RustPython
diff --git a/‎Lib/test/test_xml_etree.py‎
Lines changed: 77 additions & 25 deletions b/‎Lib/test/test_xml_etree.py‎
Lines changed: 77 additions & 25 deletions
diff --git a/‎Lib/xml/dom/expatbuilder.py‎
Lines changed: 4 additions & 4 deletions b/‎Lib/xml/dom/expatbuilder.py‎
Lines changed: 4 additions & 4 deletions
diff --git a/‎Lib/xml/dom/minidom.py‎
Lines changed: 43 additions & 13 deletions b/‎Lib/xml/dom/minidom.py‎
Lines changed: 43 additions & 13 deletions
@@ -10,7 +10,6 @@
 import html
 import io
 import itertools
-import locale
 import operator
 import os
 import pickle
@@ -130,6 +129,9 @@ def newtest(*args, **kwargs):
         return newtest
     return decorator
 
+def convlinesep(data):
+    return data.replace(b'\n', os.linesep.encode())
+
 
 class ModuleTest(unittest.TestCase):
     def test_sanity(self):
@@ -1023,17 +1025,15 @@ def test_tostring_xml_declaration(self):
     @unittest.expectedFailure
     def test_tostring_xml_declaration_unicode_encoding(self):
         elem = ET.XML('<body><tag/></body>')
-        preferredencoding = locale.getpreferredencoding()
         self.assertEqual(
-            f"<?xml version='1.0' encoding='{preferredencoding}'?>\n<body><tag /></body>",
-            ET.tostring(elem, encoding='unicode', xml_declaration=True)
+            ET.tostring(elem, encoding='unicode', xml_declaration=True),
+            "<?xml version='1.0' encoding='utf-8'?>\n<body><tag /></body>"
         )
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_tostring_xml_declaration_cases(self):
         elem = ET.XML('<body><tag>ø</tag></body>')
-        preferredencoding = locale.getpreferredencoding()
         TESTCASES = [
         #   (expected_retval,                  encoding, xml_declaration)
             # ... xml_declaration = None
@@ -1060,7 +1060,7 @@ def test_tostring_xml_declaration_cases(self):
              b"<body><tag>&#248;</tag></body>", 'US-ASCII', True),
             (b"<?xml version='1.0' encoding='ISO-8859-1'?>\n"
              b"<body><tag>\xf8</tag></body>", 'ISO-8859-1', True),
-            (f"<?xml version='1.0' encoding='{preferredencoding}'?>\n"
+            ("<?xml version='1.0' encoding='utf-8'?>\n"
              "<body><tag>ø</tag></body>", 'unicode', True),
 
         ]
@@ -1102,11 +1102,10 @@ def test_tostringlist_xml_declaration(self):
             b"<?xml version='1.0' encoding='us-ascii'?>\n<body><tag /></body>"
         )
 
-        preferredencoding = locale.getpreferredencoding()
         stringlist = ET.tostringlist(elem, encoding='unicode', xml_declaration=True)
         self.assertEqual(
             ''.join(stringlist),
-            f"<?xml version='1.0' encoding='{preferredencoding}'?>\n<body><tag /></body>"
+            "<?xml version='1.0' encoding='utf-8'?>\n<body><tag /></body>"
         )
         self.assertRegex(stringlist[0], r"^<\?xml version='1.0' encoding='.+'?>")
         self.assertEqual(['<body', '>', '<tag', ' />', '</body>'], stringlist[1:])
@@ -3914,54 +3913,107 @@ def test_encoding(self):
     @unittest.expectedFailure
     def test_write_to_filename(self):
         self.addCleanup(os_helper.unlink, TESTFN)
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         tree.write(TESTFN)
         with open(TESTFN, 'rb') as f:
-            self.assertEqual(f.read(), b'''<site />''')
+            self.assertEqual(f.read(), b'''<site>&#248;</site>''')
+
+    def test_write_to_filename_with_encoding(self):
+        self.addCleanup(os_helper.unlink, TESTFN)
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
+        tree.write(TESTFN, encoding='utf-8')
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(), b'''<site>\xc3\xb8</site>''')
+
+        tree.write(TESTFN, encoding='ISO-8859-1')
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(), convlinesep(
+                             b'''<?xml version='1.0' encoding='ISO-8859-1'?>\n'''
+                             b'''<site>\xf8</site>'''))
+
+    def test_write_to_filename_as_unicode(self):
+        self.addCleanup(os_helper.unlink, TESTFN)
+        with open(TESTFN, 'w') as f:
+            encoding = f.encoding
+        os_helper.unlink(TESTFN)
+
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
+        tree.write(TESTFN, encoding='unicode')
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(), b"<site>\xc3\xb8</site>")
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_text_file(self):
         self.addCleanup(os_helper.unlink, TESTFN)
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         with open(TESTFN, 'w', encoding='utf-8') as f:
             tree.write(f, encoding='unicode')
             self.assertFalse(f.closed)
         with open(TESTFN, 'rb') as f:
-            self.assertEqual(f.read(), b'''<site />''')
+            self.assertEqual(f.read(), b'''<site>\xc3\xb8</site>''')
+
+        with open(TESTFN, 'w', encoding='ascii', errors='xmlcharrefreplace') as f:
+            tree.write(f, encoding='unicode')
+            self.assertFalse(f.closed)
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(),  b'''<site>&#248;</site>''')
+
+        with open(TESTFN, 'w', encoding='ISO-8859-1') as f:
+            tree.write(f, encoding='unicode')
+            self.assertFalse(f.closed)
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(), b'''<site>\xf8</site>''')
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_binary_file(self):
         self.addCleanup(os_helper.unlink, TESTFN)
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         with open(TESTFN, 'wb') as f:
             tree.write(f)
             self.assertFalse(f.closed)
         with open(TESTFN, 'rb') as f:
-            self.assertEqual(f.read(), b'''<site />''')
+            self.assertEqual(f.read(), b'''<site>&#248;</site>''')
+
+    def test_write_to_binary_file_with_encoding(self):
+        self.addCleanup(os_helper.unlink, TESTFN)
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
+        with open(TESTFN, 'wb') as f:
+            tree.write(f, encoding='utf-8')
+            self.assertFalse(f.closed)
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(), b'''<site>\xc3\xb8</site>''')
+
+        with open(TESTFN, 'wb') as f:
+            tree.write(f, encoding='ISO-8859-1')
+            self.assertFalse(f.closed)
+        with open(TESTFN, 'rb') as f:
+            self.assertEqual(f.read(),
+                             b'''<?xml version='1.0' encoding='ISO-8859-1'?>\n'''
+                             b'''<site>\xf8</site>''')
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_binary_file_with_bom(self):
         self.addCleanup(os_helper.unlink, TESTFN)
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         # test BOM writing to buffered file
         with open(TESTFN, 'wb') as f:
             tree.write(f, encoding='utf-16')
             self.assertFalse(f.closed)
         with open(TESTFN, 'rb') as f:
             self.assertEqual(f.read(),
                     '''<?xml version='1.0' encoding='utf-16'?>\n'''
-                    '''<site />'''.encode("utf-16"))
+                    '''<site>\xf8</site>'''.encode("utf-16"))
         # test BOM writing to non-buffered file
         with open(TESTFN, 'wb', buffering=0) as f:
             tree.write(f, encoding='utf-16')
             self.assertFalse(f.closed)
         with open(TESTFN, 'rb') as f:
             self.assertEqual(f.read(),
                     '''<?xml version='1.0' encoding='utf-16'?>\n'''
-                    '''<site />'''.encode("utf-16"))
+                    '''<site>\xf8</site>'''.encode("utf-16"))
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
@@ -3974,10 +4026,10 @@ def test_read_from_stringio(self):
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_stringio(self):
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         stream = io.StringIO()
         tree.write(stream, encoding='unicode')
-        self.assertEqual(stream.getvalue(), '''<site />''')
+        self.assertEqual(stream.getvalue(), '''<site>\xf8</site>''')
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
@@ -3990,10 +4042,10 @@ def test_read_from_bytesio(self):
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_bytesio(self):
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         raw = io.BytesIO()
         tree.write(raw)
-        self.assertEqual(raw.getvalue(), b'''<site />''')
+        self.assertEqual(raw.getvalue(), b'''<site>&#248;</site>''')
 
     class dummy:
         pass
@@ -4011,12 +4063,12 @@ def test_read_from_user_text_reader(self):
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_user_text_writer(self):
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         stream = io.StringIO()
         writer = self.dummy()
         writer.write = stream.write
         tree.write(writer, encoding='unicode')
-        self.assertEqual(stream.getvalue(), '''<site />''')
+        self.assertEqual(stream.getvalue(), '''<site>\xf8</site>''')
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
@@ -4032,12 +4084,12 @@ def test_read_from_user_binary_reader(self):
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
     def test_write_to_user_binary_writer(self):
-        tree = ET.ElementTree(ET.XML('''<site />'''))
+        tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
         raw = io.BytesIO()
         writer = self.dummy()
         writer.write = raw.write
         tree.write(writer)
-        self.assertEqual(raw.getvalue(), b'''<site />''')
+        self.assertEqual(raw.getvalue(), b'''<site>&#248;</site>''')
 
     # TODO: RUSTPYTHON
     @unittest.expectedFailure
 
@@ -204,11 +204,11 @@ def parseFile(self, file):
                 buffer = file.read(16*1024)
                 if not buffer:
                     break
-                parser.Parse(buffer, 0)
+                parser.Parse(buffer, False)
                 if first_buffer and self.document.documentElement:
                     self._setup_subset(buffer)
                 first_buffer = False
-            parser.Parse("", True)
+            parser.Parse(b"", True)
         except ParseEscape:
             pass
         doc = self.document
@@ -637,7 +637,7 @@ def parseString(self, string):
         nsattrs = self._getNSattrs() # get ns decls from node's ancestors
         document = _FRAGMENT_BUILDER_TEMPLATE % (ident, subset, nsattrs)
         try:
-            parser.Parse(document, 1)
+            parser.Parse(document, True)
         except:
             self.reset()
             raise
@@ -697,7 +697,7 @@ def external_entity_ref_handler(self, context, base, systemId, publicId):
             self.fragment = self.document.createDocumentFragment()
             self.curNode = self.fragment
             try:
-                parser.Parse(self._source, 1)
+                parser.Parse(self._source, True)
             finally:
                 self.curNode = old_cur_node
                 self.document = old_document
 
@@ -43,10 +43,11 @@ class Node(xml.dom.Node):
     def __bool__(self):
         return True
 
-    def toxml(self, encoding=None):
-        return self.toprettyxml("", "", encoding)
+    def toxml(self, encoding=None, standalone=None):
+        return self.toprettyxml("", "", encoding, standalone)
 
-    def toprettyxml(self, indent="\t", newl="\n", encoding=None):
+    def toprettyxml(self, indent="\t", newl="\n", encoding=None,
+                    standalone=None):
         if encoding is None:
             writer = io.StringIO()
         else:
@@ -56,7 +57,7 @@ def toprettyxml(self, indent="\t", newl="\n", encoding=None):
                                       newline='\n')
         if self.nodeType == Node.DOCUMENT_NODE:
             # Can pass encoding only to document, to put it into XML header
-            self.writexml(writer, "", indent, newl, encoding)
+            self.writexml(writer, "", indent, newl, encoding, standalone)
         else:
             self.writexml(writer, "", indent, newl)
         if encoding is None:
@@ -718,6 +719,14 @@ def unlink(self):
         Node.unlink(self)
 
     def getAttribute(self, attname):
+        """Returns the value of the specified attribute.
+
+        Returns the value of the element's attribute named attname as
+        a string. An empty string is returned if the element does not
+        have such an attribute. Note that an empty string may also be
+        returned as an explicitly given attribute value, use the
+        hasAttribute method to distinguish these two cases.
+        """
         if self._attrs is None:
             return ""
         try:
@@ -823,10 +832,16 @@ def removeAttributeNode(self, node):
         # Restore this since the node is still useful and otherwise
         # unlinked
         node.ownerDocument = self.ownerDocument
+        return node
 
     removeAttributeNodeNS = removeAttributeNode
 
     def hasAttribute(self, name):
+        """Checks whether the element has an attribute with the specified name.
+
+        Returns True if the element has an attribute with the specified name.
+        Otherwise, returns False.
+        """
         if self._attrs is None:
             return False
         return name in self._attrs
@@ -837,6 +852,11 @@ def hasAttributeNS(self, namespaceURI, localName):
         return (namespaceURI, localName) in self._attrsNS
 
     def getElementsByTagName(self, name):
+        """Returns all descendant elements with the given tag name.
+
+        Returns the list of all descendant elements (not direct children
+        only) with the specified tag name.
+        """
         return _get_elements_by_tagName_helper(self, name, NodeList())
 
     def getElementsByTagNameNS(self, namespaceURI, localName):
@@ -847,22 +867,27 @@ def __repr__(self):
         return "<DOM Element: %s at %#x>" % (self.tagName, id(self))
 
     def writexml(self, writer, indent="", addindent="", newl=""):
+        """Write an XML element to a file-like object
+
+        Write the element to the writer object that must provide
+        a write method (e.g. a file or StringIO object).
+        """
         # indent = current indentation
         # addindent = indentation to add to higher levels
         # newl = newline string
         writer.write(indent+"<" + self.tagName)
 
         attrs = self._get_attributes()
-        a_names = sorted(attrs.keys())
 
-        for a_name in a_names:
+        for a_name in attrs.keys():
             writer.write(" %s=\"" % a_name)
             _write_data(writer, attrs[a_name].value)
             writer.write("\"")
         if self.childNodes:
             writer.write(">")
             if (len(self.childNodes) == 1 and
-                self.childNodes[0].nodeType == Node.TEXT_NODE):
+                self.childNodes[0].nodeType in (
+                        Node.TEXT_NODE, Node.CDATA_SECTION_NODE)):
                 self.childNodes[0].writexml(writer, '', '', '')
             else:
                 writer.write(newl)
@@ -1786,12 +1811,17 @@ def importNode(self, node, deep):
             raise xml.dom.NotSupportedErr("cannot import document type nodes")
         return _clone_node(node, deep, self)
 
-    def writexml(self, writer, indent="", addindent="", newl="", encoding=None):
-        if encoding is None:
-            writer.write('<?xml version="1.0" ?>'+newl)
-        else:
-            writer.write('<?xml version="1.0" encoding="%s"?>%s' % (
-                encoding, newl))
+    def writexml(self, writer, indent="", addindent="", newl="", encoding=None,
+                 standalone=None):
+        declarations = []
+
+        if encoding:
+            declarations.append(f'encoding="{encoding}"')
+        if standalone is not None:
+            declarations.append(f'standalone="{"yes" if standalone else "no"}"')
+
+        writer.write(f'<?xml version="1.0" {" ".join(declarations)}?>{newl}')
+
         for node in self.childNodes:
             node.writexml(writer, indent, addindent, newl)