pickletools.py 文件源码

def read_unicodestring4(f):
    r"""
    # bug 24549
    #>>> import StringIO
    #>>> s = u'abcd\uabcd'
    #>>> enc = s.encode('utf-8')
    #>>> enc
    #'abcd\xea\xaf\x8d'
    #>>> n = chr(len(enc)) + chr(0) * 3  # little-endian 4-byte length
    #>>> t = read_unicodestring4(StringIO.StringIO(n + enc + 'junk'))
    #>>> s == t
    #True
    #
    #>>> read_unicodestring4(StringIO.StringIO(n + enc[:-1]))
    #Traceback (most recent call last):
    #...
    #ValueError: expected 7 bytes in a unicodestring4, but only 6 remain
    """

    n = read_int4(f)
    if n < 0:
        raise ValueError("unicodestring4 byte count < 0: %d" % n)
    data = f.read(n)
    if len(data) == n:
        return unicode(data, 'utf-8')
    raise ValueError("expected %d bytes in a unicodestring4, but only %d "
                     "remain" % (n, len(data)))