gh-111262: Add PyDict_Pop() function [without default value nor KeyError]

Doc/c-api/dict.rst

@@ -173,6 +173,33 @@ Dictionary Objects
 
  .. versionadded:: 3.4
 
+
+.. c:function:: int PyDict_Pop(PyObject *p, PyObject *key, PyObject **result)
+
+ Remove *key* from dictionary *p* and optionally return the removed value.
+ Do not raise :exc:`KeyError` if the key missing.
+
+ - If the key is present, set *\*result* to a new reference to the removed
+ value if *result* is not ``NULL``, and return ``1``.
+ - If the key is missing, set *\*result* to ``NULL`` if *result* is not
+ ``NULL``, and return ``0``.
+ - On error, raise an exception and return ``-1``.
+
+ This is similar to :meth:`dict.pop`, but without the default value and
+ not raising :exc:`KeyError` if the key missing.
+
+ .. versionadded:: 3.13
+
+
+.. c:function:: int PyDict_PopString(PyObject *p, const char *key, PyObject **result)
+
+ Similar to :c:func:`PyDict_Pop`, but *key* is specified as a
+ :c:expr:`const char*` UTF-8 encoded bytes string, rather than a
+ :c:expr:`PyObject*`.
+
+ .. versionadded:: 3.13
+
+
 .. c:function:: PyObject* PyDict_Items(PyObject *p)
 
  Return a :c:type:`PyListObject` containing all the items from the dictionary.

Doc/whatsnew/3.13.rst

@@ -1175,6 +1175,12 @@ New Features
  Python ``list.extend()`` and ``list.clear()`` methods.
  (Contributed by Victor Stinner in :gh:`111138`.)
 
+* Add :c:func:`PyDict_Pop` and :c:func:`PyDict_PopString` functions: remove a
+ key from a dictionary and optionally return the removed value. This is
+ similar to :meth:`dict.pop`, but without the default value and not raising
+ :exc:`KeyError` if the key missing.
+ (Contributed by Stefan Behnel and Victor Stinner in :gh:`111262`.)
+
 
 Porting to Python 3.13
 ----------------------

Include/cpython/dictobject.h

@@ -46,6 +46,8 @@ static inline Py_ssize_t PyDict_GET_SIZE(PyObject *op) {
 
 PyAPI_FUNC(int) PyDict_ContainsString(PyObject *mp, const char *key);
 
+PyAPI_FUNC(int) PyDict_Pop(PyObject *dict, PyObject *key, PyObject **result);
+PyAPI_FUNC(int) PyDict_PopString(PyObject *dict, const char *key, PyObject **result);
 PyAPI_FUNC(PyObject *) _PyDict_Pop(PyObject *dict, PyObject *key, PyObject *default_value);
 
 /* Dictionary watchers */

Include/internal/pycore_dict.h

@@ -116,7 +116,11 @@ extern PyObject *_PyDict_LoadGlobal(PyDictObject *, PyDictObject *, PyObject *);
 extern int _PyDict_SetItem_Take2(PyDictObject *op, PyObject *key, PyObject *value);
 extern int _PyObjectDict_SetItem(PyTypeObject *tp, PyObject **dictptr, PyObject *name, PyObject *value);
 
-extern PyObject *_PyDict_Pop_KnownHash(PyObject *, PyObject *, Py_hash_t, PyObject *);
+extern int _PyDict_Pop_KnownHash(
+ PyDictObject *dict,
+ PyObject *key,
+ Py_hash_t hash,
+ PyObject **result);
 
 #define DKIX_EMPTY (-1)
 #define DKIX_DUMMY (-2) /* Used internally */

Lib/test/test_capi/test_dict.py

@@ -432,6 +432,93 @@ def test_dict_mergefromseq2(self):
  # CRASHES mergefromseq2({}, NULL, 0)
  # CRASHES mergefromseq2(NULL, {}, 0)
 
+ def test_dict_pop(self):
+ # Test PyDict_Pop()
+ dict_pop = _testcapi.dict_pop
+ dict_pop_null = _testcapi.dict_pop_null
+
+ # key present, get removed value
+ mydict = {"key": "value", "key2": "value2"}
+ self.assertEqual(dict_pop(mydict, "key"), (1, "value"))
+ self.assertEqual(mydict, {"key2": "value2"})
+ self.assertEqual(dict_pop(mydict, "key2"), (1, "value2"))
+ self.assertEqual(mydict, {})
+
+ # key present, ignore removed value
+ mydict = {"key": "value", "key2": "value2"}
+ self.assertEqual(dict_pop_null(mydict, "key"), 1)
+ self.assertEqual(mydict, {"key2": "value2"})
+ self.assertEqual(dict_pop_null(mydict, "key2"), 1)
+ self.assertEqual(mydict, {})
+
+ # key missing, expect removed value; empty dict has a fast path
+ self.assertEqual(dict_pop({}, "key"), (0, NULL))
+ self.assertEqual(dict_pop({"a": 1}, "key"), (0, NULL))
+
+ # key missing, ignored removed value; empty dict has a fast path
+ self.assertEqual(dict_pop_null({}, "key"), 0)
+ self.assertEqual(dict_pop_null({"a": 1}, "key"), 0)
+
+ # dict error
+ not_dict = UserDict({1: 2})
+ self.assertRaises(SystemError, dict_pop, not_dict, "key")
+ self.assertRaises(SystemError, dict_pop_null, not_dict, "key")
+
+ # key error; don't hash key if dict is empty
+ not_hashable_key = ["list"]
+ self.assertEqual(dict_pop({}, not_hashable_key), (0, NULL))
+ with self.assertRaises(TypeError):
+ dict_pop({'key': 1}, not_hashable_key)
+ dict_pop({}, NULL) # key is not checked if dict is empty
+
+ # CRASHES dict_pop(NULL, "key")
+ # CRASHES dict_pop({"a": 1}, NULL)
+
+ def test_dict_popstring(self):
+ # Test PyDict_PopString()
+ dict_popstring = _testcapi.dict_popstring
+ dict_popstring_null = _testcapi.dict_popstring_null
+
+ # key present, get removed value
+ mydict = {"key": "value", "key2": "value2"}
+ self.assertEqual(dict_popstring(mydict, "key"), (1, "value"))
+ self.assertEqual(mydict, {"key2": "value2"})
+ self.assertEqual(dict_popstring(mydict, "key2"), (1, "value2"))
+ self.assertEqual(mydict, {})
+
+ # key present, ignore removed value
+ mydict = {"key": "value", "key2": "value2"}
+ self.assertEqual(dict_popstring_null(mydict, "key"), 1)
+ self.assertEqual(mydict, {"key2": "value2"})
+ self.assertEqual(dict_popstring_null(mydict, "key2"), 1)
+ self.assertEqual(mydict, {})
+
+ # key missing; empty dict has a fast path
+ self.assertEqual(dict_popstring({}, "key"), (0, NULL))
+ self.assertEqual(dict_popstring_null({}, "key"), 0)
+ self.assertEqual(dict_popstring({"a": 1}, "key"), (0, NULL))
+ self.assertEqual(dict_popstring_null({"a": 1}, "key"), 0)
+
+ # non-ASCII key
+ non_ascii = '\U0001f40d'
+ dct = {'\U0001f40d': 123}
+ self.assertEqual(dict_popstring(dct, '\U0001f40d'.encode()), (1, 123))
+ dct = {'\U0001f40d': 123}
+ self.assertEqual(dict_popstring_null(dct, '\U0001f40d'.encode()), 1)
+
+ # dict error
+ not_dict = UserDict({1: 2})
+ self.assertRaises(SystemError, dict_popstring, not_dict, "key")
+ self.assertRaises(SystemError, dict_popstring_null, not_dict, "key")
+
+ # key error
+ self.assertRaises(UnicodeDecodeError, dict_popstring, {1: 2}, INVALID_UTF8)
+ self.assertRaises(UnicodeDecodeError, dict_popstring_null, {1: 2}, INVALID_UTF8)
+
+ # CRASHES dict_popstring(NULL, "key")
+ # CRASHES dict_popstring({}, NULL)
+ # CRASHES dict_popstring({"a": 1}, NULL)
+
 
 if __name__ == "__main__":
  unittest.main()

Misc/NEWS.d/next/C API/2023-11-10-10-21-38.gh-issue-111262.2utB5m.rst

@@ -0,0 +1,4 @@
+Add :c:func:`PyDict_Pop` and :c:func:`PyDict_PopString` functions: remove a key
+from a dictionary and optionally return the removed value. This is similar to
+:meth:`dict.pop`, but without the default value and not raising :exc:`KeyError`
+if the key missing. Patch by Stefan Behnel and Victor Stinner.

Modules/_functoolsmodule.c

@@ -1087,19 +1087,9 @@ bounded_lru_cache_wrapper(lru_cache_object *self, PyObject *args, PyObject *kwds
  The cache dict holds one reference to the link.
  We created one other reference when the link was created.
  The linked list only has borrowed references. */
- popresult = _PyDict_Pop_KnownHash(self->cache, link->key,
- link->hash, Py_None);
- if (popresult == Py_None) {
- /* Getting here means that the user function call or another
- thread has already removed the old key from the dictionary.
- This link is now an orphan. Since we don't want to leave the
- cache in an inconsistent state, we don't restore the link. */
- Py_DECREF(popresult);
- Py_DECREF(link);
- Py_DECREF(key);
- return result;
- }
- if (popresult == NULL) {
+ int res = _PyDict_Pop_KnownHash((PyDictObject*)self->cache, link->key,
+ link->hash, &popresult);
+ if (res < 0) {
  /* An error arose while trying to remove the oldest key (the one
  being evicted) from the cache. We restore the link to its
  original position as the oldest link. Then we allow the
@@ -1110,10 +1100,22 @@ bounded_lru_cache_wrapper(lru_cache_object *self, PyObject *args, PyObject *kwds
  Py_DECREF(result);
  return NULL;
  }
+ if (res == 0) {
+ /* Getting here means that the user function call or another
+ thread has already removed the old key from the dictionary.
+ This link is now an orphan. Since we don't want to leave the
+ cache in an inconsistent state, we don't restore the link. */
+ assert(popresult == NULL);
+ Py_DECREF(link);
+ Py_DECREF(key);
+ return result;
+ }
+
  /* Keep a reference to the old key and old result to prevent their
  ref counts from going to zero during the update. That will
  prevent potentially arbitrary object clean-up code (i.e. __del__)
  from running while we're still adjusting the links. */
+ assert(popresult != NULL);
  oldkey = link->key;
  oldresult = link->result;
 

Modules/_testcapi/dict.c

@@ -331,6 +331,88 @@ dict_mergefromseq2(PyObject *self, PyObject *args)
 }
 
 
+static PyObject *
+dict_pop(PyObject *self, PyObject *args)
+{
+ // Test PyDict_Pop(dict, key, &value)
+ PyObject *dict, *key;
+ if (!PyArg_ParseTuple(args, "OO", &dict, &key)) {
+ return NULL;
+ }
+ NULLABLE(dict);
+ NULLABLE(key);
+ PyObject *result = UNINITIALIZED_PTR;
+ int res = PyDict_Pop(dict, key, &result);
+ if (res < 0) {
+ assert(result == NULL);
+ return NULL;
+ }
+ if (res == 0) {
+ assert(result == NULL);
+ result = Py_NewRef(Py_None);
+ }
+ else {
+ assert(result != NULL);
+ }
+ return Py_BuildValue("iN", res, result);
+}
+
+
+static PyObject *
+dict_pop_null(PyObject *self, PyObject *args)
+{
+ // Test PyDict_Pop(dict, key, NULL)
+ PyObject *dict, *key;
+ if (!PyArg_ParseTuple(args, "OO", &dict, &key)) {
+ return NULL;
+ }
+ NULLABLE(dict);
+ NULLABLE(key);
+ RETURN_INT(PyDict_Pop(dict, key, NULL));
+}
+
+
+static PyObject *
+dict_popstring(PyObject *self, PyObject *args)
+{
+ PyObject *dict;
+ const char *key;
+ Py_ssize_t key_size;
+ if (!PyArg_ParseTuple(args, "Oz#", &dict, &key, &key_size)) {
+ return NULL;
+ }
+ NULLABLE(dict);
+ PyObject *result = UNINITIALIZED_PTR;
+ int res = PyDict_PopString(dict, key, &result);
+ if (res < 0) {
+ assert(result == NULL);
+ return NULL;
+ }
+ if (res == 0) {
+ assert(result == NULL);
+ result = Py_NewRef(Py_None);
+ }
+ else {
+ assert(result != NULL);
+ }
+ return Py_BuildValue("iN", res, result);
+}
+
+
+static PyObject *
+dict_popstring_null(PyObject *self, PyObject *args)
+{
+ PyObject *dict;
+ const char *key;
+ Py_ssize_t key_size;
+ if (!PyArg_ParseTuple(args, "Oz#", &dict, &key, &key_size)) {
+ return NULL;
+ }
+ NULLABLE(dict);
+ RETURN_INT(PyDict_PopString(dict, key, NULL));
+}
+
+
 static PyMethodDef test_methods[] = {
  {"dict_check", dict_check, METH_O},
  {"dict_checkexact", dict_checkexact, METH_O},
@@ -358,7 +440,10 @@ static PyMethodDef test_methods[] = {
  {"dict_merge", dict_merge, METH_VARARGS},
  {"dict_update", dict_update, METH_VARARGS},
  {"dict_mergefromseq2", dict_mergefromseq2, METH_VARARGS},
-
+ {"dict_pop", dict_pop, METH_VARARGS},
+ {"dict_pop_null", dict_pop_null, METH_VARARGS},
+ {"dict_popstring", dict_popstring, METH_VARARGS},
+ {"dict_popstring_null", dict_popstring_null, METH_VARARGS},
  {NULL},
 };
 

Modules/_threadmodule.c

@@ -967,11 +967,8 @@ local_clear(localobject *self)
  HEAD_UNLOCK(runtime);
  while (tstate) {
  if (tstate->dict) {
- PyObject *v = _PyDict_Pop(tstate->dict, self->key, Py_None);
- if (v != NULL) {
- Py_DECREF(v);
- }
- else {
+ if (PyDict_Pop(tstate->dict, self->key, NULL) < 0) {
+ // Silently ignore error
  PyErr_Clear();
  }
  }

Modules/socketmodule.c

@@ -392,16 +392,10 @@ remove_unusable_flags(PyObject *m)
  break;
  }
  else {
-PyObject *flag_name = PyUnicode_FromString(win_runtime_flags[i].flag_name);
-if (flag_name == NULL) {
+if (PyDict_PopString(dict, win_runtime_flags[i].flag_name,
+NULL) < 0) {
  return -1;
  }
- PyObject *v = _PyDict_Pop(dict, flag_name, Py_None);
- Py_DECREF(flag_name);
- if (v == NULL) {
- return -1;
- }
- Py_DECREF(v);
  }
  }
  return 0;