Character data

Every method returns entries of the same shape. This page describes each field, what its values mean, and when it can be empty.

Fields#

{
  literal: '日',
  meanings: ['day', 'sun', 'Japan', 'counter for days'],
  onyomi: ['ニチ', 'ジツ'],
  kunyomi: ['ひ', '-び', '-か'],
  stroke_count: 4,
  grade: 1,
  jlpt: 4,
  freq: 1
}

Field	Type	Description
literal	string	The character itself. Every entry has a unique one.
meanings	string[]	English meanings. Empty for 2,761 entries.
onyomi	string[]	On (Sino-Japanese) readings, in katakana. Empty for 960 entries.
kunyomi	string[]	Kun (native Japanese) readings, in hiragana. Empty for 3,309 entries.
stroke_count	number	Number of strokes, from 1 to 34.
grade	number \| null	School grade or official list. See Grade.
jlpt	number \| null	JLPT level, old scale. See JLPT level.
freq	number \| null	Newspaper frequency rank. See Frequency.

Grade#

grade shows where a character sits in the Japanese government's kanji lists. The value 7 isn't used.

Value	List	Meaning	Entries
1–6	Kyōiku kanji	Taught in elementary school, grade 1 to 6	1,006
8	Jōyō kanji	The rest of the general-use list, taught in secondary school	1,130
9	Jinmeiyō kanji	Approved for personal names	684
10	Jinmeiyō variants	Name kanji that are variant forms of jōyō kanji	177
null	None	Not on any official list	10,111

Grades 1–6 and 8 together make up the 2,136 jōyō (general-use) kanji. To get them all, use search({ grade: { min: 1, max: 8 } }).

JLPT level#

jlpt uses the old four-level scale that the Japanese-Language Proficiency Test used before 2010. Higher numbers are easier. This is the opposite of today's N1–N5 order, where N5 is the easiest. The current test has no official kanji lists, so the mapping below is only approximate.

Value	Meaning	Entries
4	Easiest. Roughly today's N5	103
3	Roughly today's N4	181
2	Roughly today's N3 and N2	739
1	Hardest. Roughly today's N1	1,207
null	Not on a JLPT list	10,878

Frequency#

freq ranks the 2,501 most-used characters by how often they appear in Japanese newspapers. 1 is the most frequent (日). Every other entry has null. Sort by it to put common characters first:

const byFrequency = (a, b) => (a.freq ?? Infinity) - (b.freq ?? Infinity)

Kanji.search({ jlpt: 4 }).sort(byFrequency).slice(0, 10).map(k => k.literal).join('')
// '日一国会人年大十二本'

Readings#

onyomi entries are in katakana, and kunyomi entries are in hiragana.
A hyphen marks a reading that only appears as part of a word. A leading hyphen means a suffix, as in -び for 日. A trailing hyphen means a prefix, as in なま- for 生. 281 entries have one.
Kun readings don't mark okurigana (the kana that follow the kanji), so いきる for 生 is stored as one string.
A reading can appear more than once in the same list.

The romaji search matches against these lists, hyphens included.

Source and license#

The data comes from KANJIDIC2, compiled by Jim Breen and the Electronic Dictionary Research and Development Group (EDRDG). kanji.js includes a subset of its fields. KANJIDIC2 is available under the Creative Commons Attribution-ShareAlike 4.0 license. If you redistribute the data, follow the EDRDG licence terms.