bpo-32424: Synchronize xml.etree.ElementTree.Element copy methods.

The Python implementation had Element.copy() while the C
implementation had Element.__copy__() and Element.__deepcopy__().

This synchronizes the two implementations and deprecates
elem.copy() in favor of copy.copy(elem).

Author: GPHemsley

Lib/test/test_xml_etree.py

@@ -6,6 +6,7 @@
 # monkey-patched when running the "test_xml_etree_c" test suite.
 
 import copy
+import itertools
 import functools
 import html
 import io
@@ -102,6 +103,11 @@ def newtest(*args, **kwargs):
     return decorator
 
 
+#-----------------------------------------------------------------------------
+# UNIT TESTS
+#-----------------------------------------------------------------------------
+
+
 class ModuleTest(unittest.TestCase):
     def test_sanity(self):
         # Import sanity.
@@ -115,6 +121,143 @@ def test_all(self):
         support.check__all__(self, ET, names, blacklist=("HTML_EMPTY",))
 
 
+class ElementTree_Element_UnitTest(unittest.TestCase):
+
+    def test___init__(self):
+        tag = "foo"
+        attrib = { "zix": "wyp" }
+
+        element_foo = ET.Element(tag, attrib)
+
+        with self.subTest("traits of an element"):
+            self.assertIsInstance(element_foo, ET.Element)
+            self.assertIn("tag", dir(element_foo))
+            self.assertIn("attrib", dir(element_foo))
+            self.assertIn("text", dir(element_foo))
+            self.assertIn("tail", dir(element_foo))
+
+        with self.subTest("string attributes have expected values"):
+            self.assertEqual(element_foo.tag, tag)
+            self.assertIsNone(element_foo.text)
+            self.assertIsNone(element_foo.tail)
+
+        with self.subTest("attrib is a copy"):
+            self.assertIsNot(element_foo.attrib, attrib)
+            self.assertEqual(element_foo.attrib, attrib)
+
+        with self.subTest("attrib isn't linked"):
+            attrib["bar"] = "baz"
+
+            self.assertIsNot(element_foo.attrib, attrib)
+            self.assertNotEqual(element_foo.attrib, attrib)
+
+    def test_copy(self):
+        element_foo = ET.Element("foo", { "zix": "wyp" })
+        element_foo.append(ET.Element("bar", { "baz": "qix" }))
+
+        with self.assertWarns(DeprecationWarning):
+            element_foo2 = element_foo.copy()
+
+        with self.subTest("elements are not the same"):
+            self.assertIsNot(element_foo2, element_foo)
+
+        with self.subTest("string attributes are the same"):
+            self.assertIs(element_foo2.tag, element_foo.tag)
+            self.assertIs(element_foo2.text, element_foo.text)
+            self.assertIs(element_foo2.tail, element_foo.tail)
+
+        with self.subTest("string attributes are equal"):
+            self.assertEqual(element_foo2.tag, element_foo.tag)
+            self.assertEqual(element_foo2.text, element_foo.text)
+            self.assertEqual(element_foo2.tail, element_foo.tail)
+
+        with self.subTest("children are the same"):
+            for (child1, child2) in itertools.zip_longest(element_foo, element_foo2):
+                self.assertIs(child1, child2)
+
+        with self.subTest("attrib is a copy"):
+            self.assertIsNot(element_foo2.attrib, element_foo.attrib)
+            self.assertEqual(element_foo2.attrib, element_foo.attrib)
+
+        with self.subTest("attrib isn't linked"):
+            element_foo.attrib["bar"] = "baz"
+
+            self.assertIsNot(element_foo2.attrib, element_foo.attrib)
+            self.assertNotEqual(element_foo2.attrib, element_foo.attrib)
+
+    def test___copy__(self):
+        element_foo = ET.Element("foo", { "zix": "wyp" })
+        element_foo.append(ET.Element("bar", { "baz": "qix" }))
+
+        element_foo2 = copy.copy(element_foo)
+
+        with self.subTest("elements are not the same"):
+            self.assertIsNot(element_foo2, element_foo)
+
+        with self.subTest("string attributes are the same"):
+            self.assertIs(element_foo2.tag, element_foo.tag)
+            self.assertIs(element_foo2.text, element_foo.text)
+            self.assertIs(element_foo2.tail, element_foo.tail)
+
+        with self.subTest("string attributes are equal"):
+            self.assertEqual(element_foo2.tag, element_foo.tag)
+            self.assertEqual(element_foo2.text, element_foo.text)
+            self.assertEqual(element_foo2.tail, element_foo.tail)
+
+        with self.subTest("children are the same"):
+            for (child1, child2) in itertools.zip_longest(element_foo, element_foo2):
+                self.assertIs(child1, child2)
+
+        with self.subTest("attrib is a copy"):
+            self.assertIsNot(element_foo2.attrib, element_foo.attrib)
+            self.assertEqual(element_foo2.attrib, element_foo.attrib)
+
+        with self.subTest("attrib isn't linked"):
+            element_foo.attrib["bar"] = "baz"
+
+            self.assertIsNot(element_foo2.attrib, element_foo.attrib)
+            self.assertNotEqual(element_foo2.attrib, element_foo.attrib)
+
+    def test___deepcopy__(self):
+        element_foo = ET.Element("foo", { "zix": "wyp" })
+        element_foo.append(ET.Element("bar", { "baz": "qix" }))
+
+        element_foo2 = copy.deepcopy(element_foo)
+
+        with self.subTest("elements are not the same"):
+            self.assertIsNot(element_foo2, element_foo)
+
+        # Strings should still be the same in a deep copy.
+        with self.subTest("string attributes are the same"):
+            self.assertIs(element_foo2.tag, element_foo.tag)
+            self.assertIs(element_foo2.text, element_foo.text)
+            self.assertIs(element_foo2.tail, element_foo.tail)
+
+        with self.subTest("string attributes are equal"):
+            self.assertEqual(element_foo2.tag, element_foo.tag)
+            self.assertEqual(element_foo2.text, element_foo.text)
+            self.assertEqual(element_foo2.tail, element_foo.tail)
+
+        with self.subTest("children are not the same"):
+            for (child1, child2) in itertools.zip_longest(element_foo, element_foo2):
+                self.assertIsNot(child1, child2)
+
+        with self.subTest("attrib is a copy"):
+            self.assertIsNot(element_foo2.attrib, element_foo.attrib)
+            self.assertEqual(element_foo2.attrib, element_foo.attrib)
+
+        with self.subTest("attrib isn't linked"):
+            element_foo.attrib["bar"] = "baz"
+
+            self.assertIsNot(element_foo2.attrib, element_foo.attrib)
+            self.assertNotEqual(element_foo2.attrib, element_foo.attrib)
+
+
+#-----------------------------------------------------------------------------
+# INTEGRATION TESTS
+#-----------------------------------------------------------------------------
+
+
 def serialize(elem, to_string=True, encoding='unicode', **options):
     if encoding != 'unicode':
         file = io.BytesIO()
@@ -1786,36 +1929,38 @@ def test_cyclic_gc(self):
         class Dummy:
             pass
 
-        # Test the shortest cycle: d->element->d
-        d = Dummy()
-        d.dummyref = ET.Element('joe', attr=d)
-        wref = weakref.ref(d)
-        del d
-        gc_collect()
-        self.assertIsNone(wref())
-
-        # A longer cycle: d->e->e2->d
-        e = ET.Element('joe')
-        d = Dummy()
-        d.dummyref = e
-        wref = weakref.ref(d)
-        e2 = ET.SubElement(e, 'foo', attr=d)
-        del d, e, e2
-        gc_collect()
-        self.assertIsNone(wref())
-
-        # A cycle between Element objects as children of one another
-        # e1->e2->e3->e1
-        e1 = ET.Element('e1')
-        e2 = ET.Element('e2')
-        e3 = ET.Element('e3')
-        e1.append(e2)
-        e2.append(e2)
-        e3.append(e1)
-        wref = weakref.ref(e1)
-        del e1, e2, e3
-        gc_collect()
-        self.assertIsNone(wref())
+        with self.subTest("Test the shortest cycle: d->element->d"):
+            d = Dummy()
+            d.dummyref = ET.Element('joe', attr=d)
+            wref = weakref.ref(d)
+            del d
+            gc_collect()
+            self.assertIsNone(wref())
+
+        with self.subTest("A longer cycle: d->e->e2->d"):
+            e = ET.Element('joe')
+            d = Dummy()
+            d.dummyref = e
+            wref = weakref.ref(d)
+            e2 = ET.SubElement(e, 'foo', attr=d)
+            del d, e, e2
+            gc_collect()
+            self.assertIsNone(wref())
+
+        with self.subTest(
+            "A cycle between Element objects as children of one another: "
+            "e1->e2->e3->e1"
+            ):
+            e1 = ET.Element('e1')
+            e2 = ET.Element('e2')
+            e3 = ET.Element('e3')
+            e3.append(e1)
+            e2.append(e3)
+            e1.append(e2)
+            wref = weakref.ref(e1)
+            del e1, e2, e3
+            gc_collect()
+            self.assertIsNone(wref())
 
     def test_weakref(self):
         flag = False
@@ -3143,6 +3288,7 @@ def test_main(module=None):
 
     test_classes = [
         ModuleTest,
+        ElementTree_Element_UnitTest,
         ElementSlicingTest,
         BasicElementTest,
         BadElementTest,

Lib/xml/etree/ElementTree.py

@@ -94,6 +94,7 @@
 import sys
 import re
 import warnings
+import copy
 import io
 import collections
 import collections.abc
@@ -196,12 +197,35 @@ def copy(self):
         original tree.
 
         """
+        warnings.warn(
+            "elem.copy() is deprecated. Use copy.copy(elem) instead.",
+            DeprecationWarning
+            )
+        return self.__copy__()
+
+    def __copy__(self):
         elem = self.makeelement(self.tag, self.attrib)
         elem.text = self.text
         elem.tail = self.tail
         elem[:] = self
         return elem
 
+    def __deepcopy__(self, memo):
+        tag = copy.deepcopy(self.tag, memo)
+        attrib = copy.deepcopy(self.attrib, memo)
+
+        elem = self.makeelement(tag, attrib)
+
+        elem.text = copy.deepcopy(self.text, memo)
+        elem.tail = copy.deepcopy(self.tail, memo)
+
+        for child in self:
+            elem.append(copy.deepcopy(child, memo))
+
+        memo[id(self)] = elem
+
+        return elem
+
     def __len__(self):
         return len(self._children)
 

Modules/_elementtree.c

@@ -283,10 +283,12 @@ create_new_element(PyObject* tag, PyObject* attrib)
     PyObject_GC_Track(self);
 
     if (attrib != Py_None && !is_empty_dict(attrib)) {
+        attrib = PyDict_Copy(attrib);
         if (create_extra(self, attrib) < 0) {
             Py_DECREF(self);
             return NULL;
         }
+        Py_DECREF(attrib);
     }
 
     return (PyObject*) self;
@@ -725,6 +727,24 @@ _elementtree_Element___copy___impl(ElementObject *self)
     return (PyObject*) element;
 }
 
+/*[clinic input]
+_elementtree.Element.copy
+
+[clinic start generated code]*/
+
+static PyObject *
+_elementtree_Element_copy_impl(ElementObject *self)
+/*[clinic end generated code: output=84660e8524276b22 input=1f8134305a7719a3]*/
+{
+    if (PyErr_WarnEx(PyExc_DeprecationWarning,
+                     "elem.copy() is deprecated. Use copy.copy(elem) instead.",
+                     1) < 0) {
+        return NULL;
+    }
+
+    return _elementtree_Element___copy___impl(self);
+}
+
 /* Helper for a deep copy. */
 LOCAL(PyObject *) deepcopy(PyObject *, PyObject *);
 
@@ -3791,6 +3811,7 @@ static PyMethodDef element_methods[] = {
 
     _ELEMENTTREE_ELEMENT_MAKEELEMENT_METHODDEF
 
+    _ELEMENTTREE_ELEMENT_COPY_METHODDEF
     _ELEMENTTREE_ELEMENT___COPY___METHODDEF
     _ELEMENTTREE_ELEMENT___DEEPCOPY___METHODDEF
     _ELEMENTTREE_ELEMENT___SIZEOF___METHODDEF

Modules/clinic/_elementtree.c.h

@@ -62,6 +62,23 @@ _elementtree_Element___copy__(ElementObject *self, PyObject *Py_UNUSED(ignored))
     return _elementtree_Element___copy___impl(self);
 }
 
+PyDoc_STRVAR(_elementtree_Element_copy__doc__,
+"copy($self, /)\n"
+"--\n"
+"\n");
+
+#define _ELEMENTTREE_ELEMENT_COPY_METHODDEF    \
+    {"copy", (PyCFunction)_elementtree_Element_copy, METH_NOARGS, _elementtree_Element_copy__doc__},
+
+static PyObject *
+_elementtree_Element_copy_impl(ElementObject *self);
+
+static PyObject *
+_elementtree_Element_copy(ElementObject *self, PyObject *Py_UNUSED(ignored))
+{
+    return _elementtree_Element_copy_impl(self);
+}
+
 PyDoc_STRVAR(_elementtree_Element___deepcopy____doc__,
 "__deepcopy__($self, memo, /)\n"
 "--\n"
@@ -749,4 +766,4 @@ _elementtree_XMLParser__setevents(XMLParserObject *self, PyObject *const *args,
 exit:
     return return_value;
 }
-/*[clinic end generated code: output=c5a85a88bbb5cc06 input=a9049054013a1b77]*/
+/*[clinic end generated code: output=1d2dc2ddb03bcaeb input=a9049054013a1b77]*/