Add _PyUnicode_DATA_ANY(op) private macro
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 49afa4a..2c48c82 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -131,11 +131,11 @@
#define _PyUnicode_GET_LENGTH(op) \
(assert(PyUnicode_Check(op)), \
((PyASCIIObject *)(op))->length)
+#define _PyUnicode_DATA_ANY(op) (((PyUnicodeObject*)(op))->data.any)
/* The Unicode string has been modified: reset the hash */
#define _PyUnicode_DIRTY(op) do { _PyUnicode_HASH(op) = -1; } while (0)
-
/* This dictionary holds all interned unicode strings. Note that references
to strings in this dictionary are *not* counted in the string's ob_refcnt.
When the interned string reaches a refcnt of 0 the string deallocation
@@ -441,7 +441,7 @@
_PyUnicode_STATE(unicode).compact = 0;
_PyUnicode_STATE(unicode).ready = 0;
_PyUnicode_STATE(unicode).ascii = 0;
- unicode->data.any = NULL;
+ _PyUnicode_DATA_ANY(unicode) = NULL;
_PyUnicode_LENGTH(unicode) = 0;
_PyUnicode_UTF8(unicode) = NULL;
_PyUnicode_UTF8_LENGTH(unicode) = 0;
@@ -815,7 +815,7 @@
assert(!PyUnicode_IS_COMPACT(obj));
assert(_PyUnicode_KIND(obj) == PyUnicode_WCHAR_KIND);
assert(_PyUnicode_WSTR(unicode) != NULL);
- assert(unicode->data.any == NULL);
+ assert(_PyUnicode_DATA_ANY(unicode) == NULL);
assert(_PyUnicode_UTF8(unicode) == NULL);
/* Actually, it should neither be interned nor be anything else: */
assert(_PyUnicode_STATE(unicode).interned == SSTATE_NOT_INTERNED);
@@ -830,8 +830,8 @@
return -1;
if (maxchar < 256) {
- unicode->data.any = PyObject_MALLOC(_PyUnicode_WSTR_LENGTH(unicode) + 1);
- if (!unicode->data.any) {
+ _PyUnicode_DATA_ANY(unicode) = PyObject_MALLOC(_PyUnicode_WSTR_LENGTH(unicode) + 1);
+ if (!_PyUnicode_DATA_ANY(unicode)) {
PyErr_NoMemory();
return -1;
}
@@ -842,7 +842,7 @@
_PyUnicode_LENGTH(unicode) = _PyUnicode_WSTR_LENGTH(unicode);
_PyUnicode_STATE(unicode).kind = PyUnicode_1BYTE_KIND;
if (maxchar < 128) {
- _PyUnicode_UTF8(unicode) = unicode->data.any;
+ _PyUnicode_UTF8(unicode) = _PyUnicode_DATA_ANY(unicode);
_PyUnicode_UTF8_LENGTH(unicode) = _PyUnicode_WSTR_LENGTH(unicode);
}
else {
@@ -861,7 +861,7 @@
#if SIZEOF_WCHAR_T == 2
/* We can share representations and are done. */
- unicode->data.any = _PyUnicode_WSTR(unicode);
+ _PyUnicode_DATA_ANY(unicode) = _PyUnicode_WSTR(unicode);
PyUnicode_2BYTE_DATA(unicode)[_PyUnicode_WSTR_LENGTH(unicode)] = '\0';
_PyUnicode_LENGTH(unicode) = _PyUnicode_WSTR_LENGTH(unicode);
_PyUnicode_STATE(unicode).kind = PyUnicode_2BYTE_KIND;
@@ -869,9 +869,9 @@
_PyUnicode_UTF8_LENGTH(unicode) = 0;
#else
/* sizeof(wchar_t) == 4 */
- unicode->data.any = PyObject_MALLOC(
+ _PyUnicode_DATA_ANY(unicode) = PyObject_MALLOC(
2 * (_PyUnicode_WSTR_LENGTH(unicode) + 1));
- if (!unicode->data.any) {
+ if (!_PyUnicode_DATA_ANY(unicode)) {
PyErr_NoMemory();
return -1;
}
@@ -894,8 +894,8 @@
/* in case the native representation is 2-bytes, we need to allocate a
new normalized 4-byte version. */
length_wo_surrogates = _PyUnicode_WSTR_LENGTH(unicode) - num_surrogates;
- unicode->data.any = PyObject_MALLOC(4 * (length_wo_surrogates + 1));
- if (!unicode->data.any) {
+ _PyUnicode_DATA_ANY(unicode) = PyObject_MALLOC(4 * (length_wo_surrogates + 1));
+ if (!_PyUnicode_DATA_ANY(unicode)) {
PyErr_NoMemory();
return -1;
}
@@ -914,7 +914,7 @@
#else
assert(num_surrogates == 0);
- unicode->data.any = _PyUnicode_WSTR(unicode);
+ _PyUnicode_DATA_ANY(unicode) = _PyUnicode_WSTR(unicode);
_PyUnicode_LENGTH(unicode) = _PyUnicode_WSTR_LENGTH(unicode);
_PyUnicode_UTF8(unicode) = NULL;
_PyUnicode_UTF8_LENGTH(unicode) = 0;
@@ -961,8 +961,8 @@
Py_TYPE(unicode)->tp_free((PyObject *)unicode);
}
else {
- if (unicode->data.any)
- PyObject_DEL(unicode->data.any);
+ if (_PyUnicode_DATA_ANY(unicode))
+ PyObject_DEL(_PyUnicode_DATA_ANY(unicode));
Py_TYPE(unicode)->tp_free((PyObject *)unicode);
}
}
@@ -11657,7 +11657,7 @@
/* If it is a two-block object, account for base object, and
for character block if present. */
size = sizeof(PyUnicodeObject);
- if (v->data.any)
+ if (_PyUnicode_DATA_ANY(v))
size += (PyUnicode_GET_LENGTH(v) + 1) *
PyUnicode_CHARACTER_SIZE(v);
}
@@ -12477,7 +12477,7 @@
_PyUnicode_UTF8_LENGTH(self) = 0;
_PyUnicode_UTF8(self) = NULL;
_PyUnicode_WSTR_LENGTH(self) = 0;
- self->data.any = NULL;
+ _PyUnicode_DATA_ANY(self) = NULL;
share_utf8 = 0;
share_wstr = 0;
@@ -12509,7 +12509,7 @@
goto onError;
}
- self->data.any = data;
+ _PyUnicode_DATA_ANY(self) = data;
if (share_utf8) {
_PyUnicode_UTF8_LENGTH(self) = length;
_PyUnicode_UTF8(self) = data;