Implement PEP 393.

This commit is contained in:
Martin v. Löwis 2011-09-28 07:41:54 +02:00
parent 48d49497c5
commit d63a3b8beb
102 changed files with 8153 additions and 5431 deletions

View file

@ -9,13 +9,26 @@ unicode_eq(PyObject *aa, PyObject *bb)
register PyUnicodeObject *a = (PyUnicodeObject *)aa;
register PyUnicodeObject *b = (PyUnicodeObject *)bb;
if (a->length != b->length)
if (PyUnicode_READY(a) == -1 || PyUnicode_READY(b) == -1) {
assert(0 && "unicode_eq ready fail");
return 0;
if (a->length == 0)
return 1;
if (a->str[0] != b->str[0])
}
if (PyUnicode_GET_LENGTH(a) != PyUnicode_GET_LENGTH(b))
return 0;
if (a->length == 1)
if (PyUnicode_GET_LENGTH(a) == 0)
return 1;
return memcmp(a->str, b->str, a->length * sizeof(Py_UNICODE)) == 0;
if (PyUnicode_KIND(a) != PyUnicode_KIND(b))
return 0;
/* Just comparing the first byte is enough to see if a and b differ.
* If they are 2 byte or 4 byte character most differences will happen in
* the lower bytes anyways.
*/
if (PyUnicode_1BYTE_DATA(a)[0] != PyUnicode_1BYTE_DATA(b)[0])
return 0;
if (PyUnicode_KIND(a) == PyUnicode_1BYTE_KIND &&
PyUnicode_GET_LENGTH(a) == 1)
return 1;
return memcmp(PyUnicode_1BYTE_DATA(a), PyUnicode_1BYTE_DATA(b),
PyUnicode_GET_LENGTH(a) * PyUnicode_CHARACTER_SIZE(a)) == 0;
}