diff -r 12ce8e0413f1 Lib/test/test_xml_etree.py
--- a/Lib/test/test_xml_etree.py Fri Sep 02 12:12:23 2016 +0200
+++ b/Lib/test/test_xml_etree.py Fri Sep 02 12:56:38 2016 +0200
@@ -17,6 +17,7 @@ import warnings
import weakref
from itertools import product
+from unittest import mock
from test import support
from test.support import TESTFN, findfile, import_fresh_module, gc_collect, swap_attr
@@ -157,8 +158,8 @@ class ElementTestCase:
class ElementTreeTest(unittest.TestCase):
- def serialize_check(self, elem, expected):
- self.assertEqual(serialize(elem), expected)
+ def serialize_check(self, elem, expected, **options):
+ self.assertEqual(serialize(elem, **options), expected)
def test_interface(self):
# Test element tree interface.
@@ -490,7 +491,6 @@ class ElementTreeTest(unittest.TestCase)
self.assertEqual(b"".join(ET.tostringlist(element)),
b'text')
self.assertEqual(ET.tostring(element, "ascii"),
- b"\n"
b"text")
_, ids = ET.XMLID("text")
self.assertEqual(len(ids), 0)
@@ -498,6 +498,59 @@ class ElementTreeTest(unittest.TestCase)
self.assertEqual(len(ids), 1)
self.assertEqual(ids["body"].tag, 'body')
+ def test_write_xml_declaration(self):
+ elem = ET.XML("")
+
+ # Encodings compatible with UTF-8: XML declaration is optional
+ utf8_encodings = ("ASCII", "us-ascii", "UTF-8", "utf8")
+ other_encodings = ("latin9", "GBK")
+
+ # No XML declaration by default
+ #
+ # By default, the "unicode" encoding doesn't use
+ # locale.getpreferredencoding().
+ with mock.patch('locale.getpreferredencoding', return_value='latin9'):
+ self.serialize_check(elem, '')
+ for encoding in utf8_encodings:
+ self.serialize_check(elem, b'',
+ encoding=encoding)
+
+ # Test the "unicode" encoding with XML declaration which relies on
+ # locale.getpreferredencoding()
+ with mock.patch('locale.getpreferredencoding', return_value='UTF-8'):
+ self.serialize_check(elem,
+ "\n"
+ "",
+ xml_declaration=True)
+
+ with mock.patch('locale.getpreferredencoding', return_value='latin9'):
+ self.serialize_check(elem,
+ "\n"
+ "",
+ xml_declaration=True)
+
+ # Force XML declaration
+ for encoding in utf8_encodings:
+ expected = (f"\n"
+ f"").encode(encoding)
+ self.serialize_check(elem, expected,
+ encoding=encoding,
+ xml_declaration=True)
+
+ # Non-UTF-8 encodings: XML declaration is always needed
+ for encoding in other_encodings:
+ expected = (f"\n"
+ f"").encode(encoding)
+ self.serialize_check(elem, expected,
+ encoding=encoding)
+ self.serialize_check(elem, expected,
+ encoding=encoding,
+ xml_declaration=True)
+
+ # Invalid codec name
+ with self.assertRaises(LookupError):
+ serialize(elem, encoding="xxxxx")
+
def test_iterparse(self):
# Test iterparse interface.
@@ -1636,12 +1689,10 @@ class BugsTest(unittest.TestCase):
e = ET.XML(b""
b't\xc3\xa3g')
self.assertEqual(ET.tostring(e, 'ascii'),
- b"\n"
b'tãg')
e = ET.XML(b""
b't\xe3g')
self.assertEqual(ET.tostring(e, 'ascii'),
- b"\n"
b'tãg')
def test_issue3151(self):
diff -r 12ce8e0413f1 Lib/xml/etree/ElementTree.py
--- a/Lib/xml/etree/ElementTree.py Fri Sep 02 12:12:23 2016 +0200
+++ b/Lib/xml/etree/ElementTree.py Fri Sep 02 12:56:38 2016 +0200
@@ -91,12 +91,13 @@
VERSION = "1.3.0"
-import sys
-import re
-import warnings
-import io
+import codecs
import collections
import contextlib
+import io
+import re
+import sys
+import warnings
from . import ElementPath
@@ -756,10 +757,13 @@ class ElementTree:
else:
encoding = "us-ascii"
enc_lower = encoding.lower()
+ if enc_lower != "unicode":
+ # Normalize the encoding name
+ enc_lower = codecs.lookup(enc_lower).name
with _get_writer(file_or_filename, enc_lower) as write:
if method == "xml" and (xml_declaration or
(xml_declaration is None and
- enc_lower not in ("utf-8", "us-ascii", "unicode"))):
+ enc_lower not in {"utf-8", "ascii", "unicode"})):
declared_encoding = encoding
if enc_lower == "unicode":
# Retrieve the default encoding for the xml declaration