diff options
author | Antoine Pitrou <solipsis@pitrou.net> | 2012-11-17 20:14:58 (GMT) |
---|---|---|
committer | Antoine Pitrou <solipsis@pitrou.net> | 2012-11-17 20:14:58 (GMT) |
commit | e3ae321222c32d64232c46e4062082effec79bcf (patch) | |
tree | 1afcaf1354d5d6e7ddb9d3fb3a534b43c37f1769 /Lib/test/test_codecs.py | |
parent | 1df43d33d2f17584639759519e290021b063485c (diff) | |
download | cpython-e3ae321222c32d64232c46e4062082effec79bcf.zip cpython-e3ae321222c32d64232c46e4062082effec79bcf.tar.gz cpython-e3ae321222c32d64232c46e4062082effec79bcf.tar.bz2 |
Issue #15379: Fix passing of non-BMP characters as integers for the charmap decoder (already working as unicode strings).
Patch by Serhiy Storchaka.
Diffstat (limited to 'Lib/test/test_codecs.py')
-rw-r--r-- | Lib/test/test_codecs.py | 105 |
1 files changed, 105 insertions, 0 deletions
diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py index 57669f8..f620b4c 100644 --- a/Lib/test/test_codecs.py +++ b/Lib/test/test_codecs.py @@ -1534,6 +1534,10 @@ class CharmapTest(unittest.TestCase): (u"ab", 3) ) + self.assertRaises(UnicodeDecodeError, + codecs.charmap_decode, b"\x00\x01\x02", "strict", u"ab" + ) + self.assertEqual( codecs.charmap_decode("\x00\x01\x02", "ignore", u"ab\ufffe"), (u"ab", 3) @@ -1545,6 +1549,107 @@ class CharmapTest(unittest.TestCase): (u"", len(allbytes)) ) + def test_decode_with_int2str_map(self): + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "strict", + {0: u'a', 1: u'b', 2: u'c'}), + (u"abc", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "strict", + {0: u'Aa', 1: u'Bb', 2: u'Cc'}), + (u"AaBbCc", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "strict", + {0: u'\U0010FFFF', 1: u'b', 2: u'c'}), + (u"\U0010FFFFbc", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "strict", + {0: u'a', 1: u'b', 2: u''}), + (u"ab", 3) + ) + + self.assertRaises(UnicodeDecodeError, + codecs.charmap_decode, "\x00\x01\x02", "strict", + {0: u'a', 1: u'b'} + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "replace", + {0: u'a', 1: u'b'}), + (u"ab\ufffd", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "replace", + {0: u'a', 1: u'b', 2: None}), + (u"ab\ufffd", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "ignore", + {0: u'a', 1: u'b'}), + (u"ab", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "ignore", + {0: u'a', 1: u'b', 2: None}), + (u"ab", 3) + ) + + allbytes = bytes(range(256)) + self.assertEqual( + codecs.charmap_decode(allbytes, "ignore", {}), + (u"", len(allbytes)) + ) + + def test_decode_with_int2int_map(self): + a = ord(u'a') + b = ord(u'b') + c = ord(u'c') + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "strict", + {0: a, 1: b, 2: c}), + (u"abc", 3) + ) + + # Issue #15379 + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "strict", + {0: 0x10FFFF, 1: b, 2: c}), + (u"\U0010FFFFbc", 3) + ) + + self.assertRaises(TypeError, + codecs.charmap_decode, "\x00\x01\x02", "strict", + {0: 0x110000, 1: b, 2: c} + ) + + self.assertRaises(UnicodeDecodeError, + codecs.charmap_decode, "\x00\x01\x02", "strict", + {0: a, 1: b}, + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "replace", + {0: a, 1: b}), + (u"ab\ufffd", 3) + ) + + self.assertEqual( + codecs.charmap_decode("\x00\x01\x02", "ignore", + {0: a, 1: b}), + (u"ab", 3) + ) + + class WithStmtTest(unittest.TestCase): def test_encodedfile(self): f = StringIO.StringIO("\xc3\xbc") |