Fix and deprecated the unicode_internal codec

unicode_internal codec uses Py_UNICODE instead of the real internal
representation (PEP 393: Py_UCS1, Py_UCS2 or Py_UCS4) for backward
compatibility.
This commit is contained in:
Victor Stinner 2011-11-10 20:56:30 +01:00
parent 240c55f721
commit 9f4b1e9c50
4 changed files with 42 additions and 10 deletions

View file

@ -6237,6 +6237,11 @@ _PyUnicode_DecodeUnicodeInternal(const char *s,
PyObject *errorHandler = NULL;
PyObject *exc = NULL;
if (PyErr_WarnEx(PyExc_DeprecationWarning,
"unicode_internal codecs has been deprecated",
1))
return NULL;
/* XXX overflow detection missing */
v = PyUnicode_New((size+Py_UNICODE_SIZE-1)/ Py_UNICODE_SIZE, 127);
if (v == NULL)
@ -6270,15 +6275,26 @@ _PyUnicode_DecodeUnicodeInternal(const char *s,
errors, &errorHandler,
"unicode_internal", reason,
&starts, &end, &startinpos, &endinpos, &exc, &s,
&v, &outpos)) {
&v, &outpos))
goto onError;
continue;
}
s += Py_UNICODE_SIZE;
#ifndef Py_UNICODE_WIDE
if (ch >= 0xD800 && ch <= 0xDBFF && s < end)
{
Py_UCS4 ch2 = *(Py_UNICODE*)s;
if (ch2 >= 0xDC00 && ch2 <= 0xDFFF)
{
ch = (((ch & 0x3FF)<<10) | (ch2 & 0x3FF)) + 0x10000;
s += Py_UNICODE_SIZE;
}
}
else {
if (unicode_putchar(&v, &outpos, ch) < 0)
goto onError;
s += Py_UNICODE_SIZE;
}
#endif
if (unicode_putchar(&v, &outpos, ch) < 0)
goto onError;
}
if (PyUnicode_Resize(&v, outpos) < 0)