jaconv (Japanese Converter)

Install the jaconv package using pip. Since it is implemented entirely in Python, it can be used in environments where a C compiler is unavailable.

$ pip install jaconv

The normalize(text, form) method expands unicodedata.normalize specifically for Japanese language processing. It handles common character variations like different types of dashes, hyphens, and quotation marks to ensure consistency.

import jaconv

jaconv.normalize('ティロ･フィナ〜レ', 'NFKC')
# => 'ティロ・フィナーレ'

The normalize(text, form) method is based on Python's unicodedata.normalize but includes specialized extensions for Japanese text processing. It automatically maps various Japanese-specific symbols (like different types of dashes, hyphens, and quotation marks) to standard forms.

Common mappings include:

• 〜 or ～ $\rightarrow$ ー (Chōonpu)
• Various dashes (―, ‐, ‑, etc.) $\rightarrow$ - or ー depending on context.
• Various quotation marks $\rightarrow$ standard ASCII quotes.

import jaconv

jaconv.normalize('ティロ･フィナ〜レ', 'NFKC')
# => 'ティロ・フィナーレ'

The z2h method (alias: zenkaku2hankaku) converts full-width characters to half-width. You can control which character types are converted using boolean flags:

• kana: Convert full-width kana to half-width kana (default behavior).
• ascii: Convert full-width ASCII to half-width ASCII.
• digit: Convert full-width digits to half-width digits.

import jaconv

# Full-width kana to Half-width kana
jaconv.z2h('ティロ・フィナーレ')
# => 'ﾃｨﾛ･ﾌｨﾅｰﾚ'

# Full-width ASCII to Half-width ASCII
jaconv.z2h('ａｂｃ', kana=False, ascii=True, digit=False)
# => 'abc'

# Full-width digits to Half-width digits
jaconv.z2h('１２３', kana=False, ascii=False, digit=True)
# => '123'

# Mixed conversion
jaconv.z2h('アａｂｃ１２３', kana=False, digit=True, ascii=True)
# => 'アabc123'

# Alias usage
jaconv.zenkaku2hankaku('ティロ・フィナーレａｂｃ１２３')
# => 'ﾃｨﾛ･ﾌｨﾅｰﾚａｂｃ１２３'

Use these methods to transliterate between Japanese kana and the Latin alphabet:

• kana2alphabet(text): Converts Hiragana to Alphabet.
• alphabet2kana(text): Converts Alphabet to Hiragana.
• kata2alphabet(text): Converts Katakana to Alphabet.
• alphabet2kata(text): Converts Alphabet to Katakana.

import jaconv

jaconv.kana2alphabet('じゃぱん')
# => 'japan'

jaconv.alphabet2kana('japan')
# => 'じゃぱん'

jaconv.kata2alphabet('ケツイ')
# => 'ketsui'

jaconv.alphabet2kata('namba')
# => 'ナンバ'

Use hiragana2julius(text) to convert Hiragana into Julius's phoneme format.

import jaconv

jaconv.hiragana2julius('てんきすごくいいいいいい')
# => 't e N k i s u g o k u i:'

The enlarge_smallkana(text, ignore=None) method converts small kana (like ぁ, ぃ, ぅ, ぇ, ぉ, っ, ゃ, ゅ, ょ) into their standard large counterparts.

Use the ignore parameter to specify characters that should not be enlarged.

import jaconv

# Standard enlargement
jaconv.enlarge_smallkana('わぁい')
# => 'わあい'

# Enlargement with ignore
jaconv.enlarge_smallkana('きょういっぱい', ignore='っ')
# => 'きよういっぱい'

Use the following methods to convert between Hiragana and Katakana character sets:

• hira2kata(text): Converts Hiragana to Katakana.
• hira2hkata(text): Converts Hiragana to Half-width Katakana.
• kata2hira(text): Converts Katakana to Hiragana.

import jaconv

# Hiragana to Katakana
jaconv.hira2kata('ともえまみ')
# => 'トモエマミ'

# Hiragana to Half-width Katakana
jaconv.hira2hkata('ともえまみ')
# => 'ﾄﾓｴﾏﾐ'

# Katakana to Hiragana
jaconv.kata2hira('巴マミ')
# => '巴まみ'

Use enlarge_smallkana(text, ignore=None) to convert small Hiragana or Katakana characters (like ぁ, ぃ, ぅ, ぇ, ぉ, っ, etc.) to their normal size versions.

Use the ignore parameter to specify characters that should remain small.

import jaconv

jaconv.enlarge_smallkana('わぁい')
# => 'わあい'

jaconv.enlarge_smallkana('きょういっぱい', ignore='っ')
# => 'きよういっぱい'

The hiragana2julius(text) method converts Hiragana text into the Julius phonetic transcription format.

import jaconv

jaconv.hiragana2julius('てんきすごくいいいいいい')
# => 't e N k i s u g o k u i:'