pypy/interpreter/test/test_unicodehelper.py::test_latin1_shortcut_bug

space = StdObjSpace

    def test_latin1_shortcut_bug(space):
        state = space.fromcache(CodecState)
        handler = state.encode_error_handler
    
        sin = u"a\xac\u1234\u20ac\u8000"
        sin_utf8 = sin.encode("utf-8")
>       assert utf8_encode_latin_1(space, sin_utf8, space.newtext(sin_utf8), "backslashreplace", handler) == sin.encode("latin-1", "backslashreplace")

interpreter/test/test_unicodehelper.py:92: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
objspace/std/objspace.py:413: in newtext
    return self._newtext_memo(s)
objspace/std/objspace.py:432: in _newtext_memo
    return W_UnicodeObject(s, lgt)
<609-codegen /Users/matti/build-worker-x86_64/own-macos-x86-64/build/rpython/tool/sourcetools.py:318>:5: in __init__
    return __init___original(self, utf8str, length)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = W_UnicodeObject('a\xc2\xac\xe1\x88\xb4\xe2\x82\xac\xe8\x80\x80')
utf8str = 'a\xc2\xac\xe1\x88\xb4\xe2\x82\xac\xe8\x80\x80', length = 12

    @enforceargs(utf8str=str)
    def __init__(self, utf8str, length):
        assert isinstance(utf8str, bytes)
        # TODO: how to handle surrogates
        assert length >= 0
        self._utf8 = utf8str
        self._length = length
        self._index_storage = rutf8.null_storage()
        if CHECK_ALL_STRINGS or not we_are_translated():
            # utf8str must always be a valid utf8 string, except maybe with
            # explicit surrogate characters---which .decode('utf-8') doesn't
            # special-case in Python 2, which is exactly what we want here
            try:
                if sys.maxunicode == 0xffff:
                    # can't use .decode('utf-8') because it will add surrogates
                    real_length = rutf8.check_utf8(utf8str, True)
                else:
                    real_length = len(utf8str.decode('utf-8'))
            except (rutf8.CheckError, UnicodeDecodeError):
                real_length = -999
            if length != real_length:
                from rpython.rlib.debug import debug_print
                debug_print("!!! BAD UTF8 !!!")
                debug_print(str([ord(c) for c in utf8str]))
                debug_print("length", length, "real_length", real_length)
>               raise BadUtf8
E               BadUtf8

objspace/std/unicodeobject.py:86: BadUtf8
---------- Captured stderr call ----------
!!! BAD UTF8 !!! 
[97, 194, 172, 225, 136, 180, 226, 130, 172, 232, 128, 128] 
length 12 real_length 5 

builder: own-macos-x86-64 build #1268+

test: pypy/interpreter/test/test_unicodehelper.py::test_latin1_shortcut_bug