Implement pyexpat ordered_attributes (#8305)

When `ordered_attributes` is set, report attributes to the
StartElementHandler as a flat `[name, value, ...]` list matching
CPython's my_StartElementHandler, instead of always passing a dict.

`xml.etree.ElementTree` and `xml.dom.minidom` (via expatbuilder) both
enable this flag and rely on the flat-list contract, so parsing any
element with attributes previously failed: the dict made `_start` do
integer indexing, raising KeyError which was silently swallowed by the
handler invoker, leaving the tree empty ("missing toplevel element").

Reference (CPython 3.14):
https://github.com/python/cpython/blob/3.14/Modules/pyexpat.c#L439-L443

Remove `@unittest.expectedFailure` from tests that now pass
(23 in test_xml_etree, 15 in test_minidom, 1 in test_regrtest).

Assisted-by: Claude:claude-opus-4-8

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

Lib/test/test_minidom.py

@@ -553,7 +553,6 @@ class MinidomTest(unittest.TestCase):
         self.assertEqual(str(node), repr(node))
         dom.unlink()
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testWriteXML(self):
         str = '<?xml version="1.0" ?><a b="c"/>'
         dom = parseString(str)
@@ -601,7 +600,6 @@ class MinidomTest(unittest.TestCase):
                 'lflf="&#10;&#10;" '
                 'ws="&#9;&#10;&#13; "/>')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testAltNewline(self):
         str = '<?xml version="1.0" ?>\n<a b="c"/>\n'
         dom = parseString(str)
@@ -705,7 +703,6 @@ class MinidomTest(unittest.TestCase):
 
     def testAttrListKeysNS(self): pass
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testRemoveNamedItem(self):
         doc = parseString("<doc a=''/>")
         e = doc.documentElement
@@ -715,7 +712,6 @@ class MinidomTest(unittest.TestCase):
         self.assertTrue(a1.isSameNode(a2))
         self.assertRaises(xml.dom.NotFoundErr, attrs.removeNamedItem, "a")
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testRemoveNamedItemNS(self):
         doc = parseString("<doc xmlns:a='http://xml.python.org/' a:b=''/>")
         e = doc.documentElement
@@ -788,7 +784,6 @@ class MinidomTest(unittest.TestCase):
         root.setAttribute("added", "VALUE")
         return dom, clone
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testCloneElementShallow(self):
         dom, clone = self._setupCloneElement(0)
         self.assertEqual(len(clone.childNodes), 0)
@@ -940,11 +935,9 @@ class MinidomTest(unittest.TestCase):
         self.confirm(clone.specified,
                 testName + ": cloned attribute must have specified == True")
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testCloneAttributeShallow(self):
         self.check_clone_attribute(0, "testCloneAttributeShallow")
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testCloneAttributeDeep(self):
         self.check_clone_attribute(1, "testCloneAttributeDeep")
 
@@ -1336,7 +1329,6 @@ class MinidomTest(unittest.TestCase):
         self.assertRaises(xml.dom.WrongDocumentErr, doc2.renameNode, node,
                           xml.dom.EMPTY_NAMESPACE, "foo")
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testRenameAttribute(self):
         doc = parseString("<doc a='v'/>")
         elem = doc.documentElement
@@ -1541,7 +1533,6 @@ class MinidomTest(unittest.TestCase):
         self.confirm(text is None
                 and len(elem.childNodes) == 2)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testSchemaType(self):
         doc = parseString(
             "<!DOCTYPE doc [\n"
@@ -1575,7 +1566,6 @@ class MinidomTest(unittest.TestCase):
             self.confirm(hasattr(t, "name")
                     and t.namespace == xml.dom.EMPTY_NAMESPACE)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testSetIdAttribute(self):
         doc = parseString("<doc a1='v' a2='w'/>")
         e = doc.documentElement
@@ -1607,7 +1597,6 @@ class MinidomTest(unittest.TestCase):
         self.confirm(e.isSameNode(doc.getElementById("w"))
                 and a2.isId)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testSetIdAttributeNS(self):
         NS1 = "http://xml.python.org/ns1"
         NS2 = "http://xml.python.org/ns2"
@@ -1644,7 +1633,6 @@ class MinidomTest(unittest.TestCase):
         self.confirm(e.isSameNode(doc.getElementById("w"))
                 and a2.isId)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def testSetIdAttributeNode(self):
         NS1 = "http://xml.python.org/ns1"
         NS2 = "http://xml.python.org/ns2"
@@ -1766,7 +1754,6 @@ class MinidomTest(unittest.TestCase):
         pi = doc.createProcessingInstruction("y", "z")
         pi.nodeValue = "crash"
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_minidom_attribute_order(self):
         xml_str = '<?xml version="1.0" ?><curriculum status="public" company="example"/>'
         doc = parseString(xml_str)
@@ -1774,13 +1761,11 @@ class MinidomTest(unittest.TestCase):
         doc.writexml(output)
         self.assertEqual(output.getvalue(), xml_str)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_toxml_with_attributes_ordered(self):
         xml_str = '<?xml version="1.0" ?><curriculum status="public" company="example"/>'
         doc = parseString(xml_str)
         self.assertEqual(doc.toxml(), xml_str)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_toprettyxml_with_attributes_ordered(self):
         xml_str = '<?xml version="1.0" ?><curriculum status="public" company="example"/>'
         doc = parseString(xml_str)

Lib/test/test_regrtest.py

@@ -2355,7 +2355,6 @@ class ArgsTestCase(BaseTestCase):
             self.check_executed_tests(output, testname, stats=1, parallel=True)
             self.assertNotIn('SPAM SPAM SPAM', output)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON; TypeError: int() argument must be a string, a bytes-like object or a real number, not 'NoneType'
     def test_xml(self):
         code = textwrap.dedent(r"""
             import unittest

Lib/test/test_xml_etree.py

@@ -392,7 +392,6 @@ class ElementTreeTest(unittest.TestCase):
         self.serialize_check(ET.XML("<tag><![CDATA[hello]]></tag>"),
                 '<tag>hello</tag>')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_file_init(self):
         stringfile = io.BytesIO(SAMPLE_XML.encode("utf-8"))
         tree = ET.ElementTree(file=stringfile)
@@ -508,7 +507,6 @@ class ElementTreeTest(unittest.TestCase):
         elem[:] = tuple([subelem])
         self.serialize_check(elem, '<tag><subtag key="value" /></tag>')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_parsefile(self):
         # Test parsing from file.
 
@@ -688,7 +686,6 @@ class ElementTreeTest(unittest.TestCase):
         parser2 = ET.XMLParser()
         self.assertIsInstance(parser2.target, ET.TreeBuilder)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_children(self):
         # Test Element children iteration
 
@@ -1091,7 +1088,6 @@ class ElementTreeTest(unittest.TestCase):
         self.assertEqual(str(cm.exception),
                 'undefined entity &entity;: line 4, column 10')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_namespace(self):
         # Test namespace issues.
 
@@ -1335,7 +1331,6 @@ class ElementTreeTest(unittest.TestCase):
 class IterparseTest(unittest.TestCase):
     # Test iterparse interface.
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_basic(self):
         iterparse = ET.iterparse
 
@@ -1361,7 +1356,6 @@ class IterparseTest(unittest.TestCase):
             ])
         it.close()
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_external_file(self):
         with open(SIMPLE_XMLFILE, 'rb') as source:
             it = ET.iterparse(source)
@@ -1374,7 +1368,6 @@ class IterparseTest(unittest.TestCase):
                 ])
             self.assertEqual(it.root.tag, 'root')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_events(self):
         iterparse = ET.iterparse
 
@@ -1475,7 +1468,6 @@ class IterparseTest(unittest.TestCase):
         with self.assertRaises(FileNotFoundError):
             ET.iterparse("nonexistent")
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_resource_warnings_not_exhausted(self):
         # Not exhausting the iterator still closes the underlying file (bpo-43292)
         it = ET.iterparse(SIMPLE_XMLFILE)
@@ -1514,7 +1506,6 @@ class IterparseTest(unittest.TestCase):
             del it
             gc_collect()
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_close_not_exhausted(self):
         iterparse = ET.iterparse
 
@@ -2041,7 +2032,6 @@ class XIncludeTest(unittest.TestCase):
         else:
             return None
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_xinclude_default(self):
         from xml.etree import ElementInclude
         doc = self.xinclude_loader('default.xml')
@@ -2056,7 +2046,6 @@ class XIncludeTest(unittest.TestCase):
             '</root>\n'
             '</document>')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_xinclude(self):
         from xml.etree import ElementInclude
 
@@ -2121,7 +2110,6 @@ class XIncludeTest(unittest.TestCase):
             '  </ns0:include>\n'
             '</div>') # C5
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_xinclude_repeated(self):
         from xml.etree import ElementInclude
 
@@ -2129,7 +2117,6 @@ class XIncludeTest(unittest.TestCase):
         ElementInclude.include(document, self.xinclude_loader)
         self.assertEqual(1+4*2, len(document.findall(".//p")))
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_xinclude_failures(self):
         from xml.etree import ElementInclude
 
@@ -2234,7 +2221,6 @@ class BugsTest(unittest.TestCase):
         elem.set("123", 123)
         check(elem) # attribute value
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_bug_xmltoolkit25(self):
         # typo in ElementTree.findtext
 
@@ -2258,7 +2244,6 @@ class BugsTest(unittest.TestCase):
             ET.dump(tree)
             self.assertEqual(stdout.getvalue(), '<doc><table><tbody /></table></doc>\n')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_bug_xmltoolkit39(self):
         # non-ascii element and attribute names doesn't work
 
@@ -2345,7 +2330,6 @@ class BugsTest(unittest.TestCase):
             xmltoolkit63()
         self.assertEqual(sys.getrefcount(None), count)
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_bug_200708_newline(self):
         # Preserve newlines in attributes.
 
@@ -2461,7 +2445,6 @@ class BugsTest(unittest.TestCase):
                 b"<?xml version='1.0' encoding='ascii'?>\n"
                 b'<body>t&#227;g</body>')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_issue6565(self):
         elem = ET.XML("<body><tag/></body>")
         self.assertEqual(summarize_list(elem), ['tag'])
@@ -2533,7 +2516,6 @@ class BugsTest(unittest.TestCase):
         root = ET.XML(xml)
         self.assertEqual(root.get('b'), text.decode('utf-8'))
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_expat224_utf8_bug(self):
         # bpo-31170: Expat 2.2.3 had a bug in its UTF-8 decoder.
         # Check that Expat 2.2.4 fixed the bug.
@@ -3355,7 +3337,6 @@ class ElementTreeTypeTest(unittest.TestCase):
 
 
 class ElementFindTest(unittest.TestCase):
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_find_simple(self):
         e = ET.XML(SAMPLE_XML)
         self.assertEqual(e.find('tag').tag, 'tag')
@@ -3379,7 +3360,6 @@ class ElementFindTest(unittest.TestCase):
         # Issue #16922
         self.assertEqual(ET.XML('<tag><empty /></tag>').findtext('empty'), '')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_find_xpath(self):
         LINEAR_XML = '''
         <body>
@@ -3402,7 +3382,6 @@ class ElementFindTest(unittest.TestCase):
         self.assertRaisesRegex(SyntaxError, 'XPath', e.find, './tag[last()-0]')
         self.assertRaisesRegex(SyntaxError, 'XPath', e.find, './tag[last()+1]')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_findall(self):
         e = ET.XML(SAMPLE_XML)
         e[2] = ET.XML(SAMPLE_SECTION)
@@ -3591,7 +3570,6 @@ class ElementFindTest(unittest.TestCase):
         with self.assertRaisesRegex(SyntaxError, 'cannot use absolute path'):
             e.findall('/tag')
 
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_find_through_ElementTree(self):
         e = ET.XML(SAMPLE_XML)
         self.assertEqual(ET.ElementTree(e).find('tag').tag, 'tag')
@@ -4576,7 +4554,6 @@ class NoAcceleratorTest(unittest.TestCase):
 # --------------------------------------------------------------------
 
 class BoolTest(unittest.TestCase):
-    @unittest.expectedFailure  # TODO: RUSTPYTHON
     def test_warning(self):
         e = ET.fromstring('<a style="new"></a>')
         msg = (

crates/stdlib/src/pyexpat.rs

@@ -43,7 +43,8 @@ macro_rules! create_bool_property {
 #[pymodule(name = "pyexpat")]
 mod _pyexpat {
     use crate::vm::{
-        Context, Py, PyObjectRef, PyPayload, PyRef, PyResult, TryFromObject, VirtualMachine,
+        AsObject, Context, Py, PyObjectRef, PyPayload, PyRef, PyResult, TryFromObject,
+        VirtualMachine,
         builtins::{PyBytesRef, PyException, PyModule, PyStr, PyStrRef, PyType, PyUtf8StrRef},
         extend_module,
         function::{ArgBytesLike, ArgPrimitiveIndex, Either, IntoFuncArgs, OptionalArg},
@@ -355,19 +356,30 @@ mod _pyexpat {
                     XmlEvent::StartElement {
                         name, attributes, ..
                     } => {
-                        let dict = vm.ctx.new_dict();
-                        for attribute in attributes {
-                            let attr_name = self.make_name(&attribute.name);
-                            dict.set_item(
-                                attr_name.as_str(),
-                                vm.ctx.new_str(attribute.value).into(),
-                                vm,
-                            )
-                            .unwrap();
-                        }
+                        let ordered = self.ordered_attributes.read().is(&vm.ctx.true_value);
+                        // Build the container.
+                        let attrs: PyObjectRef = if ordered {
+                            let mut items = Vec::with_capacity(attributes.len() * 2);
+                            for attribute in attributes {
+                                items.push(vm.ctx.new_str(self.make_name(&attribute.name)).into());
+                                items.push(vm.ctx.new_str(attribute.value).into());
+                            }
+                            vm.ctx.new_list(items).into()
+                        } else {
+                            let dict = vm.ctx.new_dict();
+                            for attribute in attributes {
+                                dict.set_item(
+                                    self.make_name(&attribute.name).as_str(),
+                                    vm.ctx.new_str(attribute.value).into(),
+                                    vm,
+                                )
+                                .unwrap();
+                            }
+                            dict.into()
+                        };
 
                         let name_str = PyStr::from(self.make_name(&name)).into_ref(&vm.ctx);
-                        invoke_handler(vm, &self.start_element, (name_str, dict));
+                        invoke_handler(vm, &self.start_element, (name_str, attrs));
                     }
                     XmlEvent::EndElement { name, .. } => {
                         let name_str = PyStr::from(self.make_name(&name)).into_ref(&vm.ctx);