def cjk(str):
    '''
    rudimentary CJK detection
    seems like a lot of it is between these two ranges.
    cf. http://www.alanwood.net/unicode/fontsbyrange.html
    '''
    ranges = [(12272,12287), (19968, 40959)]
    for i in str:
        j = ord(i)
        for (low, high) in ranges:
            if j > low and j < high:
                return True
    return False