Mirror
/
cpython
mirror of https://github.com/python/cpython
4
1
Fork 0
Code Issues Releases Wiki Activity

[3.11] gh-91810: ElementTree: Use text file's encoding by default in XML declaration (GH-91903) (GH-92663) 

ElementTree method write() and function tostring() now use the text file's
encoding ("UTF-8" if not available) instead of locale encoding in XML
declaration when encoding="unicode" is specified.
(cherry picked from commit 707839b0fe)


Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>

Automerge-Triggered-By: GH:serhiy-storchaka
This commit is contained in:
Miss Islington (bot) 2022-05-11 10:36:52 -07:00 committed by GitHub
parent 5ea8a93e1a
commit 20fec2c265
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 29 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
Lib/test/test_xml_etree.py
View File

@ -10,7 +10,6 @@ import functools
import html import html
import io import io
import itertools import itertools
import locale
import operator import operator
import os import os
import pickle import pickle
@ -978,15 +977,13 @@ class ElementTreeTest(unittest.TestCase):
def test_tostring_xml_declaration_unicode_encoding(self): def test_tostring_xml_declaration_unicode_encoding(self):
elem = ET.XML('<body><tag/></body>') elem = ET.XML('<body><tag/></body>')
preferredencoding = locale.getpreferredencoding()
self.assertEqual( 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>"
) )
def test_tostring_xml_declaration_cases(self): def test_tostring_xml_declaration_cases(self):
elem = ET.XML('<body><tag>ø</tag></body>') elem = ET.XML('<body><tag>ø</tag></body>')
preferredencoding = locale.getpreferredencoding()
TESTCASES = [ TESTCASES = [
# (expected_retval, encoding, xml_declaration) # (expected_retval, encoding, xml_declaration)
# ... xml_declaration = None # ... xml_declaration = None
@ -1013,7 +1010,7 @@ class ElementTreeTest(unittest.TestCase):
b"<body><tag>&#248;</tag></body>", 'US-ASCII', True), b"<body><tag>&#248;</tag></body>", 'US-ASCII', True),
(b"<?xml version='1.0' encoding='ISO-8859-1'?>\n" (b"<?xml version='1.0' encoding='ISO-8859-1'?>\n"
b"<body><tag>\xf8</tag></body>", 'ISO-8859-1', True), 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), "<body><tag>ø</tag></body>", 'unicode', True),
] ]
@ -1051,11 +1048,10 @@ class ElementTreeTest(unittest.TestCase):
b"<?xml version='1.0' encoding='us-ascii'?>\n<body><tag /></body>" b"<?xml version='1.0' encoding='us-ascii'?>\n<body><tag /></body>"
) )
preferredencoding = locale.getpreferredencoding()
stringlist = ET.tostringlist(elem, encoding='unicode', xml_declaration=True) stringlist = ET.tostringlist(elem, encoding='unicode', xml_declaration=True)
self.assertEqual( self.assertEqual(
''.join(stringlist), ''.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.assertRegex(stringlist[0], r"^<\?xml version='1.0' encoding='.+'?>")
self.assertEqual(['<body', '>', '<tag', ' />', '</body>'], stringlist[1:]) self.assertEqual(['<body', '>', '<tag', ' />', '</body>'], stringlist[1:])
@ -3740,17 +3736,16 @@ class IOTest(unittest.TestCase):
encoding = f.encoding encoding = f.encoding
os_helper.unlink(TESTFN) os_helper.unlink(TESTFN)
try:
'\xf8'.encode(encoding)
except UnicodeEncodeError:
self.skipTest(f'default file encoding {encoding} not supported')
tree = ET.ElementTree(ET.XML('''<site>\xf8</site>''')) tree = ET.ElementTree(ET.XML('''<site>\xf8</site>'''))
tree.write(TESTFN, encoding='unicode') tree.write(TESTFN, encoding='unicode')
with open(TESTFN, 'rb') as f: with open(TESTFN, 'rb') as f:
data = f.read() data = f.read()
expected = "<site>\xf8</site>".encode(encoding, 'xmlcharrefreplace') expected = "<site>\xf8</site>".encode(encoding, 'xmlcharrefreplace')
self.assertEqual(data, expected) if encoding.lower() in ('utf-8', 'ascii'):
self.assertEqual(data, expected)
else:
self.assertIn(b"<?xml version='1.0' encoding=", data)
self.assertIn(expected, data)
def test_write_to_text_file(self): def test_write_to_text_file(self):
self.addCleanup(os_helper.unlink, TESTFN) self.addCleanup(os_helper.unlink, TESTFN)
@ -3765,13 +3760,17 @@ class IOTest(unittest.TestCase):
tree.write(f, encoding='unicode') tree.write(f, encoding='unicode')
self.assertFalse(f.closed) self.assertFalse(f.closed)
with open(TESTFN, 'rb') as f: with open(TESTFN, 'rb') as f:
self.assertEqual(f.read(), b'''<site>&#248;</site>''') self.assertEqual(f.read(), convlinesep(
b'''<?xml version='1.0' encoding='ascii'?>\n'''
b'''<site>&#248;</site>'''))
with open(TESTFN, 'w', encoding='ISO-8859-1') as f: with open(TESTFN, 'w', encoding='ISO-8859-1') as f:
tree.write(f, encoding='unicode') tree.write(f, encoding='unicode')
self.assertFalse(f.closed) self.assertFalse(f.closed)
with open(TESTFN, 'rb') as f: with open(TESTFN, 'rb') as f:
self.assertEqual(f.read(), b'''<site>\xf8</site>''') self.assertEqual(f.read(), convlinesep(
b'''<?xml version='1.0' encoding='ISO-8859-1'?>\n'''
b'''<site>\xf8</site>'''))
def test_write_to_binary_file(self): def test_write_to_binary_file(self):
self.addCleanup(os_helper.unlink, TESTFN) self.addCleanup(os_helper.unlink, TESTFN)

23
Lib/xml/etree/ElementTree.py
View File

@ -728,16 +728,10 @@ class ElementTree:
encoding = "utf-8" encoding = "utf-8"
else: else:
encoding = "us-ascii" encoding = "us-ascii"
enc_lower = encoding.lower() with _get_writer(file_or_filename, encoding) as (write, declared_encoding):
with _get_writer(file_or_filename, enc_lower) as write:
if method == "xml" and (xml_declaration or if method == "xml" and (xml_declaration or
(xml_declaration is None and (xml_declaration is None and
enc_lower not in ("utf-8", "us-ascii", "unicode"))): declared_encoding.lower() not in ("utf-8", "us-ascii"))):
declared_encoding = encoding
if enc_lower == "unicode":
# Retrieve the default encoding for the xml declaration
import locale
declared_encoding = locale.getpreferredencoding()
write("<?xml version='1.0' encoding='%s'?>\n" % ( write("<?xml version='1.0' encoding='%s'?>\n" % (
declared_encoding,)) declared_encoding,))
if method == "text": if method == "text":
@ -762,19 +756,20 @@ def _get_writer(file_or_filename, encoding):
write = file_or_filename.write write = file_or_filename.write
except AttributeError: except AttributeError:
# file_or_filename is a file name # file_or_filename is a file name
if encoding == "unicode": if encoding.lower() == "unicode":
file = open(file_or_filename, "w") file = open(file_or_filename, "w",
errors="xmlcharrefreplace")
else: else:
file = open(file_or_filename, "w", encoding=encoding, file = open(file_or_filename, "w", encoding=encoding,
errors="xmlcharrefreplace") errors="xmlcharrefreplace")
with file: with file:
yield file.write yield file.write, file.encoding
else: else:
# file_or_filename is a file-like object # file_or_filename is a file-like object
# encoding determines if it is a text or binary writer # encoding determines if it is a text or binary writer
if encoding == "unicode": if encoding.lower() == "unicode":
# use a text writer as is # use a text writer as is
yield write yield write, getattr(file_or_filename, "encoding", None) or "utf-8"
else: else:
# wrap a binary writer with TextIOWrapper # wrap a binary writer with TextIOWrapper
with contextlib.ExitStack() as stack: with contextlib.ExitStack() as stack:
@ -805,7 +800,7 @@ def _get_writer(file_or_filename, encoding):
# Keep the original file open when the TextIOWrapper is # Keep the original file open when the TextIOWrapper is
# destroyed # destroyed
stack.callback(file.detach) stack.callback(file.detach)
yield file.write yield file.write, encoding
def _namespaces(elem, default_namespace=None): def _namespaces(elem, default_namespace=None):
# identify namespaces used in this tree # identify namespaces used in this tree

5
Misc/NEWS.d/next/Library/2022-04-25-10-23-01.gh-issue-91810.DOHa6B.rst Normal file
View File

@ -0,0 +1,5 @@
:class:`~xml.etree.ElementTree.ElementTree` method
:meth:`~xml.etree.ElementTree.ElementTree.write` and function
:func:`~xml.etree.ElementTree.tostring` now use the text file's encoding
("UTF-8" if not available) instead of locale encoding in XML declaration
when ``encoding="unicode"`` is specified.