normalize_kana

gistfile1.py

# -*- coding: utf-8 -*-

FULL_KATA_OFFSET = ord(u"ァ") - ord(u"ぁ")
half_kata_to_hira = u"をぁぃぅぇぉゃゅょっーあいうえおかきくけこさしすせそたちつてとなにぬねのはひふへほまみむめもやゆよ"

def normalize_kana(s):
    """Normalizes all kana in the given string to full-width hiragana.

    Args:
        s       String to be normalized.
    """
    # Converting to a list allows us to edit the string in-place.
    rv = list(s)

    for i in xrange(len(rv)):
        ch = rv[i]
        if not ch:
            continue
        ordch = ord(ch) 
        if ordch in xrange(ord(u"ァ"), ord(u"ン") + 1):
            # Full-width katakana
            rv[i] = unichr(ordch - FULL_KATA_OFFSET)

        elif ordch in xrange(ord(u"ｦ"), ord(u"ﾝ") + 1):
            # Half-width katakana.  This unicode block is laid out differently.
            hira = half_kata_to_hira[ordch - ord(u"ｦ")]
            if i < len(rv) - 1:
                if rv[i + 1] == u"ﾞ":
                    hira = unichr(ord(hira) + 1)
                    rv[i + 1] = u""
                elif rv[i + 1] == u"ﾟ":
                    hira = unichr(ord(hira) + 2)
                    rv[i + 1] = u""
            rv[i] = hira

    return ''.join(rv)


if __name__ == '__main__':
    
    assert normalize_kana(u"テスト") == u"てすと"
    assert normalize_kana(u"リセット") == u"りせっと"
    assert normalize_kana(u"ﾎﾟ") == u"ぽ"
    assert normalize_kana(u"ｦ") == u"を"

    print "tests pass"