Message1527

Hi, I'm Japanese.
assume XX is Japanese string and its code are u'\u3042'
YY is another one u'\u304b'.

exec() destroys this string to meanless bytes.

>>> exec("print [hex(ord(x)) for x in u'XX']")
['0x3f', '0x3f']
>>> exec("print [hex(ord(x)) for x in u'YY']")
['0x3f', '0x3f']

if I don't use multibyte string, It works well

>>> exec("print [hex(ord(x)) for x in u'\u3042']")
['0x3042']

and printing multibyte also works well

>>> print u"XX"
??
>>> print u"\u3042"
XX

so I think the bug is in parsing process of Jython code.