def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# Use the correct exception
tests = [
("\u3042", "\\u3042"),
("\n", "\\x0a"),
("a", "\\x61"),
("\x00", "\\x00"),
("\xff", "\\xff"),
("\u0100", "\\u0100"),
("\uffff", "\\uffff"),
("\U00010000", "\\U00010000"),
("\U0010ffff", "\\U0010ffff"),
# Lone surrogates
("\ud800", "\\ud800"),
("\udfff", "\\udfff"),
("\ud800\udfff", "\\ud800\\udfff"),
]
for s, r in tests:
with self.subTest(str=s):
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "a" + s + "b",
1, 1 + len(s), "ouch")),
(r, 1 + len(s))
)
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeTranslateError("a" + s + "b",
1, 1 + len(s), "ouch")),
(r, 1 + len(s))
)
tests = [
(b"a", "\\x61"),
(b"\n", "\\x0a"),
(b"\x00", "\\x00"),
(b"\xff", "\\xff"),
]
for b, r in tests:
with self.subTest(bytes=b):
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeDecodeError("ascii", bytearray(b"a" + b + b"b"),
1, 2, "ouch")),
(r, 2)
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# Use the correct exception
tests = [
("\u3042", "\\u3042"),
("\n", "\\x0a"),
("a", "\\x61"),
("\x00", "\\x00"),
("\xff", "\\xff"),
("\u0100", "\\u0100"),
("\uffff", "\\uffff"),
("\U00010000", "\\U00010000"),
("\U0010ffff", "\\U0010ffff"),
# Lone surrogates
("\ud800", "\\ud800"),
("\udfff", "\\udfff"),
("\ud800\udfff", "\\ud800\\udfff"),
]
for s, r in tests:
with self.subTest(str=s):
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "a" + s + "b",
1, 1 + len(s), "ouch")),
(r, 1 + len(s))
)
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeTranslateError("a" + s + "b",
1, 1 + len(s), "ouch")),
(r, 1 + len(s))
)
tests = [
(b"a", "\\x61"),
(b"\n", "\\x0a"),
(b"\x00", "\\x00"),
(b"\xff", "\\xff"),
]
for b, r in tests:
with self.subTest(bytes=b):
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeDecodeError("ascii", bytearray(b"a" + b + b"b"),
1, 2, "ouch")),
(r, 2)
)
def backslashescape_errors(exception):
if isinstance(exception, UnicodeDecodeError):
bad_data = exception.object[exception.start:exception.end]
escaped = "".join(r"\x%02x" % x for x in bad_data)
return escaped, exception.end
return codecs.backslashreplace_errors(exception)
def escape_invis(decode_error):
decode_error.end = decode_error.start + 1
if CONTROL_CHARS.match(
decode_error.object[decode_error.start:decode_error.end]):
return codecs.backslashreplace_errors(decode_error)
return decode_error.object[decode_error.start:decode_error.end].encode(
'utf-8'), decode_error.end
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(TypeError, codecs.backslashreplace_errors, 42)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(TypeError, codecs.backslashreplace_errors,
UnicodeError("ouch"))
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError, codecs.backslashreplace_errors,
UnicodeDecodeError("ascii", bytearray(b"\xff"), 0, 1, "ouch"))
self.assertRaises(TypeError, codecs.backslashreplace_errors,
UnicodeTranslateError("\u3042", 0, 1, "ouch"))
# Use the correct exception
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\u3042", 0, 1, "ouch")),
("\\u3042", 1))
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\x00", 0, 1, "ouch")),
("\\x00", 1))
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\xff", 0, 1, "ouch")),
("\\xff", 1))
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\u0100", 0, 1, "ouch")),
("\\u0100", 1))
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\uffff", 0, 1, "ouch")),
("\\uffff", 1))
# 1 on UCS-4 builds, 2 on UCS-2
len_wide = len("\U00010000")
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\U00010000", 0, len_wide,
"ouch")), ("\\U00010000", len_wide))
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\U0010ffff", 0, len_wide,
"ouch")), ("\\U0010ffff", len_wide))
# Lone surrogates (regardless of unicode width)
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\ud800", 0, 1, "ouch")),
("\\ud800", 1))
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", "\udfff", 0, 1, "ouch")),
("\\udfff", 1))
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(TypeError, codecs.backslashreplace_errors, 42)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(TypeError, codecs.backslashreplace_errors, UnicodeError("ouch"))
# "backslashreplace" can only be used for encoding
self.assertRaises(TypeError, codecs.backslashreplace_errors, UnicodeDecodeError("ascii", "\xff", 0, 1, "ouch"))
self.assertRaises(TypeError, codecs.backslashreplace_errors, UnicodeTranslateError(u"\u3042", 0, 1, "ouch"))
# Use the correct exception
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u3042", 0, 1, "ouch")), (u"\\u3042", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\x00", 0, 1, "ouch")), (u"\\x00", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\xff", 0, 1, "ouch")), (u"\\xff", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u0100", 0, 1, "ouch")), (u"\\u0100", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\uffff", 0, 1, "ouch")), (u"\\uffff", 1)
)
if sys.maxunicode > 0xFFFF:
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U00010000", 0, 1, "ouch")),
(u"\\U00010000", 1),
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U0010ffff", 0, 1, "ouch")),
(u"\\U0010ffff", 1),
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeDecodeError("ascii", "\xff", 0, 1, "ouch")
)
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeTranslateError(u"\u3042", 0, 1, "ouch")
)
# Use the correct exception
tests = [
(u"\u3042", u"\\u3042"),
(u"\n", u"\\x0a"),
(u"a", u"\\x61"),
(u"\x00", u"\\x00"),
(u"\xff", u"\\xff"),
(u"\u0100", u"\\u0100"),
(u"\uffff", u"\\uffff"),
# Lone surrogates
(u"\ud800", u"\\ud800"),
(u"\udfff", u"\\udfff"),
]
if sys.maxunicode > 0xffff:
tests += [
(u"\U00010000", u"\\U00010000"),
(u"\U0010ffff", u"\\U0010ffff"),
]
else:
tests += [
(u"\U00010000", u"\\ud800\\udc00"),
(u"\U0010ffff", u"\\udbff\\udfff"),
]
for s, r in tests:
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", u"a" + s + u"b",
1, 1 + len(s), "ouch")),
(r, 1 + len(s))
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeDecodeError("ascii", "\xff", 0, 1, "ouch")
)
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeTranslateError(u"\u3042", 0, 1, "ouch")
)
# Use the correct exception
tests = [
(u"\u3042", u"\\u3042"),
(u"\n", u"\\x0a"),
(u"a", u"\\x61"),
(u"\x00", u"\\x00"),
(u"\xff", u"\\xff"),
(u"\u0100", u"\\u0100"),
(u"\uffff", u"\\uffff"),
# Lone surrogates
(u"\ud800", u"\\ud800"),
(u"\udfff", u"\\udfff"),
]
if sys.maxunicode > 0xffff:
tests += [
(u"\U00010000", u"\\U00010000"),
(u"\U0010ffff", u"\\U0010ffff"),
]
else:
tests += [
(u"\U00010000", u"\\ud800\\udc00"),
(u"\U0010ffff", u"\\udbff\\udfff"),
]
for s, r in tests:
self.assertEqual(
codecs.backslashreplace_errors(
UnicodeEncodeError("ascii", u"a" + s + u"b",
1, 1 + len(s), "ouch")),
(r, 1 + len(s))
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeDecodeError("ascii", "\xff", 0, 1, "ouch")
)
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeTranslateError(u"\u3042", 0, 1, "ouch")
)
if test_support.due_to_ironpython_bug("http://tkbgitvstfat01:8080/WorkItemTracking/WorkItem.aspx?artifactMoniker=303935"):
return
# Use the correct exception
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u3042", 0, 1, "ouch")),
(u"\\u3042", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\x00", 0, 1, "ouch")),
(u"\\x00", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\xff", 0, 1, "ouch")),
(u"\\xff", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u0100", 0, 1, "ouch")),
(u"\\u0100", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\uffff", 0, 1, "ouch")),
(u"\\uffff", 1)
)
if sys.maxunicode>0xffff:
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U00010000", 0, 1, "ouch")),
(u"\\U00010000", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U0010ffff", 0, 1, "ouch")),
(u"\\U0010ffff", 1)
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeDecodeError("ascii", "\xff", 0, 1, "ouch")
)
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeTranslateError(u"\u3042", 0, 1, "ouch")
)
if test_support.due_to_ironpython_bug("http://tkbgitvstfat01:8080/WorkItemTracking/WorkItem.aspx?artifactMoniker=303935"):
return
# Use the correct exception
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u3042", 0, 1, "ouch")),
(u"\\u3042", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\x00", 0, 1, "ouch")),
(u"\\x00", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\xff", 0, 1, "ouch")),
(u"\\xff", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u0100", 0, 1, "ouch")),
(u"\\u0100", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\uffff", 0, 1, "ouch")),
(u"\\uffff", 1)
)
if sys.maxunicode>0xffff:
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U00010000", 0, 1, "ouch")),
(u"\\U00010000", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U0010ffff", 0, 1, "ouch")),
(u"\\U0010ffff", 1)
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
42
)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeError("ouch")
)
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeDecodeError("ascii", "\xff", 0, 1, "ouch")
)
self.assertRaises(
TypeError,
codecs.backslashreplace_errors,
UnicodeTranslateError(u"\u3042", 0, 1, "ouch")
)
# Use the correct exception
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u3042", 0, 1, "ouch")),
(u"\\u3042", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\x00", 0, 1, "ouch")),
(u"\\x00", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\xff", 0, 1, "ouch")),
(u"\\xff", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\u0100", 0, 1, "ouch")),
(u"\\u0100", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\uffff", 0, 1, "ouch")),
(u"\\uffff", 1)
)
if sys.maxunicode>0xffff:
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U00010000", 0, 1, "ouch")),
(u"\\U00010000", 1)
)
self.assertEquals(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", u"\U0010ffff", 0, 1, "ouch")),
(u"\\U0010ffff", 1)
)
def test_badandgoodbackslashreplaceexceptions(self):
# "backslashreplace" complains about a non-exception passed in
self.assertRaises(TypeError, codecs.backslashreplace_errors, 42)
# "backslashreplace" complains about the wrong exception types
self.assertRaises(TypeError, codecs.backslashreplace_errors, UnicodeError("ouch"))
# "backslashreplace" can only be used for encoding
self.assertRaises(
TypeError, codecs.backslashreplace_errors, UnicodeDecodeError("ascii", bytearray(b"\xff"), 0, 1, "ouch")
)
self.assertRaises(TypeError, codecs.backslashreplace_errors, UnicodeTranslateError("\u3042", 0, 1, "ouch"))
# Use the correct exception
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\u3042", 0, 1, "ouch")), ("\\u3042", 1)
)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\x00", 0, 1, "ouch")), ("\\x00", 1)
)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\xff", 0, 1, "ouch")), ("\\xff", 1)
)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\u0100", 0, 1, "ouch")), ("\\u0100", 1)
)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\uffff", 0, 1, "ouch")), ("\\uffff", 1)
)
# 1 on UCS-4 builds, 2 on UCS-2
len_wide = len("\U00010000")
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\U00010000", 0, len_wide, "ouch")),
("\\U00010000", len_wide),
)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\U0010ffff", 0, len_wide, "ouch")),
("\\U0010ffff", len_wide),
)
# Lone surrogates (regardless of unicode width)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\ud800", 0, 1, "ouch")), ("\\ud800", 1)
)
self.assertEqual(
codecs.backslashreplace_errors(UnicodeEncodeError("ascii", "\udfff", 0, 1, "ouch")), ("\\udfff", 1)
)