kr.bydelta.koala.package.scala Maven / Gradle / Ivy
package kr.bydelta
import scala.annotation.tailrec
/**
* 통합 Scala 한국어 분석기, Koala
*
* 각 Package의 Parser와 Tagger를 참조하세요.
*/
package object koala {
/** '가' 위치 **/
final val HANGUL_START: Int = '가'
/** '힣' 위치 **/
final val HANGUL_END: Int = '힣'
/** 종성 범위 **/
final val JONGSUNG_RANGE = 0x001C
/** 중성 범위 **/
final val JUNGSUNG_RANGE = 0x024C
protected[koala] final val ALPHA_PRON = Seq(
"에이치", "더블유", "에이", "에프", "아이", "제이", "케이",
"에스", "브이", "엑스", "와이", "제트", "비", "씨", "디",
"이", "지", "엘", "엠", "엔", "오", "피", "큐", "알",
"티", "유"
)
protected[koala] final val ALPHA_PRON_ORDER = Seq(
'H', 'W', 'A', 'F', 'I', 'J', 'K', 'S', 'V', 'X', 'Y', 'Z',
'B', 'C', 'D', 'E', 'G', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'T', 'U'
)
/**
* 통합 의존구문분석표기로 변경.
*
* @param tag 통합표기로 변경할 의존구문분석 결과.
* @return 통합표기.
*/
def KKMAdepTag(tag: String): FunctionalTag.Value = {
tag match {
case "목적어" | "(주어,목적)대상" => FunctionalTag.Object
case "주어" => FunctionalTag.Subject
case "부사어" | "이유" => FunctionalTag.Adjunct
case "보어" | "인용" => FunctionalTag.Complement
case "수식" => FunctionalTag.Modifier
case "동일" | "보조 연결" | "의존 연결" | "대등 연결" | "체언 연결" | "연결" => FunctionalTag.Conjunctive
case _ =>
FunctionalTag.Undefined
}
}
/**
* 통합 의존구문분석표기로 변경.
*
* @param tag 통합표기로 변경할 의존구문분석 결과.
* @return 통합표기.
*/
def HNNdepTag(tag: String): FunctionalTag.Value = {
tag match {
case "SBJ" => FunctionalTag.Subject
case "OBJ" => FunctionalTag.Object
case "CMP" => FunctionalTag.Complement
case "VMOD" | "NMOD" | "MOD" => FunctionalTag.Modifier
case "ADV" | "AJT" => FunctionalTag.Adjunct
case "CNJ" => FunctionalTag.Conjunctive
case "INT" => FunctionalTag.Interjective
case "PRN" => FunctionalTag.Parenthetical
case _ => FunctionalTag.Undefined
}
}
/**
* 원본품사로 변환.
*
* @param tag 원본품사로 변환할 통합표기.
* @return 변환된 품사.
*/
def tagToKomoran(tag: _root_.kr.bydelta.koala.POS.Value): String = {
tag match {
case POS.NNM => "NNB"
case POS.XSM | POS.XSO => "XSN"
case POS.XPV => "XR"
case POS.SY => "SW"
case POS.UN | POS.UV | POS.UE => "NA"
case _ => tag.toString
}
}
/**
* 원본품사로 변환.
*
* @param tag 원본품사로 변환할 통합표기.
* @return 변환된 품사.
*/
def tagToTwt(tag: _root_.kr.bydelta.koala.POS.Value): String = {
tag match {
case POS.NNG | POS.NNB |
POS.NNM | POS.NP => "Noun"
case POS.NNP => "ProperNoun"
case POS.NR | POS.SN => "Number"
case POS.VV | POS.VX |
POS.VCP | POS.VCN => "Verb"
case POS.VA => "Adjective"
case POS.MM => "Determiner"
case POS.MAG | POS.MAJ => "Adverb"
case POS.IC => "Exclamation"
case POS.JKB | POS.JKC |
POS.JKG | POS.JKO |
POS.JKQ | POS.JKS |
POS.JKV | POS.JX => "Josa"
case POS.JC => "Conjunction"
case POS.EP => "PreEomi"
case POS.EF | POS.EC |
POS.ETM | POS.ETN => "Eomi"
case POS.XPN => "NounPrefix"
case POS.XPV => "VerbPrefix"
case POS.XSA | POS.XSM |
POS.XSN | POS.XSO | POS.XSV => "Suffix"
case POS.SF => "Punctuation"
case POS.SS | POS.SP |
POS.SE | POS.SY | POS.XR => "Others"
case POS.UE | POS.UN | POS.UV => "Unknown"
case POS.SL => "Foreign"
}
}
/**
* 원본품사로 변환.
*
* @param tag 원본품사로 변환할 통합표기.
* @return 변환된 품사.
*/
def tagToEunjeon(tag: _root_.kr.bydelta.koala.POS.Value): String = {
tag match {
case POS.NNM => "NNBC"
case POS.SS => "SSO"
case POS.SP => "SC"
case POS.XPV => "XR"
case POS.XSM | POS.XSO => "XSN"
case POS.UN | POS.UE | POS.UV => "UNKNOWN"
case x => x.toString
}
}
/**
* 원본품사로 변환.
*
* @param tag 원본품사로 변환할 통합표기.
* @return 변환된 품사.
*/
def tagToHNN(tag: _root_.kr.bydelta.koala.POS.Value): String = {
(tag match {
case POS.NNG | POS.UN => "NCN"
case POS.NNP => "NQQ"
case POS.NNB => "NBN"
case POS.NNM => "NBU"
case POS.NR => "NNC"
case POS.NP => "NPD"
case POS.VV | POS.UV => "PVG"
case POS.VA => "PAA"
case POS.VX => "PX"
case POS.VCP => "JP"
case POS.VCN => "PAA"
case POS.MM => "MMA"
case POS.IC => "II"
case POS.JKS => "JCS"
case POS.JKC => "JCC"
case POS.JKG => "JCM"
case POS.JKO => "JCO"
case POS.JKB => "JCA"
case POS.JKV => "JCV"
case POS.JKQ => "JCR"
case POS.JC => "JCT"
case POS.JX => "JXF"
case POS.EC => "ECC"
case POS.XSO | POS.XSN => "XSNX"
case POS.XSV => "XSVN"
case POS.XSA => "XSMN"
case POS.XSM => "XSAS"
case POS.SS => "SL"
case POS.SL => "F"
case POS.SN => "NNC"
case POS.SY | POS.XR | POS.UE => "SY"
case POS.XPN | POS.XPV => "XP"
case x => x.toString
}).toLowerCase
}
/**
* 원본품사로 변환.
*
* @param tag 원본품사로 변환할 통합표기.
* @return 변환된 품사.
*/
def tagToKKMA(tag: _root_.kr.bydelta.koala.POS.Value): String = {
tag match {
case POS.VX => "VXV"
case POS.MM => "MDT"
case POS.MAJ => "MAC"
case POS.JKB => "JKM"
case POS.JKV => "JKI"
case POS.EP => "EPT"
case POS.EF => "EFN"
case POS.EC => "ECE"
case POS.ETM => "ETD"
case POS.SY => "SW"
case POS.SL => "OL"
case POS.SN => "ON"
case x => x.toString
}
}
/**
* 통합품사로 변환.
*
* @param tag 통합품사로 변환할 원본표기.
* @return 변환된 통합품사.
*/
def fromKomoranTag(tag: String): POS.Value = {
tag match {
case "SW" | "SO" => POS.SY
case "NA" => POS.UE
case "SL" | "SH" => POS.SL
case _ => POS withName tag
}
}
/**
* 통합품사로 변환.
*
* @param tag 통합품사로 변환할 원본표기.
* @return 변환된 통합품사.
*/
def fromTwtTag(tag: String): POS.Value = {
tag match {
case "Noun" => POS.NNG
case "ProperNoun" => POS.NNP
case "Number" => POS.NR
case "Verb" => POS.VV
case "Adjective" => POS.VA
case "Determiner" => POS.MM
case "Adverb" => POS.MAG
case "Exclamation" => POS.IC
case "Josa" => POS.JX
case "Conjunction" => POS.JC
case "PreEomi" => POS.EP
case "Eomi" => POS.EF
case "NounPrefix" => POS.XPN
case "VerbPrefix" => POS.XPV
case "Suffix" => POS.XSO
case "Punctuation" => POS.SF
case "Unknown" => POS.UE
case "Foreign" | "Alpha" => POS.SL
case _ => POS.SY
}
}
/**
* 통합품사로 변환.
*
* @param tag 통합품사로 변환할 원본표기.
* @return 변환된 통합품사.
*/
def fromEunjeonTag(tag: String): POS.Value = {
tag.toUpperCase match {
case "NNBC" => POS.NNM
case "SC" => POS.SP
case "SSO" | "SSC" => POS.SS
case "SL" | "SH" => POS.SL
case "UNKNOWN" => POS.UE
case x => POS withName x
}
}
/**
* 통합품사로 변환.
*
* @param tag 통합품사로 변환할 원본표기.
* @return 변환된 통합품사.
*/
def fromHNNTag(tag: String): POS.Value = {
tag match {
case x if x startsWith "nc" => POS.NNG
case x if x startsWith "nq" => POS.NNP
case "nbn" | "nbs" => POS.NNB
case "nbu" => POS.NNM
case x if x startsWith "nn" => POS.NR
case x if x startsWith "np" => POS.NP
case x if x startsWith "pv" => POS.VV
case x if x startsWith "pa" => POS.VA
case "px" => POS.VX
case "jp" => POS.VCP
case x if x startsWith "mm" => POS.MM
case "mad" | "mag" => POS.MAG
case "maj" => POS.MAJ
case "ii" => POS.IC
case "jcs" => POS.JKS
case "jcc" => POS.JKC
case "jcm" => POS.JKG
case "jco" => POS.JKO
case "jca" => POS.JKB
case "jcv" => POS.JKV
case "jcr" => POS.JKQ
case "jct" | "jcj" => POS.JC
case "jxc" | "jxf" => POS.JX
case x if x startsWith "ec" => POS.EC
case "xp" => POS.XPN
case x if x startsWith "xsn" => POS.XSN
case x if x startsWith "xsv" => POS.XSV
case x if x startsWith "xsm" => POS.XSA
case x if x startsWith "xsa" => POS.XSM
case "sl" | "sr" => POS.SS
case "sd" | "su" | "sy" => POS.SY
case "f" => POS.SL
case x => POS withName x.toUpperCase
}
}
/**
* 통합품사로 변환.
*
* @param tag 통합품사로 변환할 원본표기.
* @return 변환된 통합품사.
*/
def fromKKMATag(tag: String): POS.Value = {
tag.toUpperCase match {
case "VXV" | "VXA" => POS.VX
case "MDT" | "MDN" => POS.MM
case "MAC" => POS.MAJ
case "JKM" => POS.JKB
case "JKI" => POS.JKV
case x if x startsWith "EP" => POS.EP
case x if x startsWith "EF" => POS.EF
case x if x startsWith "EC" => POS.EC
case "ETD" => POS.ETM
case "SO" | "SW" => POS.SY
case "OL" | "OH" => POS.SL
case "ON" => POS.SN
case "EMO" => POS.SY //Emoticons
case x => POS withName x
}
}
@tailrec
final def pronounceAlphabet(x: String, acc: String = ""): String =
if (x.isEmpty) acc
else {
pronounceAlphabet(x.tail, acc + ALPHA_PRON(ALPHA_PRON_ORDER.indexOf(x.head)))
}
@tailrec
final def writeAlphabet(x: String, acc: String = ""): String =
if (x.isEmpty) acc
else {
ALPHA_PRON.indexWhere(x.startsWith) match {
case -1 => acc + x
case ind => writeAlphabet(x.substring(ALPHA_PRON(ind).length), acc + ALPHA_PRON_ORDER(ind))
}
}
@tailrec
final def isAlphabetPronounced(word: String): Boolean =
if (word.isEmpty) true
else {
ALPHA_PRON.find(word.startsWith) match {
case Some(p) => isAlphabetPronounced(word.substring(p.length))
case None => false
}
}
def reconstructKorean(cho: Int = 11, jung: Int = 0, jong: Int = 0): Char =
(HANGUL_START + cho * JUNGSUNG_RANGE + jung * JONGSUNG_RANGE + jong).toChar
/**
* 한국어를 구성하는 문자의 연산.
*
* @param ch 검사할 문자.
*/
implicit class KoreanCharacterExtension(ch: Char) {
/**
* (Code modified from Seunjeon package)
* 종성으로 끝나는지 확인.
*
* @return 종성으로 끝난다면, true를, 없다면 false를 반환.
*/
def endsWithJongsung = getJongsungCode != 0
/**
* (Code modified from Seunjeon package)
* 종성 종료 코드
*
* @return 종성으로 끝난다면, 해당 위치를, 없다면 0을 반환.
*/
def getJongsungCode = (ch - HANGUL_START) % JONGSUNG_RANGE
/**
* (Code modified from Seunjeon package)
* 중성 종료 코드
*
* @return 중성으로 끝난다면, 해당 위치를, 없다면 0을 반환.
*/
def getJungsungCode = (ch - HANGUL_START) % JUNGSUNG_RANGE / JONGSUNG_RANGE
/**
* (Code modified from Seunjeon package)
* 초성 종료 코드
*
* @return 초성으로 끝난다면, 해당 위치를, 없다면 0을 반환.
*/
def getChosungCode = (ch - HANGUL_START) / JUNGSUNG_RANGE
/**
* (Code modified from Seunjeon package)
* 한글 문자로 끝나는 지 확인.
*
* @return True: 종료 문자가 한글일 경우.
*/
def isHangul = {
((HANGUL_START <= ch && ch <= HANGUL_END)
|| (0x1100 <= ch && ch <= 0x11FF)
|| (0x3130 <= ch && ch <= 0x318F))
}
/**
* (Code modified from Seunjeon package)
* 한글의 불완전 문자인지 확인
*
* @return True: 종료 문자가 한글일 경우.
*/
def isIncompleteHangul = {
((0x1100 <= ch && ch <= 0x11FF)
|| (0x3130 <= ch && ch <= 0x318F))
}
/**
* 완성된 문자의 범위인지 확인.
*
* @return True: 완성된 문자일 경우.
*/
def isCompleteHangul = HANGUL_START <= ch && ch <= HANGUL_END
}
/**
* 한국어를 구성하는 문자열의 연산.
*
* @param word 검사할 문자열.
*/
implicit class KoreanStringExtension(word: String) {
/**
* (Code modified from Seunjeon package)
* 종성으로 끝나는지 확인.
*
* @return 종성으로 끝난다면, true를, 없다면 false를 반환.
*/
def endsWithJongsung = word.last.endsWithJongsung
/**
* (Code modified from Seunjeon package)
* 한글 문자로 끝나는 지 확인.
*
* @return True: 종료 문자가 한글일 경우.
*/
def endsWithHangul = word.last.isHangul
def filterNonHangul = {
word.replaceAll("(?U)[^가-힣\\s]+", " ").trim.split("(?U)\\s+").toSeq
}
}
}© 2015 - 2025 Weber Informatics LLC | Privacy Policy