Skip to content

Commit 6726626

Browse files
vstinnervstinner
authored
gh-108314: Add PyDict_ContainsString() function (#108323)
1 parent c163d7f commit 6726626

File tree

10 files changed

+111
-35
lines changed

10 files changed

+111
-35
lines changed

‎Doc/c-api/dict.rst

Lines changed: 16 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -55,6 +55,15 @@ Dictionary Objects
5555
This is equivalent to the Python expression ``key in p``.
5656
5757
58+
.. c:function:: int PyDict_ContainsString(PyObject *p, const char *key)
59+
60+
This is the same as :c:func:`PyDict_Contains`, but *key* is specified as a
61+
:c:expr:`const char*` UTF-8 encoded bytes string, rather than a
62+
:c:expr:`PyObject*`.
63+
64+
.. versionadded:: 3.13
65+
66+
5867
.. c:function:: PyObject* PyDict_Copy(PyObject *p)
5968
6069
Return a new dictionary that contains the same key-value pairs as *p*.
@@ -73,7 +82,7 @@ Dictionary Objects
7382
.. index:: single: PyUnicode_FromString()
7483
7584
Insert *val* into the dictionary *p* using *key* as a key. *key* should
76-
be a :c:expr:`const char*`. The key object is created using
85+
be a :c:expr:`const char*` UTF-8 encoded bytes string. The key object is created using
7786
``PyUnicode_FromString(key)``. Return ``0`` on success or ``-1`` on
7887
failure. This function *does not* steal a reference to *val*.
7988
@@ -88,7 +97,8 @@ Dictionary Objects
8897
8998
.. c:function:: int PyDict_DelItemString(PyObject *p, const char *key)
9099
91-
Remove the entry in dictionary *p* which has a key specified by the string *key*.
100+
Remove the entry in dictionary *p* which has a key specified by the UTF-8
101+
encoded bytes string *key*.
92102
If *key* is not in the dictionary, :exc:`KeyError` is raised.
93103
Return ``0`` on success or ``-1`` on failure.
94104
@@ -136,7 +146,8 @@ Dictionary Objects
136146
.. c:function:: PyObject* PyDict_GetItemString(PyObject *p, const char *key)
137147
138148
This is the same as :c:func:`PyDict_GetItem`, but *key* is specified as a
139-
:c:expr:`const char*`, rather than a :c:expr:`PyObject*`.
149+
:c:expr:`const char*` UTF-8 encoded bytes string, rather than a
150+
:c:expr:`PyObject*`.
140151
141152
.. note::
142153
@@ -150,7 +161,8 @@ Dictionary Objects
150161
.. c:function:: int PyDict_GetItemStringRef(PyObject *p, const char *key, PyObject **result)
151162
152163
Similar than :c:func:`PyDict_GetItemRef`, but *key* is specified as a
153-
:c:expr:`const char*`, rather than a :c:expr:`PyObject*`.
164+
:c:expr:`const char*` UTF-8 encoded bytes string, rather than a
165+
:c:expr:`PyObject*`.
154166
155167
.. versionadded:: 3.13
156168

‎Doc/whatsnew/3.13.rst

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -862,6 +862,11 @@ New Features
862862
not needed.
863863
(Contributed by Victor Stinner in :gh:`106004`.)
864864

865+
* Added :c:func:`PyDict_ContainsString` function: same as
866+
:c:func:`PyDict_Contains`, but *key* is specified as a :c:expr:`const char*`
867+
UTF-8 encoded bytes string, rather than a :c:expr:`PyObject*`.
868+
(Contributed by Victor Stinner in :gh:`108314`.)
869+
865870
* Add :c:func:`Py_IsFinalizing` function: check if the main Python interpreter is
866871
:term:`shutting down <interpreter shutdown>`.
867872
(Contributed by Victor Stinner in :gh:`108014`.)

‎Include/cpython/dictobject.h

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -55,6 +55,7 @@ static inline Py_ssize_t PyDict_GET_SIZE(PyObject *op) {
5555
}
5656
#define PyDict_GET_SIZE(op) PyDict_GET_SIZE(_PyObject_CAST(op))
5757

58+
PyAPI_FUNC(int) PyDict_ContainsString(PyObject *mp, const char *key);
5859
PyAPI_FUNC(int) _PyDict_ContainsId(PyObject *, _Py_Identifier *);
5960

6061
PyAPI_FUNC(PyObject *) _PyDict_NewPresized(Py_ssize_t minused);

‎Lib/test/test_capi/test_dict.py

Lines changed: 20 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -8,6 +8,7 @@
88

99

1010
NULL = None
11+
INVALID_UTF8 = b'\xff'
1112

1213
class DictSubclass(dict):
1314
def __getitem__(self, key):
@@ -137,7 +138,7 @@ def test_dict_getitemstring(self):
137138
self.assertEqual(getitemstring(dct2, b'a'), 1)
138139
self.assertIs(getitemstring(dct2, b'b'), KeyError)
139140

140-
self.assertIs(getitemstring({}, b'\xff'), KeyError)
141+
self.assertIs(getitemstring({}, INVALID_UTF8), KeyError)
141142
self.assertIs(getitemstring(42, b'a'), KeyError)
142143
self.assertIs(getitemstring([], b'a'), KeyError)
143144
# CRASHES getitemstring({}, NULL)
@@ -173,7 +174,7 @@ def test_dict_getitemstringref(self):
173174
self.assertIs(getitemstring(dct2, b'b'), KeyError)
174175

175176
self.assertRaises(SystemError, getitemstring, 42, b'a')
176-
self.assertRaises(UnicodeDecodeError, getitemstring, {}, b'\xff')
177+
self.assertRaises(UnicodeDecodeError, getitemstring, {}, INVALID_UTF8)
177178
self.assertRaises(SystemError, getitemstring, [], b'a')
178179
# CRASHES getitemstring({}, NULL)
179180
# CRASHES getitemstring(NULL, b'a')
@@ -213,6 +214,21 @@ def test_dict_contains(self):
213214
# CRASHES contains(42, 'a')
214215
# CRASHES contains(NULL, 'a')
215216

217+
def test_dict_contains_string(self):
218+
contains_string = _testcapi.dict_containsstring
219+
dct = {'a': 1, '\U0001f40d': 2}
220+
self.assertTrue(contains_string(dct, b'a'))
221+
self.assertFalse(contains_string(dct, b'b'))
222+
self.assertTrue(contains_string(dct, '\U0001f40d'.encode()))
223+
self.assertRaises(UnicodeDecodeError, contains_string, dct, INVALID_UTF8)
224+
225+
dct2 = DictSubclass(dct)
226+
self.assertTrue(contains_string(dct2, b'a'))
227+
self.assertFalse(contains_string(dct2, b'b'))
228+
229+
# CRASHES contains({}, NULL)
230+
# CRASHES contains(NULL, b'a')
231+
216232
def test_dict_setitem(self):
217233
setitem = _testcapi.dict_setitem
218234
dct = {}
@@ -245,7 +261,7 @@ def test_dict_setitemstring(self):
245261
setitemstring(dct2, b'a', 5)
246262
self.assertEqual(dct2, {'a': 5})
247263

248-
self.assertRaises(UnicodeDecodeError, setitemstring, {}, b'\xff', 5)
264+
self.assertRaises(UnicodeDecodeError, setitemstring, {}, INVALID_UTF8, 5)
249265
self.assertRaises(SystemError, setitemstring, UserDict(), b'a', 5)
250266
self.assertRaises(SystemError, setitemstring, 42, b'a', 5)
251267
# CRASHES setitemstring({}, NULL, 5)
@@ -287,7 +303,7 @@ def test_dict_delitemstring(self):
287303
self.assertEqual(dct2, {'c': 2})
288304
self.assertRaises(KeyError, delitemstring, dct2, b'b')
289305

290-
self.assertRaises(UnicodeDecodeError, delitemstring, {}, b'\xff')
306+
self.assertRaises(UnicodeDecodeError, delitemstring, {}, INVALID_UTF8)
291307
self.assertRaises(SystemError, delitemstring, UserDict({'a': 1}), b'a')
292308
self.assertRaises(SystemError, delitemstring, 42, b'a')
293309
# CRASHES delitemstring({}, NULL)

‎Misc/NEWS.d/next/C API/2023-08-22-18-45-20.gh-issue-108314.nOlmwq.rst

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,4 @@
1+
Add :c:func:`PyDict_ContainsString` function: same as
2+
:c:func:`PyDict_Contains`, but *key* is specified as a :c:expr:`const char*`
3+
UTF-8 encoded bytes string, rather than a :c:expr:`PyObject*`.
4+
Patch by Victor Stinner.

‎Modules/_ctypes/_ctypes.c

Lines changed: 20 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -5139,8 +5139,6 @@ static PyObject *
51395139
Pointer_get_contents(CDataObject *self, void *closure)
51405140
{
51415141
StgDictObject *stgdict;
5142-
PyObject *keep, *ptr_probe;
5143-
CDataObject *ptr2ptr;
51445142

51455143
if (*(void **)self->b_ptr == NULL) {
51465144
PyErr_SetString(PyExc_ValueError,
@@ -5151,29 +5149,37 @@ Pointer_get_contents(CDataObject *self, void *closure)
51515149
stgdict = PyObject_stgdict((PyObject *)self);
51525150
assert(stgdict); /* Cannot be NULL for pointer instances */
51535151

5154-
keep = GetKeepedObjects(self);
5152+
PyObject *keep = GetKeepedObjects(self);
51555153
if (keep != NULL) {
51565154
// check if it's a pointer to a pointer:
51575155
// pointers will have '0' key in the _objects
5158-
ptr_probe = PyDict_GetItemString(keep, "0");
5159-
5160-
if (ptr_probe != NULL) {
5161-
ptr2ptr = (CDataObject*) PyDict_GetItemString(keep, "1");
5162-
if (ptr2ptr == NULL) {
5156+
int ptr_probe = PyDict_ContainsString(keep, "0");
5157+
if (ptr_probe < 0) {
5158+
return NULL;
5159+
}
5160+
if (ptr_probe) {
5161+
PyObject *item;
5162+
if (PyDict_GetItemStringRef(keep, "1", &item) < 0) {
5163+
return NULL;
5164+
}
5165+
if (item == NULL) {
51635166
PyErr_SetString(PyExc_ValueError,
5164-
"Unexpected NULL pointer in _objects");
5167+
"Unexpected NULL pointer in _objects");
51655168
return NULL;
51665169
}
5167-
// don't construct a new object,
5168-
// return existing one instead to preserve refcount
5170+
#ifndef NDEBUG
5171+
CDataObject *ptr2ptr = (CDataObject *)item;
5172+
// Don't construct a new object,
5173+
// return existing one instead to preserve refcount.
5174+
// Double-check that we are returning the same thing.
51695175
assert(
51705176
*(void**) self->b_ptr == ptr2ptr->b_ptr ||
51715177
*(void**) self->b_value.c == ptr2ptr->b_ptr ||
51725178
*(void**) self->b_ptr == ptr2ptr->b_value.c ||
51735179
*(void**) self->b_value.c == ptr2ptr->b_value.c
5174-
); // double-check that we are returning the same thing
5175-
Py_INCREF(ptr2ptr);
5176-
return (PyObject *) ptr2ptr;
5180+
);
5181+
#endif
5182+
return item;
51775183
}
51785184
}
51795185

‎Modules/_testcapi/dict.c

Lines changed: 14 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -74,6 +74,19 @@ dict_contains(PyObject *self, PyObject *args)
7474
RETURN_INT(PyDict_Contains(obj, key));
7575
}
7676

77+
static PyObject *
78+
dict_containsstring(PyObject *self, PyObject *args)
79+
{
80+
PyObject *obj;
81+
const char *key;
82+
Py_ssize_t size;
83+
if (!PyArg_ParseTuple(args, "Oz#", &obj, &key, &size)) {
84+
return NULL;
85+
}
86+
NULLABLE(obj);
87+
RETURN_INT(PyDict_ContainsString(obj, key));
88+
}
89+
7790
static PyObject *
7891
dict_size(PyObject *self, PyObject *obj)
7992
{
@@ -349,6 +362,7 @@ static PyMethodDef test_methods[] = {
349362
{"dict_getitemref", dict_getitemref, METH_VARARGS},
350363
{"dict_getitemstringref", dict_getitemstringref, METH_VARARGS},
351364
{"dict_contains", dict_contains, METH_VARARGS},
365+
{"dict_containsstring", dict_containsstring, METH_VARARGS},
352366
{"dict_setitem", dict_setitem, METH_VARARGS},
353367
{"dict_setitemstring", dict_setitemstring, METH_VARARGS},
354368
{"dict_delitem", dict_delitem, METH_VARARGS},

‎Objects/dictobject.c

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3741,6 +3741,18 @@ PyDict_Contains(PyObject *op, PyObject *key)
37413741
return (ix != DKIX_EMPTY && value != NULL);
37423742
}
37433743

3744+
int
3745+
PyDict_ContainsString(PyObject *op, const char *key)
3746+
{
3747+
PyObject *key_obj = PyUnicode_FromString(key);
3748+
if (key_obj == NULL) {
3749+
return -1;
3750+
}
3751+
int res = PyDict_Contains(op, key_obj);
3752+
Py_DECREF(key_obj);
3753+
return res;
3754+
}
3755+
37443756
/* Internal version of PyDict_Contains used when the hash value is already known */
37453757
int
37463758
_PyDict_Contains_KnownHash(PyObject *op, PyObject *key, Py_hash_t hash)

‎Python/pylifecycle.c

Lines changed: 5 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -2220,10 +2220,11 @@ add_main_module(PyInterpreterState *interp)
22202220
}
22212221
Py_DECREF(ann_dict);
22222222

2223-
if (_PyDict_GetItemStringWithError(d, "__builtins__") == NULL) {
2224-
if (PyErr_Occurred()) {
2225-
return _PyStatus_ERR("Failed to test __main__.__builtins__");
2226-
}
2223+
int has_builtins = PyDict_ContainsString(d, "__builtins__");
2224+
if (has_builtins < 0) {
2225+
return _PyStatus_ERR("Failed to test __main__.__builtins__");
2226+
}
2227+
if (!has_builtins) {
22272228
PyObject *bimod = PyImport_ImportModule("builtins");
22282229
if (bimod == NULL) {
22292230
return _PyStatus_ERR("Failed to retrieve builtins module");

‎Python/pythonrun.c

Lines changed: 14 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -413,10 +413,11 @@ _PyRun_SimpleFileObject(FILE *fp, PyObject *filename, int closeit,
413413
PyObject *dict = PyModule_GetDict(main_module); // borrowed ref
414414

415415
int set_file_name = 0;
416-
if (_PyDict_GetItemStringWithError(dict, "__file__") == NULL) {
417-
if (PyErr_Occurred()) {
418-
goto done;
419-
}
416+
int has_file = PyDict_ContainsString(dict, "__file__");
417+
if (has_file < 0) {
418+
goto done;
419+
}
420+
if (!has_file) {
420421
if (PyDict_SetItemString(dict, "__file__", filename) < 0) {
421422
goto done;
422423
}
@@ -1713,13 +1714,17 @@ run_eval_code_obj(PyThreadState *tstate, PyCodeObject *co, PyObject *globals, Py
17131714
_PyRuntime.signals.unhandled_keyboard_interrupt = 0;
17141715

17151716
/* Set globals['__builtins__'] if it doesn't exist */
1716-
if (globals != NULL && _PyDict_GetItemStringWithError(globals, "__builtins__") == NULL) {
1717-
if (PyErr_Occurred() ||
1718-
PyDict_SetItemString(globals, "__builtins__",
1719-
tstate->interp->builtins) < 0)
1720-
{
1717+
if (globals != NULL) {
1718+
int has_builtins = PyDict_ContainsString(globals, "__builtins__");
1719+
if (has_builtins < 0) {
17211720
return NULL;
17221721
}
1722+
if (!has_builtins) {
1723+
if (PyDict_SetItemString(globals, "__builtins__",
1724+
tstate->interp->builtins) < 0) {
1725+
return NULL;
1726+
}
1727+
}
17231728
}
17241729

17251730
v = PyEval_EvalCode((PyObject*)co, globals, locals);

0 commit comments

Comments
 (0)
pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy