Package {HanNLP}

Type: Package
Title: Korean Morphological Analysis and Hangul Utilities
Version: 0.1.1
Maintainer: Jonghyeon Park <jng.hyn.park@gmail.com>
Description: Provides Korean morphological analysis, part-of-speech (POS) tagging, noun extraction, Hangul (Korean script) conversion utilities, concordance search, mutual information statistics, and user dictionary management tools for Korean text research. The morphological analyzer backend is derived from the 'KoNLP' package and the 'HanNanum' analyzer, and is reimplemented in native C so that no Java runtime is required. Bundled dictionary and statistical resources allow analysis to work out of the box, and optional dictionaries can be loaded at runtime from the 'Sejong' and 'NIADic' packages when they are installed and loaded.
License: GPL-3
Copyright: See inst/COPYRIGHTS for bundled third-party code and data.
URL: https://github.com/ShapeLayer/HanNLP
BugReports: https://github.com/ShapeLayer/HanNLP/issues
Encoding: UTF-8
Depends: R (≥ 4.0.0)
Imports: utils
NeedsCompilation: yes
Packaged: 2026-09-21 09:22:36 UTC; runner
Author: Jonghyeon Park [aut, cre], Heewon Jeon [aut], Taekyung Kim [ctb], Semantic Web Research Center, KAIST [cph] (Copyright details and third-party contributors are documented in inst/COPYRIGHTS.)
Repository: CRAN
Date/Publication: 2026-09-30 09:40:19 UTC

Compose Hangul syllables from Jamo or keystrokes

Description

Compose complete Hangul syllables from compatibility Jamo or two-beolsik keystroke sequences.

Usage

HangulAutomata(input, isKeystroke = FALSE, isForceConv = FALSE)

Arguments

input

input Jamo sequences or two-beolsik keystrokes

isKeystroke

whether input is two-beolsik keystrokes instead of Jamo

isForceConv

force conversion for incomplete sequences

Value

complete Hangul syllable string

KAIST tag to Sejong tag

Description

KoNLP-compatible mapping from KAIST tags to Sejong tags.

Format

A named character vector whose names are KAIST tag names and whose values are the corresponding Sejong tag names.

Value

A named character vector mapping KAIST tags to Sejong tags.

Source

Derived from the tag mapping distributed with 'KoNLP'.

Hannanum morphological analyzer interface function

Description

Analyze Korean sentences using the native HanNLP analyzer backend.

Usage

MorphAnalyzer(sentences, autoSpacing = FALSE)

Arguments

sentences

input character vector

autoSpacing

retained for KoNLP API compatibility; currently ignored

Value

named list of eojeol to morpheme/tag strings for one sentence, list for multiple sentences

Examples

text <- intToUtf8(c(0xd55c, 0xae00, 0x20, 0xd615, 0xd0dc, 0xc18c))
MorphAnalyzer(text)

POS tagging by using 9 KAIST tags

Description

Tag Korean sentences using 9 KAIST tags with the native HanNLP analyzer backend.

Usage

SimplePos09(sentences, autoSpacing = FALSE)

Arguments

sentences

input character vector

autoSpacing

retained for KoNLP API compatibility; currently ignored

Value

named list of eojeol to morpheme/tag strings for one sentence, list for multiple sentences

POS tagging by using 22 KAIST tags

Description

Tag Korean sentences using 22 KAIST tags with the native HanNLP analyzer backend.

Usage

SimplePos22(sentences, autoSpacing = FALSE)

Arguments

sentences

input character vector

autoSpacing

retained for KoNLP API compatibility; currently ignored

Value

named list of eojeol to morpheme/tag strings for one sentence, list for multiple sentences

Sejong tag to KAIST tag

Description

KoNLP-compatible mapping from Sejong tags to KAIST tags.

Format

A named character vector whose names are Sejong tag names and whose values are the corresponding KAIST tag names.

Value

A named character vector mapping Sejong tags to KAIST tags.

Source

Derived from the tag mapping distributed with 'KoNLP'.

Backup current user dictionary

Description

Backup the writable HanNLP user dictionary.

Usage

backupUsrDic(ask = TRUE)

Arguments

ask

ask to confirm backup

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible logical

Build HanNLP user dictionary

Description

Build the writable HanNLP user dictionary from user-supplied entries and optional external sources. The Sejong dictionary is read from the installed Sejong package or the path specified by HANNLP_SEJONG_HANDIC. The woorimalsam and insighter dictionaries require the optional NIADic package to be installed and loaded (library(NIADic)) in the current session; HanNLP does not bundle NIADic data.

Usage

buildDictionary(
  ext_dic = "",
  category_dic_nms = "",
  user_dic = data.frame(),
  replace_usr_dic = FALSE,
  verbose = FALSE
)

Arguments

ext_dic

optional external dictionary names: sejong, woorimalsam, or insighter

category_dic_nms

retained for KoNLP API compatibility

user_dic

user dictionary data.frame with term and KAIST tag columns

replace_usr_dic

replace existing user dictionary instead of appending

verbose

print detail progress

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible data.frame of the resulting user dictionary

Concordance for input text file

Description

Return concordance text for an input file.

Usage

concordance_file(filename, pattern, encoding = getOption("encoding"), span = 5)

Arguments

filename

file name

pattern

pattern of central words

encoding

file encoding

span

number of characters to include around each match

Value

character vector of matched concordance strings

Concordance for input text vector

Description

Return concordance text for an input pattern and span.

Usage

concordance_str(string, pattern, span = 5)

Arguments

string

input text as character vector or single character

pattern

pattern of central words

span

number of characters to include around each match

Value

list of matched concordance strings, with empty elements removed

Convert Hangul string to Jamos

Description

Convert Hangul syllables to compatibility Jamo sequences.

Usage

convertHangulStringToJamos(hangul)

Arguments

hangul

Hangul string

Value

Jamo sequences

Examples

hangul <- intToUtf8(c(0xd55c, 0xae00))
convertHangulStringToJamos(hangul)

Convert Hangul string to keystrokes

Description

Convert Hangul syllables to two-beolsik keystroke sequences.

Usage

convertHangulStringToKeyStrokes(hangul, isFullwidth = TRUE)

Arguments

hangul

Hangul sentence

isFullwidth

specify returned character will be Fullwidth ASCII or Halfwidth ASCII

Value

Keystroke sequence

Tag name converter

Description

Convert tags between KAIST and Sejong tag sets.

Usage

convertTag(fromTag, toTag, tag)

Arguments

fromTag

tag set name to convert from, either K or S

toTag

desired tag set name, either K or S

tag

tag name to search

Value

converted tag vector

Examples

convertTag("K", "S", "ncn")

Keystroke misspell cost table

Description

KoNLP-compatible edit weight metadata. HanNLP does not use this table in runtime algorithms yet.

Format

A 66 by 66 numeric matrix of edit (substitution) costs indexed by keystroke characters on both dimensions.

Value

A 66 by 66 numeric matrix of keystroke edit costs.

Source

Derived from the edit weight table distributed with 'KoNLP'.

Noun extractor for Hangul

Description

Extract nouns from Korean sentences using the native HanNLP analyzer backend.

Usage

extractNoun(sentences, autoSpacing = FALSE)

Arguments

sentences

input character vector

autoSpacing

retained for KoNLP API compatibility; currently ignored

Value

character vector for one sentence, list for multiple sentences

Examples

local({
  old <- Sys.getenv("HANNLP_USER_DATA_DIR", unset = NA)
  tmp <- tempfile("hannlp-user-data-")
  on.exit({
    if (is.na(old)) Sys.unsetenv("HANNLP_USER_DATA_DIR") else Sys.setenv(HANNLP_USER_DATA_DIR = old)
    unlink(tmp, recursive = TRUE, force = TRUE)
  })
  Sys.setenv(HANNLP_USER_DATA_DIR = tmp)
  text <- intToUtf8(c(0xd55c, 0xae00, 0x20, 0xd615, 0xd0dc, 0xc18c))
  extractNoun(text)
})

Get Dictionary

Description

Return a HanNLP dictionary as a data.frame. The external Sejong dictionary requires an installed source package or a configured file path. The NIADic-based dictionaries require the loaded optional NIADic package.

Usage

get_dictionary(dic_name)

Arguments

dic_name

one of user_dic, system_dic, analyzed_dic, sejong, woorimalsam, or insighter. The woorimalsam and insighter dictionaries require the optional NIADic package to be installed and loaded (library(NIADic)) in the current session; HanNLP does not bundle NIADic data.

Value

data.frame with dictionary terms and tags

Check if sentence is all ASCII

Description

Check whether each input string contains only ASCII characters.

Usage

is.ascii(sentence)

Arguments

sentence

input characters

Value

logical vector

Check if sentence is all Hangul

Description

Check whether each input string contains only Hangul syllables or compatibility Jamo.

Usage

is.hangul(sentence)

Arguments

sentence

input characters

Value

logical vector

Examples

hangul <- intToUtf8(c(0xd55c, 0xae00))
is.hangul(c(hangul, "abc"))

Check if sentence is all Jaeum

Description

Check whether each input string contains only Korean consonant Jamo.

Usage

is.jaeum(sentence)

Arguments

sentence

input characters

Value

logical vector

Check if sentence is all Jamo

Description

Check whether each input string contains only compatibility Jamo.

Usage

is.jamo(sentence)

Arguments

sentence

input characters

Value

logical vector

Check if sentence is all Moeum

Description

Check whether each input string contains only Korean vowel Jamo.

Usage

is.moeum(sentence)

Arguments

sentence

input characters

Value

logical vector

Append or replace the HanNLP user dictionary

Description

Append to or replace the writable HanNLP user dictionary.

Usage

mergeUserDic(newUserDic, append = TRUE, verbose = FALSE, ask = FALSE)

Arguments

newUserDic

new user dictionary as data.frame

append

append to existing dictionary or replace it

verbose

retained for KoNLP API compatibility

ask

ask to backup current dictionary

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible data.frame of the resulting user dictionary

Mutual information for input text

Description

Return mutual information or t-scores for adjacent word bigrams.

Usage

mutualinformation(text, query = "", method = c("mutual", "tscores"))

Arguments

text

input character vector

query

term to filter bigrams

method

calculation method, either mutual or tscores

Value

named numeric vector

Reload all dictionaries

Description

HanNLP loads dictionaries per analysis call, so this function is retained for KoNLP API compatibility.

Usage

reloadAllDic()

Value

invisible TRUE

Reload user dictionaries

Description

Compatibility wrapper for reloading user dictionaries.

Usage

reloadUserDic(whichDics)

Arguments

whichDics

character vector of dictionary consumers

Value

invisible TRUE

Restore backed up user dictionary

Description

Restore the writable HanNLP user dictionary from backup.

Usage

restoreUsrDic(ask = TRUE)

Arguments

ask

ask to confirm restore

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible logical

Summary of dictionaries

Description

Show summary, head, and tail of current or backup user dictionaries.

Usage

statDic(which = "current", n = 6)

Arguments

which

one of current or backup

n

number of rows for head and tail

Value

list containing summary, head, and tail

Tag names

Description

KoNLP-compatible KAIST/HanNanum tag names used by HanNLP dictionary utilities.

Format

A named character vector mapping each supported tag name to itself, used to validate tag entries.

Value

A named character vector of supported tag names.

Source

Derived from the KAIST/HanNanum tag set distributed with 'KoNLP'.

Use NIA dictionary

Description

Replace the writable user dictionary with NIADic entries. HanNLP does not bundle or distribute NIADic data because it is released under CC BY-SA, which is incompatible with the GPL. This function works only when the optional NIADic package is installed and its namespace is already loaded in the current R session (e.g. after library(NIADic)); otherwise it errors.

Usage

useNIADic(
  which_dic = c("woorimalsam", "insighter"),
  category_dic_nms = "all",
  backup = TRUE
)

Arguments

which_dic

NIADic dictionaries to load: woorimalsam, insighter, or both

category_dic_nms

retained for KoNLP API compatibility

backup

backup current dictionary before switching

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible data.frame of the resulting user dictionary

Use bundled Sejong-style dictionary

Description

Replace the writable user dictionary with Sejong dictionary entries from an installed Sejong package or HANNLP_SEJONG_HANDIC.

Usage

useSejongDic(backup = TRUE)

Arguments

backup

backup current dictionary before switching

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible result

Use system default dictionary

Description

Restore the writable user dictionary to HanNLP's bundled default user dictionary.

Usage

useSystemDic(backup = TRUE)

Arguments

backup

backup current dictionary before switching

Details

Before writing dictionary files, explicitly set the environment variable HANNLP_USER_DATA_DIR to the desired directory. There is no default writable directory. Backups are stored in its ‘backup’ subdirectory. For temporary use, choose a directory under tempdir().

Value

invisible TRUE