rewrite of dict iteration code at a more C-ish level in order to make the...

rewrite of dict iteration code at a more C-ish level in order to make the actual code C compile time dependent

--HG--
extra : transplant_source : R%BCF%17%5D%7E%FA%B3%C4q%D2%28%CF%0Di%8D%FC%0B%E4%7B

Cython/Compiler/ExprNodes.py

@@ -9752,25 +9752,9 @@ static CYTHON_INLINE int __Pyx_ErrOccurredWithGIL(void) {
 
 #------------------------------------------------------------------------------------
 
-raise_noneattr_error_utility_code = UtilityCode(
-proto = """
-static CYTHON_INLINE void __Pyx_RaiseNoneAttributeError(const char* attrname);
-""",
-impl = '''
-static CYTHON_INLINE void __Pyx_RaiseNoneAttributeError(const char* attrname) {
-    PyErr_Format(PyExc_AttributeError, "'NoneType' object has no attribute '%s'", attrname);
-}
-''')
-
-raise_noneindex_error_utility_code = UtilityCode(
-proto = """
-static CYTHON_INLINE void __Pyx_RaiseNoneIndexingError(void);
-""",
-impl = '''
-static CYTHON_INLINE void __Pyx_RaiseNoneIndexingError(void) {
-    PyErr_SetString(PyExc_TypeError, "'NoneType' object is unsubscriptable");
-}
-''')
+raise_noneattr_error_utility_code = UtilityCode.load_cached("RaiseNoneAttrError", "ObjectHandling.c")
+raise_noneindex_error_utility_code = UtilityCode.load_cached("RaiseNoneIndexingError", "ObjectHandling.c")
+raise_none_iter_error_utility_code = UtilityCode.load_cached("RaiseNoneIterError", "ObjectHandling.c")
 
 raise_noneindex_memview_error_utility_code = UtilityCode(
     proto = """
@@ -9782,16 +9766,6 @@ static CYTHON_INLINE void __Pyx_RaiseNoneMemviewIndexingError(void) {
 }
 ''')
 
-raise_none_iter_error_utility_code = UtilityCode(
-proto = """
-static CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void);
-""",
-impl = '''
-static CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void) {
-    PyErr_SetString(PyExc_TypeError, "'NoneType' object is not iterable");
-}
-''')
-
 raise_unbound_local_error_utility_code = UtilityCode(
 proto = """
 static CYTHON_INLINE void __Pyx_RaiseUnboundLocalError(const char *varname);
@@ -10029,92 +10003,14 @@ impl = """
 
 #------------------------------------------------------------------------------------
 
-raise_too_many_values_to_unpack = UtilityCode(
-proto = """
-static CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected);
-""",
-impl = '''
-static CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected) {
-    PyErr_Format(PyExc_ValueError,
-                 "too many values to unpack (expected %"PY_FORMAT_SIZE_T"d)", expected);
-}
-''')
-
-raise_need_more_values_to_unpack = UtilityCode(
-proto = """
-static CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index);
-""",
-impl = '''
-static CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index) {
-    PyErr_Format(PyExc_ValueError,
-                 "need more than %"PY_FORMAT_SIZE_T"d value%s to unpack",
-                 index, (index == 1) ? "" : "s");
-}
-''')
+raise_too_many_values_to_unpack = UtilityCode.load_cached("RaiseTooManyValuesToUnpack", "ObjectHandling.c")
+raise_need_more_values_to_unpack = UtilityCode.load_cached("RaiseNeedMoreValuesToUnpack", "ObjectHandling.c")
 
 #------------------------------------------------------------------------------------
 
-tuple_unpacking_error_code = UtilityCode(
-proto = """
-static void __Pyx_UnpackTupleError(PyObject *, Py_ssize_t index); /*proto*/
-""",
-impl = """
-static void __Pyx_UnpackTupleError(PyObject *t, Py_ssize_t index) {
-    if (t == Py_None) {
-      __Pyx_RaiseNoneNotIterableError();
-    } else if (PyTuple_GET_SIZE(t) < index) {
-      __Pyx_RaiseNeedMoreValuesError(PyTuple_GET_SIZE(t));
-    } else {
-      __Pyx_RaiseTooManyValuesError(index);
-    }
-}
-""",
-requires = [raise_none_iter_error_utility_code,
-            raise_need_more_values_to_unpack,
-            raise_too_many_values_to_unpack]
-)
-
-unpacking_utility_code = UtilityCode(
-proto = """
-static PyObject *__Pyx_UnpackItem(PyObject *, Py_ssize_t index); /*proto*/
-""",
-impl = """
-static PyObject *__Pyx_UnpackItem(PyObject *iter, Py_ssize_t index) {
-    PyObject *item;
-    if (!(item = PyIter_Next(iter))) {
-        if (!PyErr_Occurred()) {
-            __Pyx_RaiseNeedMoreValuesError(index);
-        }
-    }
-    return item;
-}
-""",
-requires = [raise_need_more_values_to_unpack]
-)
-
-iternext_unpacking_end_utility_code = UtilityCode(
-proto = """
-static int __Pyx_IternextUnpackEndCheck(PyObject *retval, Py_ssize_t expected); /*proto*/
-""",
-impl = """
-static int __Pyx_IternextUnpackEndCheck(PyObject *retval, Py_ssize_t expected) {
-    if (unlikely(retval)) {
-        Py_DECREF(retval);
-        __Pyx_RaiseTooManyValuesError(expected);
-        return -1;
-    } else if (PyErr_Occurred()) {
-        if (likely(PyErr_ExceptionMatches(PyExc_StopIteration))) {
-            PyErr_Clear();
-            return 0;
-        } else {
-            return -1;
-        }
-    }
-    return 0;
-}
-""",
-requires = [raise_too_many_values_to_unpack]
-)
+tuple_unpacking_error_code = UtilityCode.load_cached("UnpackTupleError", "ObjectHandling.c")
+unpacking_utility_code = UtilityCode.load_cached("UnpackItem", "ObjectHandling.c")
+iternext_unpacking_end_utility_code = UtilityCode.load_cached("UnpackItemEndCheck", "ObjectHandling.c")
 
 #------------------------------------------------------------------------------------
 

Cython/Compiler/Nodes.py

@@ -5772,45 +5772,98 @@ class DictIterationNextNode(Node):
     # Helper node for calling PyDict_Next() inside of a WhileStatNode
     # and checking the dictionary size for changes.  Created in
     # Optimize.py.
-    child_attrs = ['dict_obj', 'expected_size', 'pos_index_addr', 'key_addr', 'value_addr']
+    child_attrs = ['dict_obj', 'expected_size', 'pos_index_var',
+                   'coerced_key_var', 'coerced_value_var', 'coerced_tuple_var',
+                   'key_target', 'value_target', 'tuple_target', 'is_dict_flag']
 
-    def __init__(self, dict_obj, expected_size, pos_index_addr, key_addr, value_addr):
+    coerced_key_var = key_ref = None
+    coerced_value_var = value_ref = None
+    coerced_tuple_var = tuple_ref = None
+
+    def __init__(self, dict_obj, expected_size, pos_index_var,
+                 key_target, value_target, tuple_target, is_dict_flag):
         Node.__init__(
             self, dict_obj.pos,
             dict_obj = dict_obj,
             expected_size = expected_size,
-            pos_index_addr = pos_index_addr,
-            key_addr = key_addr,
-            value_addr = value_addr,
+            pos_index_var = pos_index_var,
+            key_target = key_target,
+            value_target = value_target,
+            tuple_target = tuple_target,
+            is_dict_flag = is_dict_flag,
+            is_temp = True,
             type = PyrexTypes.c_bint_type)
 
     def analyse_expressions(self, env):
+        import UtilNodes
         self.dict_obj.analyse_types(env)
         self.expected_size.analyse_types(env)
-        self.pos_index_addr.analyse_types(env)
-        self.key_addr.analyse_types(env)
-        self.value_addr.analyse_types(env)
+        if self.pos_index_var: self.pos_index_var.analyse_types(env)
+        if self.key_target:
+            self.key_target.analyse_target_types(env)
+            self.key_ref = UtilNodes.ResultRefNode(pos=self.key_target.pos, is_temp=True,
+                                                   type=PyrexTypes.py_object_type)
+            self.coerced_key_var = self.key_ref.coerce_to(self.key_target.type, env)
+        if self.value_target:
+            self.value_target.analyse_target_types(env)
+            self.value_ref = UtilNodes.ResultRefNode(pos=self.value_target.pos, is_temp=True,
+                                                     type=PyrexTypes.py_object_type)
+            self.coerced_value_var = self.value_ref.coerce_to(self.value_target.type, env)
+        if self.tuple_target:
+            self.tuple_target.analyse_target_types(env)
+            self.tuple_ref = UtilNodes.ResultRefNode(pos=self.tuple_target.pos, is_temp=True,
+                                                     type=PyrexTypes.py_object_type)
+            self.coerced_tuple_var = self.tuple_ref.coerce_to(self.tuple_target.type, env)
+        self.is_dict_flag.analyse_types(env)
 
     def generate_function_definitions(self, env, code):
         self.dict_obj.generate_function_definitions(env, code)
 
     def generate_execution_code(self, code):
+        code.globalstate.use_utility_code(UtilityCode.load_cached("dict_iter", "Optimize.c"))
         self.dict_obj.generate_evaluation_code(code)
-        code.putln("if (unlikely(%s != PyDict_Size(%s))) {" % (
-            self.expected_size.result(),
+
+        assignments = []
+        temp_addresses = []
+        for var, result, target in [(self.key_ref, self.coerced_key_var, self.key_target),
+                                    (self.value_ref, self.coerced_value_var, self.value_target),
+                                    (self.tuple_ref, self.coerced_tuple_var, self.tuple_target)]:
+            if target is None:
+                addr = 'NULL'
+            else:
+                temp = code.funcstate.allocate_temp(PyrexTypes.py_object_type, True)
+                var.result_code = temp
+                assignments.append((temp, result, target))
+                addr = '&%s' % temp
+            temp_addresses.append(addr)
+
+        result_temp = code.funcstate.allocate_temp(PyrexTypes.c_int_type, False)
+        code.putln("%s = __Pyx_dict_iter_next(%s, %s, &%s, %s, %s, %s, %s);" % (
+            result_temp,
             self.dict_obj.py_result(),
+            self.expected_size.result(),
+            self.pos_index_var.result(),
+            temp_addresses[0],
+            temp_addresses[1],
+            temp_addresses[2],
+            self.is_dict_flag.result()
         ))
-        code.putln('PyErr_SetString(PyExc_RuntimeError, "dictionary changed size during iteration"); %s' % (
-            code.error_goto(self.pos)))
-        code.putln("}")
-        self.pos_index_addr.generate_evaluation_code(code)
+        code.putln("if (unlikely(%s == 0)) break;" % result_temp)
+        code.putln(code.error_goto_if("%s == -1" % result_temp, self.pos))
+        code.funcstate.release_temp(result_temp)
 
-        code.putln("if (!PyDict_Next(%s, %s, %s, %s)) break;" % (
-            self.dict_obj.py_result(),
-            self.pos_index_addr.result(),
-            self.key_addr.result(),
-            self.value_addr.result()))
+        # evaluate all coercions before the assignments
+        for temp, result, target in assignments:
+            code.put_gotref(temp)
+            result.generate_evaluation_code(code)
+            if not result.type.is_pyobject:
+                code.put_decref_clear(temp, PyrexTypes.py_object_type)
+                code.funcstate.release_temp(temp)
 
+        for temp, result, target in assignments:
+            target.generate_assignment_code(result, code)
+            if result.type.is_pyobject:
+                code.funcstate.release_temp(temp)
 
 def ForStatNode(pos, **kw):
     if 'iterator' in kw:

Cython/Compiler/UtilNodes.py

@@ -120,7 +120,7 @@ class ResultRefNode(AtomicExprNode):
     subexprs = []
     lhs_of_first_assignment = False
 
-    def __init__(self, expression=None, pos=None, type=None, may_hold_none=True):
+    def __init__(self, expression=None, pos=None, type=None, may_hold_none=True, is_temp=False):
         self.expression = expression
         self.pos = None
         self.may_hold_none = may_hold_none
@@ -132,6 +132,8 @@ class ResultRefNode(AtomicExprNode):
             self.pos = pos
         if type is not None:
             self.type = type
+        if is_temp:
+            self.is_temp = True
         assert self.pos is not None
 
     def clone_node(self):

Cython/Utility/ObjectHandling.c

+
+/////////////// RaiseNoneAttrError.proto ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNoneAttributeError(const char* attrname);
+
+/////////////// RaiseNoneAttrError ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNoneAttributeError(const char* attrname) {
+    PyErr_Format(PyExc_AttributeError, "'NoneType' object has no attribute '%s'", attrname);
+}
+
+/////////////// RaiseNoneIndexingError.proto ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNoneIndexingError(void);
+
+/////////////// RaiseNoneIndexingError ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNoneIndexingError(void) {
+    PyErr_SetString(PyExc_TypeError, "'NoneType' object is unsubscriptable");
+}
+
+/////////////// RaiseNoneIterError.proto ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void);
+
+/////////////// RaiseNoneIterError ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void) {
+    PyErr_SetString(PyExc_TypeError, "'NoneType' object is not iterable");
+}
+
+/////////////// RaiseTooManyValuesToUnpack.proto ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected);
+
+/////////////// RaiseTooManyValuesToUnpack ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected) {
+    PyErr_Format(PyExc_ValueError,
+                 "too many values to unpack (expected %"PY_FORMAT_SIZE_T"d)", expected);
+}
+
+/////////////// RaiseNeedMoreValuesToUnpack.proto ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index);
+
+/////////////// RaiseNeedMoreValuesToUnpack ///////////////
+
+static CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index) {
+    PyErr_Format(PyExc_ValueError,
+                 "need more than %"PY_FORMAT_SIZE_T"d value%s to unpack",
+                 index, (index == 1) ? "" : "s");
+}
+
+/////////////// UnpackTupleError.proto ///////////////
+
+static void __Pyx_UnpackTupleError(PyObject *, Py_ssize_t index); /*proto*/
+
+/////////////// UnpackTupleError ///////////////
+//@requires: RaiseNoneIterError
+//@requires: RaiseNeedMoreValuesToUnpack
+//@requires: RaiseTooManyValuesToUnpack
+
+static void __Pyx_UnpackTupleError(PyObject *t, Py_ssize_t index) {
+    if (t == Py_None) {
+      __Pyx_RaiseNoneNotIterableError();
+    } else if (PyTuple_GET_SIZE(t) < index) {
+      __Pyx_RaiseNeedMoreValuesError(PyTuple_GET_SIZE(t));
+    } else {
+      __Pyx_RaiseTooManyValuesError(index);
+    }
+}
+
+/////////////// UnpackItem.proto ///////////////
+
+static PyObject *__Pyx_UnpackItem(PyObject *, Py_ssize_t index); /*proto*/
+
+/////////////// UnpackItem ///////////////
+//@requires: RaiseNeedMoreValuesToUnpack
+
+static PyObject *__Pyx_UnpackItem(PyObject *iter, Py_ssize_t index) {
+    PyObject *item;
+    if (!(item = PyIter_Next(iter))) {
+        if (!PyErr_Occurred()) {
+            __Pyx_RaiseNeedMoreValuesError(index);
+        }
+    }
+    return item;
+}
+
+/////////////// UnpackItemEndCheck.proto ///////////////
+
+static int __Pyx_IternextUnpackEndCheck(PyObject *retval, Py_ssize_t expected); /*proto*/
+
+/////////////// UnpackItemEndCheck ///////////////
+//@requires: RaiseTooManyValuesToUnpack
+
+static int __Pyx_IternextUnpackEndCheck(PyObject *retval, Py_ssize_t expected) {
+    if (unlikely(retval)) {
+        Py_DECREF(retval);
+        __Pyx_RaiseTooManyValuesError(expected);
+        return -1;
+    } else if (PyErr_Occurred()) {
+        if (likely(PyErr_ExceptionMatches(PyExc_StopIteration))) {
+            PyErr_Clear();
+            return 0;
+        } else {
+            return -1;
+        }
+    }
+    return 0;
+}
+
+/////////////// UnpackTuple2.proto ///////////////
+
+static CYTHON_INLINE int __Pyx_unpack_tuple2(PyObject* tuple, PyObject** value1, PyObject** value2,
+                                             int is_tuple, int has_known_size, int decref_tuple);
+
+/////////////// UnpackTuple2 ///////////////
+//@requires: UnpackItem
+//@requires: UnpackItemEndCheck
+//@requires: UnpackTupleError
+
+static CYTHON_INLINE int __Pyx_unpack_tuple2(PyObject* tuple, PyObject** pvalue1, PyObject** pvalue2,
+                                             int is_tuple, int has_known_size, int decref_tuple) {
+    PyObject *value1 = NULL, *value2 = NULL, *iter = NULL;
+    if (!is_tuple && unlikely(!PyTuple_Check(tuple))) {
+        iter = PyObject_GetIter(tuple);
+        if (unlikely(!iter)) goto bad;
+        if (decref_tuple)
+            Py_DECREF(tuple);
+        value1 = __Pyx_UnpackItem(iter, 0);
+        if (unlikely(!value1)) goto bad;
+        value2 = __Pyx_UnpackItem(iter, 1);
+        if (unlikely(!value2)) goto bad;
+        if (!has_known_size && unlikely(__Pyx_IternextUnpackEndCheck(PyIter_Next(iter), 2))) goto bad;
+        Py_DECREF(iter);
+    } else {
+        if (!has_known_size && unlikely(PyTuple_GET_SIZE(tuple) != 2)) {
+            __Pyx_UnpackTupleError(tuple, 2);
+            goto bad;
+        }
+        value1 = PyTuple_GET_ITEM(tuple, 0);
+        value2 = PyTuple_GET_ITEM(tuple, 1);
+#if CYTHON_COMPILING_IN_CPYTHON
+        if (decref_tuple) {
+            PyTuple_SET_ITEM(tuple, 0, NULL);
+            PyTuple_SET_ITEM(tuple, 1, NULL);
+        } else
+#endif
+        {
+            Py_INCREF(value1);
+            Py_INCREF(value2);
+        }
+        if (decref_tuple)
+            Py_DECREF(tuple);
+    }
+    *pvalue1 = value1;
+    *pvalue2 = value2;
+    return 0;
+bad:
+    Py_XDECREF(iter);
+    Py_XDECREF(value1);
+    Py_XDECREF(value2);
+    if (decref_tuple) Py_DECREF(tuple);
+    return -1;
+}

Cython/Utility/Optimize.c

@@ -299,6 +299,87 @@ static CYTHON_INLINE PyObject* __Pyx_PyDict_Clear(PyObject* d) {
     return Py_None;
 }
 
+/////////////// dict_iter.proto ///////////////
+
+static CYTHON_INLINE PyObject* __Pyx_dict_iterator(PyObject* dict, int keys, int values, int is_dict,
+                                                   PyObject* method_name, Py_ssize_t* p_orig_length, int* p_is_dict);
+static CYTHON_INLINE int __Pyx_dict_iter_next(PyObject* dict_or_iter, Py_ssize_t orig_length, Py_ssize_t* ppos,
+                                              PyObject** pkey, PyObject** pvalue, PyObject** pitem, int is_dict);
+
+/////////////// dict_iter ///////////////
+//@requires: ObjectHandling.c::UnpackTuple2
+
+static CYTHON_INLINE PyObject* __Pyx_dict_iterator(PyObject* dict, int keys, int values, int is_dict,
+                                                   PyObject* method_name, Py_ssize_t* p_orig_length, int* p_is_dict) {
+#if !CYTHON_COMPILING_IN_PYPY
+    if (is_dict || likely(PyDict_CheckExact(dict))) {
+        *p_is_dict = 1;
+        *p_orig_length = PyDict_Size(dict);
+        Py_INCREF(dict);
+        return dict;
+    }
+#endif
+    *p_is_dict = 0;
+    if (!method_name) {
+        return PyObject_GetIter(dict);
+    } else {
+        return PyObject_CallMethodObjArgs(dict, method_name, NULL);
+    }
+}
+
+static CYTHON_INLINE int __Pyx_dict_iter_next(PyObject* dict_or_iter, Py_ssize_t orig_length, Py_ssize_t* ppos,
+                                              PyObject** pkey, PyObject** pvalue, PyObject** pitem, int is_dict) {
+#if !CYTHON_COMPILING_IN_PYPY
+    if (is_dict) {
+        PyObject *key, *value;
+        if (unlikely(orig_length != PyDict_Size(dict_or_iter))) {
+            PyErr_SetString(PyExc_RuntimeError, "dictionary changed size during iteration");
+            return -1;
+        }
+        if (unlikely(!PyDict_Next(dict_or_iter, ppos, &key, &value))) {
+            return 0;
+        }
+        if (pitem) {
+            PyObject* tuple = PyTuple_New(2);
+            if (unlikely(!tuple)) {
+                return -1;
+            }
+            Py_INCREF(key);
+            Py_INCREF(value);
+            PyTuple_SET_ITEM(tuple, 0, key);
+            PyTuple_SET_ITEM(tuple, 1, value);
+            *pitem = tuple;
+        } else {
+            if (pkey) {
+                Py_INCREF(key);
+                *pkey = key;
+            }
+            if (pvalue) {
+                Py_INCREF(value);
+                *pvalue = value;
+            }
+        }
+        return 1;
+    }
+#endif
+    {
+        PyObject* next_item = PyIter_Next(dict_or_iter);
+        if (unlikely(!next_item)) {
+            return PyErr_Occurred() ? -1 : 0;
+        }
+        if (pitem) {
+            *pitem = next_item;
+        } else if (pkey && pvalue) {
+            if (__Pyx_unpack_tuple2(next_item, pkey, pvalue, is_dict, is_dict, 1))
+                return -1;
+        } else if (pkey) {
+            *pkey = next_item;
+        } else {
+            *pvalue = next_item;
+        }
+        return 1;
+    }
+}
 
 /////////////// pyobject_as_double.proto ///////////////
 

tests/run/iterdict.pyx

@@ -56,6 +56,15 @@ def iteritems_int(dict d):
     [(10, 0), (11, 1), (12, 2), (13, 3)]
     >>> iteritems_int({})
     []
+    >>> iteritems_int({'a': 1})
+    Traceback (most recent call last):
+    TypeError: an integer is required
+    >>> iteritems_int({1: 'b'})
+    Traceback (most recent call last):
+    TypeError: an integer is required
+    >>> iteritems_int({'a': 'b'})
+    Traceback (most recent call last):
+    TypeError: an integer is required
     """
     cdef int k,v
     l = []
@@ -113,6 +122,9 @@ def iterkeys_int(dict d):
     [10, 11, 12, 13]
     >>> iterkeys_int({})
     []
+    >>> iterkeys_int({'a': 'b'})
+    Traceback (most recent call last):
+    TypeError: an integer is required
     """
     cdef int k
     l = []
@@ -146,6 +158,9 @@ def iterdict_int(dict d):
     [10, 11, 12, 13]
     >>> iterdict_int({})
     []
+    >>> iterdict_int({'a': 'b'})
+    Traceback (most recent call last):
+    TypeError: an integer is required
     """
     cdef int k
     l = []
@@ -211,6 +226,9 @@ def itervalues_int(dict d):
     [0, 1, 2, 3]
     >>> itervalues_int({})
     []
+    >>> itervalues_int({'a': 'b'})
+    Traceback (most recent call last):
+    TypeError: an integer is required
     """
     cdef int v
     l = []