄀ)ᅡ at the end of a pattern is equivalent to not just 가 but [가-갛] in precomposed form. To match a syllabic block without a final consonant at the end of a pattern, use both vowel + [^ᆨ-ᇂ] and vowel + $
For example, to only match 가 (and not [각-갛]) at the end of a pattern, use both ᄀ[ᅡ][^ᆨ-ᇂ] and ᄀ[ᅡ]$
--]]
function p.respell(text, mode)
--[[
IMPORTANT: Before adding a replacement, be sure to check if it can ALWAYS be applied in ALL contexts.
Good example: 싫증 → 실@증
Bad example: 문자 → 문@자 (affects words like 방문자 (pronounced [방문자], not [방문짜]))
--]]
-- Validate mode param
mode = mw.ustring.lower(mode)
assert(mode == "mr" or mode == "rr", "mode must be either mr or rr")
-- MR only: ㅢ → ㅣ, except 의, 희
if mode == "mr" then
text = gsub(text, "([ᄀ-ᄊᄌ-ᄑ])ᅴ", "%1ᅵ")
end
local replacements = {
-- Converting some Compatibility Jamo (U+31xx) to Hangul Jamo (U+11xx) for linguistic contexts
-- As Compatibility Jamo do not differentiate between initial/final, we convert only several jamo where initial/final is a relevant distinction
{"ㄴ([ᄀ-ᄒ])", "ᆫ%1"}, -- -ㄴ다
{"ㄹ([ᄀ-ᄒ])", "ᆯ%1"}, -- -ㄹ까, -ㄹ래
{"ㄹ@([ᄀᄃᄇᄉᄌ])", "ᆯ@%1"}, -- -ㄹ지
{"%-ㄹ", "-ᆯ"}, -- -ㄹ (with actual hyphen)
{"ㅂ([ᄀ-ᄒ])", "ᆸ%1"}, -- -ㅂ니다, -ㅂ시다
-- cases where ㄺㄱ is pronounced [ㄹㄲ]
-- verb/adjective stem ending in ㄺ + ending/suffix beginning with ㄱ (e.g. 맑고 [말꼬], 긁개 [글깨])
{"(ᄀ[ᅡᅮᅳ])ᆰᄀ", "%1ᆯ@ᄀ"}, -- 갉-, 굵-, 긁-
{"(ᄂ[ᅡᅳ])ᆰᄀ", "%1ᆯ@ᄀ"}, -- 낡-, 늙-
{"([ᄆᄇ][ᅡᅮ])ᆰᄀ", "%1ᆯ@ᄀ"}, -- 맑-, 묽-, 밝-, 붉-
{"(ᄋ[ᅡᅥᅩᅵ])ᆰᄀ", "%1ᆯ@ᄀ"}, -- 앍-, 얽-, 옭-, 읽-
-- otherwise, ㄺㄱ is pronounced [ㄱㄲ]; handled later in neutralization of syllable-final consonants
-- Exception to palatalization:
-- 곧이어 [고디어] resyllabification
{"(ᄀ[ᅩ])ᆮᄋ(ᅵᄋ[ᅥ][^ᆨ-ᇂ])", "%1ᄃ%2"},
{"(ᄀ[ᅩ])ᆮᄋ(ᅵᄋ[ᅥ])$", "%1ᄃ%2"},
-- palatalization and ㅈ + -히-
-- 여/혀 (contractions of 이어/히어) captured where it is valid (namely after verbs/adjectives)
{"ᆮᄋ(ᅵ[ᆫᆯᆸ])", "ᄌ%1"}, -- e.g. 해돋이 [해도지]
{"ᆮᄋ(ᅵ[^ᆨ-ᇂ])", "ᄌ%1"}, -- ㄷ여 does not occur in dictionary
{"ᆮᄋ(ᅵ)$", "ᄌ%1"},
{"[ᆮᆽ]ᄒ(ᅵ[ᆫᆯᆷᆸ])", "ᄎ%1"}, -- e.g. 굳히다 [구치다], 꽂히다 [꼬치다]
{"[ᆮᆽ]ᄒ(ᅧᆻ)", "ᄎ%1"}, -- ㅈ + -히- included here because functionally same outcome
{"[ᆮᆽ]ᄒ([ᅧᅵ][^ᆨ-ᇂ])", "ᄎ%1"}, -- other ㅈ + ㅎ cases are ㅈㅎ → ㄷㅎ → ㅌ
{"[ᆮᆽ]ᄒ([ᅧᅵ])$", "ᄎ%1"},
{"ᆴᄋ(ᅵ[ᆫᆯᆷᆸ])", "ᆯᄎ%1"}, -- e.g. 훑이다 [훌치다]
{"ᆴᄋ(ᅧᆻ)", "ᆯᄎ%1"}, -- Other ㄾㅇ → ㄹㅌ
{"ᆴᄋ([ᅧᅵ][^ᆨ-ᇂ])", "ᆯᄎ%1"},
{"ᆴᄋ([ᅧᅵ])$", "ᆯᄎ%1"},
{"ᇀᄋ(ᅵ[ᆫᆯᆷᆸ])", "ᄎ%1"}, -- e.g. 붙이다 [부치다]
{"ᇀᄋ(ᅧᆻ)", "ᄎ%1"}, -- Other ㅌㅇ → ㅌ, or ㄷ (use $) or ㄴㄴ (use @)
{"ᇀᄋ([ᅧᅵ][^ᆨ-ᇂ])", "ᄎ%1"},
{"ᇀᄋ([ᅧᅵ])$", "ᄎ%1"},
-- {ㄵ, ㄺ, ㄼ} + -히-
{"ᆬᄒ(ᅵ[ᆫᆯᆷᆸ])", "ᆫᄎ%1"}, -- e.g. 앉히다 [안치다]
{"ᆬᄒ(ᅧᆻ)", "ᆫᄎ%1"},
{"ᆬᄒ([ᅧᅵ][^ᆨ-ᇂ])", "ᆫᄎ%1"},
{"ᆬᄒ([ᅧᅵ])$", "ᆫᄎ%1"},
{"ᆰᄒ(ᅵ[ᆫᆯᆷᆸ])", "ᆯᄏ%1"}, -- e.g. 밝히다 [발키다]
{"ᆰᄒ(ᅧᆻ)", "ᆯᄏ%1"},
{"ᆰᄒ([ᅧᅵ][^ᆨ-ᇂ])", "ᆯᄏ%1"},
{"ᆰᄒ([ᅧᅵ])$", "ᆯᄏ%1"},
{"ᆲᄒ(ᅵ[ᆫᆯᆷᆸ])", "ᆯᄑ%1"}, -- e.g. 넓히다 [널피다], 밟히다 [발피다]
{"ᆲᄒ(ᅧᆻ)", "ᆯᄑ%1"},
{"ᆲᄒ([ᅧᅵ][^ᆨ-ᇂ])", "ᆯᄑ%1"},
{"ᆲᄒ([ᅧᅵ])$", "ᆯᄑ%1"},
-- cases where 넓- is pronounced [넙] before consonant
{"(ᄂ[ᅥ])ᆲ([ᄁᄄ-ᄈᄊᄍ-ᄒ])", "%1ᆸ%2"}, -- all except ㄱ, ㄴ, ㄷ, ㅅ, ㅇ, ㅈ (exceptions handled below)
{"(ᄂ[ᅥ])ᆲ(ᄃ[ᅡ]ᄃ[ᅳ]ᆷ)", "%1ᆸ%2"}, -- 넓다듬이; all other 넓다 are [널따]
{"(ᄂ[ᅥ])ᆲ(ᄃ[ᅮ]ᆼ)", "%1ᆸ%2"}, -- 넓둥글다; last remaining 넓ㄷ with [넙]
{"(ᄂ[ᅥ])ᆲ(ᄉ[ᅡ]ᆯᄆ[ᅮ]ᆫ)", "%1ᆸ%2"}, -- 넓살문; only 넓ㅅ word with this pronunciation; compare 넓습니다 [널씀니다]
{"(ᄂ[ᅥ])ᆲ(ᄌ[ᅥᅮ]ᆨ)", "%1ᆸ%2"}, -- 넓적-, 넓죽-
-- 밟- is [밥] before consonant (except syllable-initial ㅇ; ㅎ covered earlier)
{"(ᄇ[ᅡ])ᆲ([^ᄋ])", "%1ᆸ%2"},
{"(ᄇ[ᅡ])ᆲ$", "%1ᆸ"},
-- 여덟 + particle (tensification does not occur as it is a noun; only relevant to MR)
{"(ᄋ[ᅧ]ᄃ[ᅥ])ᆲ([ᄀᄃᄇᄉᄌ])", "%1ᆯ%2"},
-- tensification after ㄵ, ㄼ, ㄾ (only relevant to MR)
-- ㄵ, ㄼ, ㄾ are only used in verb/adjective stems (except 여덟 handled above), and they tensify the ending following them (e.g. 앉고 [안꼬], 짧다 [짤따], 핥지 [할찌])
-- syllable-final ㄻ is not included because it does not always cause tensification (e.g. 굶기다 [굼기다]); ㄻ should be handled with @ manually as appropriate
-- syllable-initial ㅂ is not included because no verb/adjective ending starts with it
-- syllable-initial ㅅ is handled later in additional ㅎ combinations
{"([ᆬᆲᆴ])([ᄀᄃᄌ])", "%1@%2"},
-- Exceptions to neutralization + resyllabification
-- ㄴ-addition exceptions
{"([ᆨ-ᇂ])ᄋ([ᅧᅵ]ᇁ)", "%1ᄂ%2"}, -- 옆, 잎
{"([ᆨ-ᇂ])ᄋ(ᅲᆾ)", "%1ᄂ%2"}, -- 윷
-- Other exceptions
{"(ᄋ[ᅥ])ᆹᄋ(ᅢ[ᆫᆯᆷᆸᆻ])", "%1ᆸᄉ%2"}, -- 없애다 [업쌔다]
{"(ᄋ[ᅥ])ᆹᄋ(ᅢ[^ᆨ-ᇂ])", "%1ᆸᄉ%2"},
{"(ᄋ[ᅥ])ᆹᄋ(ᅢ)$", "%1ᆸᄉ%2"},
{"(ᄆ[ᅡᅥ])ᆺᄋ(ᅵᆻ)", "%1ᄉ%2"}, -- 맛있다 [마싣따], 멋있다 [머싣따]
{"(ᄇ[ᅡ])ᆮᄋ(ᅵᆺ)ᄀ", "%1ᄌ%2@ᄀ"}, -- 받잇ㄱ [바지ㄲ~바짇ㄲ]
{"(ᄇ[ᅡ])ᆮᄋ(ᅵᆺ[ᄃᄌ])", "%1ᄌ%2"}, -- 받잇 + {ㄷ, ㅈ} [바지~바짇] + {[ㄸ], [ㅉ]}
{"(ᄇ[ᅮ])ᇀᄋ(ᅵᆺ)ᄀ", "%1ᄎ%2@ᄀ"}, -- 붙잇ㄱ [부치ㄲ~부칟ㄲ]
{"(ᄋ[ᅡ])ᇁᄋ(ᅦᆺ)(ᄀ[ᅥ]ᆺ)", "%1ᄑ%2@%3"}, -- 앞엣것 [아페껃~아펟껃]
{"(ᄋ[ᅩ])ᆺᄋ(ᅦᆺᄂ[ᅵ][ᆫᆯᆸ])", "%1ᄉ%2"}, -- 옷엣니 [오센니]
{"(ᄋ[ᅩ])ᆺᄋ(ᅦᆺᄂ[ᅵ][^ᆨ-ᇂ])", "%1ᄉ%2"},
{"(ᄋ[ᅩ])ᆺᄋ(ᅦᆺᄂ[ᅵ])$", "%1ᄉ%2"},
{"(ᄒ[ᅡ]ᆫᄉ[ᅩ])ᇀᄋ(ᅦᆺ)(ᄇ[ᅡ]ᆸ)", "%1ᄐ%2@%3"}, -- 한솥엣밥 [한소테빱~한소텓빱]
-- Some neutralization + resyllabification automatically detected here
-- Processing happens later in neutralization + resyllabification
-- Other cases users must detect on their own and add $ to fix
{"([ᆩᆪᆬᆰ-ᆵᆹ-ᆻᆽ-ᇂ])(ᄋ[ᅡᅥᅧ][ᆨ-ᆺᆼ-ᇂ])", "%1$%2"}, -- except 아, 았, 어, 었, 여, 였
{"([ᆩᆪᆬᆰ-ᆵᆹ-ᆻᆽ-ᇂ])(ᄋ[ᅦ][ᆨ-ᆪᆬ-ᆮᆰ-ᇂ])", "%1$%2"}, -- except 에, 엔, 엘
{"([ᆩᆪᆬᆰ-ᆵᆹ-ᆻᆽ-ᇂ])(ᄋ[ᅭᅴ][ᆨ-ᇂ])", "%1$%2"}, -- except 요, 의 (w/o final consonant)
{"([ᆩᆪᆬᆰ-ᆵᆹ-ᆻᆽ-ᇂ])(ᄋ[ᅳᅵ][ᆨ-ᆪᆬ-ᆮᆰ-ᆶᆹ-ᇂ])", "%1$%2"}, -- except 으, 은, 을, 음, 읍, 이, 인, 일, 임, 입
{"([ᆩᆪᆬᆰ-ᆵᆹ-ᆻᆽ-ᇂ])(ᄋ[ᅢ-ᅤᅨ-ᅬᅮ-ᅲ])", "%1$%2"},
-- @ for ㄴ-addition
-- Can't go with other @/$ irregularities because needs to occur before resyllabification
{"([ᆨ-ᇂ])@ᄋ([ᅣᅤᅧᅨᅭᅲᅵ])", "%1ᄂ%2"}, -- e.g. 색연필 [생년필], 물엿 [물렫]
-- Resyllabification
-- {ㄱ, ㄷ, ㄹ/ㅀ, ㅂ} + ㅇ is handled later in neutralization + resyllabification; {ㄴ/ㄶ, ㅁ} + ㅇ is not needed for RR & MR
{"ᆩᄋ", "ᄁ"},
{"ᆪᄋ", "ᆨᄉ"},
{"ᆬᄋ", "ᆫᄌ"},
{"ᆰᄋ", "ᆯᄀ"},
{"ᆱᄋ", "ᆯᄆ"},
{"ᆲᄋ", "ᆯᄇ"},
{"ᆳᄋ", "ᆯᄉ"},
{"ᆴᄋ", "ᆯᄐ"},
{"ᆵᄋ", "ᆯᄑ"},
{"ᆹᄋ", "ᆸᄉ"},
{"ᆺᄋ", "ᄉ"},
{"ᆻᄋ", "ᄊ"},
{"ᆽᄋ", "ᄌ"},
{"ᆾᄋ", "ᄎ"},
{"ᆿᄋ", "ᄏ"},
{"ᇀᄋ", "ᄐ"},
{"ᇁᄋ", "ᄑ"},
{"ᇂᄋ", "ᄋ"}, -- silent; 좋아 [조아]
-- 싫증 [실쯩] (only ㅀㅈ [ㄹㅉ] exception)
{"(ᄉ[ᅵ])ᆶ(ᄌ[ᅳ]ᆼ)", "%1ᆯ@%2"},
-- convert ㅎ combinations
-- trivia: {ㄶ, ㅀ, ㅎ} + ㅂ don't actually exist, but added for completeness (syllable-final ㅎ is for aspiration anyway)
{"ᆭᄀ", "ᆫᄏ"},
{"ᆭᄃ", "ᆫᄐ"},
{"ᆭᄇ", "ᆫᄑ"},
{"ᆭᄌ", "ᆫᄎ"},
{"ᆶᄀ", "ᆯᄏ"},
{"ᆶᄃ", "ᆯᄐ"},
{"ᆶᄇ", "ᆯᄑ"},
{"ᆶᄌ", "ᆯᄎ"},
{"ᇂᄀ", "ᄏ"},
{"ᇂᄃ", "ᄐ"},
{"ᇂᄇ", "ᄑ"},
{"ᇂᄌ", "ᄎ"}
}
text = gsub_iterate(text, replacements)
if mode == "mr" then
replacements = {
-- additional ㅎ combinations
-- also including {ㄵ, ㄼ, ㄾ} + ㅅ here because functionally same outcome
{"[ᆬᆭ]ᄉ", "ᆫᄊ"},
{"[ᆲᆴᆶ]ᄉ", "ᆯᄊ"},
{"ᇂᄉ", "ᄊ"},
-- @ for written epenthetic ㅅ + ㄱ/ㅂ
{"ᆺ@ᄀ", "ᄁ"},
{"ᆺ@ᄇ", "ᄈ"}
}
text = gsub_iterate(text, replacements)
else
text = gsub(text, "ᇂᄉ", "ᄉ")
end
-- Neutralization of syllable-final consonants
-- after this, the 27 syllable-final consonants are reduced to 7 (ㄱ, ㄴ, ㄷ, ㄹ, ㅁ, ㅂ, ㅇ)
replacements = {
{"[ᆩᆪᆰᆿ]", "ᆨ"},
{"[ᆬᆭ]", "ᆫ"},
{"[ᆺᆻᆽᆾᇀᇂ]", "ᆮ"},
{"[ᆲᆳᆴᆶ]", "ᆯ"},
{"ᆱ", "ᆷ"},
{"[ᆵᆹᇁ]", "ᆸ"}
}
text = gsub_iterate(text, replacements)
-- Remaining @/$ irregularities
if mode == "mr" then
text = gsub(text, "([ᅡ-ᅵᆫᆷᆼ])@ᄉ", "%1ᄊ") -- Tensification of ㅅ
else
replacements = {
{"ᆨ@ᄒ", "ᄏ"},
{"ᆮ@ᄒ", "ᄐ"},
{"ᆸ@ᄒ", "ᄑ"}
}
text = gsub_iterate(text, replacements)
end
replacements = {
-- ㄴㄹ [ㄴㄴ], e.g. 음운론 [으문논]
{"ᆫ@ᄅ", "ᆫᄂ"},
-- neutralization + resyllabification
{"ᆨ%$?ᄋ", "ᄀ"},
{"ᆮ%$?ᄋ", "ᄃ"}, -- e.g. 웃어른 [우더른]
{"ᆯ%$?ᄋ", "ᄅ"},
{"ᆸ%$?ᄋ", "ᄇ"},
{"%$", ""} -- remove remaining $s
}
text = gsub_iterate(text, replacements)
-- MR only: voicing of ㄱ, ㄷ, ㅂ, ㅈ
if mode == "mr" then
replacements = {
{"ᆫᄀ", "ᆫ'`ᄀ"}, -- n'g
{"([ᅡ-ᅵᆫᆯᆷᆼ])([ᄀᄃᄇᄌ])", "%1`%2"},
{"([ᅡ-ᅵᆫᆯᆷᆼ])%*([ᄀᄃᄇᄌ])", "%1-`%2"} -- * is for additional hyphen in romanization only (voicing is retained after hyphen)
}
text = gsub_iterate(text, replacements)
end
-- remove remaining @s
-- to reflect tensification in MR, @ needs to be retained while voicing of ㄱ, ㄷ, ㅂ, ㅈ is being applied (@ prevents voicing)
-- e.g. 한자 [한짜] in MR is hancha (neither hanja nor hantcha)
text = gsub(text, "@", "")
-- Assimilation
replacements = {
{"[ᆨᆼ][ᄂᄅ]", "ᆼᄂ"},
{"ᆨᄆ", "ᆼᄆ"},
{"ᆫᄅ", "ᆯᄅ"},
{"ᆮ[ᄂᄅ]", "ᆫᄂ"},
{"ᆮᄆ", "ᆫᄆ"},
{"ᆯᄂ", "ᆯᄅ"},
{"[ᆷᆸ][ᄂᄅ]", "ᆷᄂ"},
{"ᆸᄆ", "ᆷᄆ"}
}
text = gsub_iterate(text, replacements)
-- MR only: no {kkk, ttt, ppp, sss/ts/tss, ttch}
if mode == "mr" then
replacements = {
{"ᆨᄁ", "ᄁ"},
{"ᆮᄄ", "ᄄ"},
{"ᆸᄈ", "ᄈ"},
{"ᆮ[ᄉᄊ]", "ᄊ"},
{"ᆮᄍ", "ᄍ"}
}
text = gsub_iterate(text, replacements)
end
-- Drop the semivowel y (/j/) after ㅈ, ㅉ, ㅊ
replacements = {
{"([ᄌ-ᄎ])ᅣ", "%1ᅡ"},
{"([ᄌ-ᄎ])ᅤ", "%1ᅢ"},
{"([ᄌ-ᄎ])ᅧ", "%1ᅥ"},
{"([ᄌ-ᄎ])ᅨ", "%1ᅦ"},
{"([ᄌ-ᄎ])ᅭ", "%1ᅩ"},
{"([ᄌ-ᄎ])ᅲ", "%1ᅮ"}
}
return gsub_iterate(text, replacements)
end
function p.romanize_rr(text)
local replacements = {
{"ᆯᄅ", "ᆯl"},
-- Vowels
{"[ᅡㅏ]", "a"},
{"[ᅢㅐ]", "ae"},
{"[ᅣㅑ]", "ya"},
{"[ᅤㅒ]", "yae"},
{"[ᅥㅓ]", "eo"},
{"[ᅦㅔ]", "e"},
{"[ᅧㅕ]", "yeo"},
{"[ᅨㅖ]", "ye"},
{"[ᅩㅗ]", "o"},
{"[ᅪㅘ]", "wa"},
{"[ᅫㅙ]", "wae"},
{"[ᅬㅚ]", "oe"},
{"[ᅭㅛ]", "yo"},
{"[ᅮㅜ]", "u"},
{"[ᅯㅝ]", "wo"},
{"[ᅰㅞ]", "we"},
{"[ᅱㅟ]", "wi"},
{"[ᅲㅠ]", "yu"},
{"[ᅳㅡ]", "eu"},
{"[ᅴㅢ]", "ui"},
{"[ᅵㅣ]", "i"},
-- to prevent input like 'ㅇ' (with ') from becoming italic markup (as [ᄋㅇ] is simply removed later)
{"'([ᄋㅇ]+)'", "'%1'"},
-- Consonants
{"[ᄀㄱ]", "g"},
{"[ᄁㄲ]", "kk"},
{"ㄳ", "ks"},
{"[ᄂᆫㄴ]", "n"},
{"ㄵ", "nj"},
{"ㄶ", "nh"},
{"[ᄃㄷ]", "d"},
{"[ᄄㄸ]", "tt"},
{"[ᄅㄹ]", "r"},
{"ᆯ", "l"},
{"ㄺ", "lg"},
{"ㄻ", "lm"},
{"ㄼ", "lb"},
{"ㄽ", "ls"},
{"ㄾ", "lt"},
{"ㄿ", "lp"},
{"ㅀ", "lh"},
{"[ᄆᆷㅁ]", "m"},
{"[ᄇㅂ]", "b"},
{"[ᄈㅃ]", "pp"},
{"ㅄ", "ps"},
{"[ᄉㅅ]", "s"},
{"[ᄊㅆ]", "ss"},
{"[ᄋㅇ]", ""},
{"ᆼ", "ng"},
{"[ᄌㅈ]", "j"},
{"[ᄍㅉ]", "jj"},
{"[ᄎㅊ]", "ch"},
{"[ᄏᆨㅋ]", "k"},
{"[ᄐᆮㅌ]", "t"},
{"[ᄑᆸㅍ]", "p"},
{"[ᄒㅎ]", "h"}
}
return gsub_iterate(text, replacements)
end
function p.romanize_mr(text)
local replacements = {
-- First some quirks related to MR
{"ᆯᄅ", "ᆯl"},
{"ᆯ%*ᄅ", "ᆯ-l"}, -- ㄹ-ㄹ (with hyphen in between) should probably be l-l rather than l-r
{"ᆯᄒ", "rᄒ"},
{"ᄉ[ᅱ]", "shᅱ"}, -- ㅅ is sh before ㅟ; otherwise sh is not used
-- Vowels
{"[ᅡㅏ]", "a"},
{"[ᅢㅐ]", "ae"},
{"[ᅣㅑ]", "ya"},
{"[ᅤㅒ]", "yae"},
{"[ᅥㅓ]", "ŏ"},
{"[ᅦㅔ]", "e"},
{"[ᅧㅕ]", "yŏ"},
{"[ᅨㅖ]", "ye"},
{"[ᅩㅗ]", "o"},
{"[ᅪㅘ]", "wa"},
{"[ᅫㅙ]", "wae"},
{"[ᅬㅚ]", "oe"},
{"[ᅭㅛ]", "yo"},
{"[ᅮㅜ]", "u"},
{"[ᅯㅝ]", "wŏ"},
{"[ᅰㅞ]", "we"},
{"[ᅱㅟ]", "wi"},
{"[ᅲㅠ]", "yu"},
{"[ᅳㅡ]", "ŭ"},
{"[ᅴㅢ]", "ŭi"},
{"[ᅵㅣ]", "i"},
-- e → ë when e is not part of digraphs ae & oe (e.g. ㅐ is ae, but ㅏ에 is aë)
{"([ao])ᄋe", "%1ë"},
-- to prevent input like 'ㅇ' (with ') from becoming italic markup (as [ᄋㅇ] is simply removed later)
{"'([ᄋㅇ]+)'", "'%1'"},
-- Consonants
{"`ᄀ", "g"},
{"`ᄃ", "d"},
{"`ᄇ", "b"},
{"`ᄌ", "j"},
{"`", ""},
{"[ᄀᆨㄱ]", "k"},
{"[ᄁㄲ]", "kk"},
{"ㄳ", "ks"},
{"[ᄂᆫㄴ]", "n"},
{"ㄵ", "nj"},
{"ㄶ", "nh"},
{"[ᄃᆮㄷ]", "t"},
{"[ᄄㄸ]", "tt"},
{"[ᄅㄹ]", "r"},
{"ᆯ", "l"},
{"ㄺ", "lg"},
{"ㄻ", "lm"},
{"ㄼ", "lb"},
{"ㄽ", "ls"},
{"ㄾ", "lt'"},
{"ㄿ", "lp'"},
{"ㅀ", "rh"},
{"[ᄆᆷㅁ]", "m"},
{"[ᄇᆸㅂ]", "p"},
{"[ᄈㅃ]", "pp"},
{"ㅄ", "ps"},
{"[ᄉㅅ]", "s"},
{"[ᄊㅆ]", "ss"},
{"[ᄋㅇ]", ""},
{"ᆼ", "ng"},
{"[ᄌㅈ]", "ch"},
{"[ᄍㅉ]", "tch"},
{"[ᄎㅊ]", "ch'"},
{"[ᄏㅋ]", "k'"},
{"[ᄐㅌ]", "t'"},
{"[ᄑㅍ]", "p'"},
{"[ᄒㅎ]", "h"}
}
return gsub_iterate(text, replacements)
end
return p
