Skip to content

Commit f5894dd

Browse files
serhiy-storchakaserhiy-storchaka
committed
Issue #28701: Replace _PyUnicode_CompareWithId with _PyUnicode_EqualToASCIIId.
The latter function is more readable, faster and doesn't raise exceptions. Based on patch by Xiang Zhang.
1 parent f4934ea commit f5894dd

5 files changed

Lines changed: 65 additions & 8 deletions

File tree

‎Include/unicodeobject.h‎

Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2000,12 +2000,31 @@ PyAPI_FUNC(int) PyUnicode_Compare(
20002000
);
20012001

20022002
#ifndef Py_LIMITED_API
2003+
/* Compare a string with an identifier and return -1, 0, 1 for less than,
2004+
equal, and greater than, respectively.
2005+
Raise an exception and return -1 on error. */
2006+
20032007
PyAPI_FUNC(int) _PyUnicode_CompareWithId(
20042008
PyObject *left, /* Left string */
20052009
_Py_Identifier *right /* Right identifier */
20062010
);
2011+
2012+
/* Test whether a unicode is equal to ASCII identifier. Return 1 if true,
2013+
0 otherwise. Return 0 if any argument contains non-ASCII characters.
2014+
Any error occurs inside will be cleared before return. */
2015+
2016+
PyAPI_FUNC(int) _PyUnicode_EqualToASCIIId(
2017+
PyObject *left, /* Left string */
2018+
_Py_Identifier *right /* Right identifier */
2019+
);
20072020
#endif
20082021

2022+
/* Compare a Unicode object with C string and return -1, 0, 1 for less than,
2023+
equal, and greater than, respectively. It is best to pass only
2024+
ASCII-encoded strings, but the function interprets the input string as
2025+
ISO-8859-1 if it contains non-ASCII characters.
2026+
Raise an exception and return -1 on error. */
2027+
20092028
PyAPI_FUNC(int) PyUnicode_CompareWithASCIIString(
20102029
PyObject *left,
20112030
const char *right /* ASCII-encoded string */

‎Objects/typeobject.c‎

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -858,7 +858,7 @@ type_repr(PyTypeObject *type)
858858
return NULL;
859859
}
860860

861-
if (mod != NULL && _PyUnicode_CompareWithId(mod, &PyId_builtins))
861+
if (mod != NULL && !_PyUnicode_EqualToASCIIId(mod, &PyId_builtins))
862862
rtn = PyUnicode_FromFormat("<class '%U.%U'>", mod, name);
863863
else
864864
rtn = PyUnicode_FromFormat("<class '%s'>", type->tp_name);
@@ -2386,7 +2386,7 @@ type_new(PyTypeObject *metatype, PyObject *args, PyObject *kwds)
23862386
if (!valid_identifier(tmp))
23872387
goto error;
23882388
assert(PyUnicode_Check(tmp));
2389-
if (_PyUnicode_CompareWithId(tmp, &PyId___dict__) == 0) {
2389+
if (_PyUnicode_EqualToASCIIId(tmp, &PyId___dict__)) {
23902390
if (!may_add_dict || add_dict) {
23912391
PyErr_SetString(PyExc_TypeError,
23922392
"__dict__ slot disallowed: "
@@ -2417,7 +2417,7 @@ type_new(PyTypeObject *metatype, PyObject *args, PyObject *kwds)
24172417
for (i = j = 0; i < nslots; i++) {
24182418
tmp = PyTuple_GET_ITEM(slots, i);
24192419
if ((add_dict &&
2420-
_PyUnicode_CompareWithId(tmp, &PyId___dict__) == 0) ||
2420+
_PyUnicode_EqualToASCIIId(tmp, &PyId___dict__)) ||
24212421
(add_weak &&
24222422
_PyUnicode_EqualToASCIIString(tmp, "__weakref__")))
24232423
continue;
@@ -3490,7 +3490,7 @@ object_repr(PyObject *self)
34903490
Py_XDECREF(mod);
34913491
return NULL;
34923492
}
3493-
if (mod != NULL && _PyUnicode_CompareWithId(mod, &PyId_builtins))
3493+
if (mod != NULL && !_PyUnicode_EqualToASCIIId(mod, &PyId_builtins))
34943494
rtn = PyUnicode_FromFormat("<%U.%U object at %p>", mod, name, self);
34953495
else
34963496
rtn = PyUnicode_FromFormat("<%s object at %p>",
@@ -7107,7 +7107,7 @@ super_getattro(PyObject *self, PyObject *name)
71077107
(i.e. super, or a subclass), not the class of su->obj. */
71087108
if (PyUnicode_Check(name) &&
71097109
PyUnicode_GET_LENGTH(name) == 9 &&
7110-
_PyUnicode_CompareWithId(name, &PyId___class__) == 0)
7110+
_PyUnicode_EqualToASCIIId(name, &PyId___class__))
71117111
goto skip;
71127112

71137113
mro = starttype->tp_mro;
@@ -7319,7 +7319,7 @@ super_init(PyObject *self, PyObject *args, PyObject *kwds)
73197319
for (i = 0; i < n; i++) {
73207320
PyObject *name = PyTuple_GET_ITEM(co->co_freevars, i);
73217321
assert(PyUnicode_Check(name));
7322-
if (!_PyUnicode_CompareWithId(name, &PyId___class__)) {
7322+
if (_PyUnicode_EqualToASCIIId(name, &PyId___class__)) {
73237323
Py_ssize_t index = co->co_nlocals +
73247324
PyTuple_GET_SIZE(co->co_cellvars) + i;
73257325
PyObject *cell = f->f_localsplus[index];

‎Objects/unicodeobject.c‎

Lines changed: 38 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -10869,6 +10869,44 @@ _PyUnicode_EqualToASCIIString(PyObject *unicode, const char *str)
1086910869
memcmp(PyUnicode_1BYTE_DATA(unicode), str, len) == 0;
1087010870
}
1087110871

10872+
int
10873+
_PyUnicode_EqualToASCIIId(PyObject *left, _Py_Identifier *right)
10874+
{
10875+
PyObject *right_uni;
10876+
Py_hash_t hash;
10877+
10878+
assert(_PyUnicode_CHECK(left));
10879+
assert(right->string);
10880+
10881+
if (PyUnicode_READY(left) == -1) {
10882+
/* memory error or bad data */
10883+
PyErr_Clear();
10884+
return non_ready_unicode_equal_to_ascii_string(left, right->string);
10885+
}
10886+
10887+
if (!PyUnicode_IS_ASCII(left))
10888+
return 0;
10889+
10890+
right_uni = _PyUnicode_FromId(right); /* borrowed */
10891+
if (right_uni == NULL) {
10892+
/* memory error or bad data */
10893+
PyErr_Clear();
10894+
return _PyUnicode_EqualToASCIIString(left, right->string);
10895+
}
10896+
10897+
if (left == right_uni)
10898+
return 1;
10899+
10900+
if (PyUnicode_CHECK_INTERNED(left))
10901+
return 0;
10902+
10903+
assert(_PyUnicode_HASH(right_uni) != 1);
10904+
hash = _PyUnicode_HASH(left);
10905+
if (hash != -1 && hash != _PyUnicode_HASH(right_uni))
10906+
return 0;
10907+
10908+
return unicode_compare_eq(left, right_uni);
10909+
}
1087210910

1087310911
#define TEST_COND(cond) \
1087410912
((cond) ? Py_True : Py_False)

‎Python/errors.c‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -934,7 +934,7 @@ PyErr_WriteUnraisable(PyObject *obj)
934934
goto done;
935935
}
936936
else {
937-
if (_PyUnicode_CompareWithId(moduleName, &PyId_builtins) != 0) {
937+
if (!_PyUnicode_EqualToASCIIId(moduleName, &PyId_builtins)) {
938938
if (PyFile_WriteObject(moduleName, f, Py_PRINT_RAW) < 0)
939939
goto done;
940940
if (PyFile_WriteString(".", f) < 0)

‎Python/pythonrun.c‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -747,7 +747,7 @@ print_exception(PyObject *f, PyObject *value)
747747
err = PyFile_WriteString("<unknown>", f);
748748
}
749749
else {
750-
if (_PyUnicode_CompareWithId(moduleName, &PyId_builtins) != 0)
750+
if (!_PyUnicode_EqualToASCIIId(moduleName, &PyId_builtins))
751751
{
752752
err = PyFile_WriteObject(moduleName, f, Py_PRINT_RAW);
753753
err += PyFile_WriteString(".", f);

0 commit comments

Comments
 (0)