Regex for Japanese

regex-japanese.txt

Regex for matching ALL Japanese common & uncommon Kanji (4e00 – 9fcf) ~ The Big Kahuna!
([一-龯])

Regex for matching Hirgana or Katakana
([ぁ-んァ-ン])

Regex for matching Non-Hirgana or Non-Katakana
([^ぁ-んァ-ン])

Regex for matching Hirgana or Katakana or basic punctuation (、。’)
([ぁ-んァ-ン\w])

Regex for matching Hirgana or Katakana and random other characters
([ぁ-んァ-ン！：／])

Regex for matching Hirgana
([ぁ-ん])

Regex for matching full-width Katakana (zenkaku 全角)
([ァ-ン])

Regex for matching half-width Katakana (hankaku 半角)
([ｧ-ﾝﾞﾟ])

Regex for matching full-width Numbers (zenkaku 全角)
([０-９])

Regex for matching full-width Letters (zenkaku 全角)
([Ａ-ｚ])

Regex for matching Hiragana codespace characters (includes non phonetic characters)
([ぁ-ゞ])

Regex for matching full-width (zenkaku) Katakana codespace characters (includes non phonetic characters)
([ァ-ヶ])

Regex for matching half-width (hankaku) Katakana codespace characters (this is an old character set so the order is inconsistent with the hiragana)
([ｦ-ﾟ])

Regex for matching Japanese Post Codes
/^¥d{3}¥-¥d{4}$/
/^¥d{3}-¥d{4}$|^¥d{3}-¥d{2}$|^¥d{3}$/

Regex for matching Japanese mobile phone numbers (keitai bangou)
/^¥d{3}-¥d{4}-¥d{4}$|^¥d{11}$/
/^0¥d0-¥d{4}-¥d{4}$/

Regex for matching Japanese fixed line phone numbers
/^[0-9-]{6,9}$|^[0-9-]{12}$/
/^¥d{1,4}-¥d{4}$|^¥d{2,5}-¥d{1,4}-¥d{4}$/

Credit to:
http://crunchytoast.com/2009/12/12/japanese-regex-alzheimers-and-why-cant-i-remember/

Actually the regexes for hiragana and katakana are too restrictive. e.g. the katakana regex would not match words containing the 長音符, such as ユーザー and コンピューター.

Here are some more comprehensive regexes for hiragana and katakana:

Hiragana = [ぁ-ゔゞ゛゜ー]  // 0x3041-0x3094, 0x309E, 0x309B, 0x309C, 0x30FC
Katakana = [ァ-・ヽヾ゛゜ー]  // 0x30A1-0x30FB, 0x30FD, &#x30FE, 0x309B, 0x309C, 0x30FC
Hiragana or katakana = [ぁ-ゔゞァ-・ヽヾ゛゜ー]  // 0x3041-0x3094, 0x309E, 0x30A1-0x30FB, 0x30FD, &#x30FE, 0x309B, 0x309C, 0x30FC

Note that the last 3 chars in each set (0x309B, 0x309C, 0x30FC) are shared between hiragana and katakana).

Just updated full-width Katakana from「30A1」～「30FE」 (Unicode：30FB).

Regex for matching full-width Katakana (zenkaku 全角)
([ァ-ン])

Replace to:

([ァ-ヾ])

reg for combination of hiragana, katakana, number?

Warning (this happened to myself): if you remember the range of kanji by heart ("ichi (一) kara masakari (鉞) made"), you'll get it wrong. The kanji here is not the "common" masakari 鉞 but an uncommon variant 龯, so copy-paste it to get it right!

This doesn't cover all kanjis. Simple example: 𧓈

There is a much easier way to do this:

/\p{Script=Han}|\p{Script=Katakana}|\p{Script=Hiragana}/u

see https://www.regular-expressions.info/unicode.html #Unicode Scripts

Japanese imperial date regex:
Example: 令和5年2月 24 日

([令和|平成|昭和|大正|明治]{2})(\d+)年[\s]?(\d{1,2})[\s]?月[\s]?(\d{1,2})[\s]?日

@Jaha96 That doesn't catch all: the first year is commonly marked as 元年 instead of 1年. 令和元年 = year 2019, for example. This case was widely disregarded in many libraries, but in actual life, it was very common to see it written that way.

I'm working on Android and \d matches ０(U+FF10), too.

This doesn't cover all kanjis. Simple example: 𧓈

To be fair those kanjis are extremely rare and are not used (they would not show up in dictionnaires or rikaichan like extensions) and 99.99% Japanese would not know about them:
https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_Extension_B

Now you can match them with: [𠀀-𪛟]
and to match everything you would simply do: [𠀀-𪛟]|[一-龯]

@cb372, your list comes close to covering all the kana, but a few characters are still missing. You got 「ゞ」 but missed 「ゝ」 and 「ゟ」, and a few others. I believe this would cover all Hiragana and Katakana separately:

Hiragana = [ぁ-ゖ゛-ゟー]
Katakana = [゠-ヿ]

Combined Hiragana & Katakana would be:

Hiragana+Katakana = [ぁ-ゖ゛-ゟ゠-ヿ]

I used the above hiragana+katakana regex to validate the kana portions of the downloadable version of JMDICT and can confirm that apart from a few errors in the JMDICT data, the kana validation works.