diff options
| author | Serhiy Storchaka <storchaka@gmail.com> | 2013-01-08 21:40:52 (GMT) |
|---|---|---|
| committer | Serhiy Storchaka <storchaka@gmail.com> | 2013-01-08 21:40:52 (GMT) |
| commit | ae3b32ad6bd6326e14b4b4316af6edea8dfc9173 (patch) | |
| tree | 6fb22559c8b1d83e2d2f9051a982b660613541f5 /Lib/test/test_codecs.py | |
| parent | 040c3c81c1d2fafe4bec9bec9e1e448c84ae7dd8 (diff) | |
| parent | 48e188e57313813bd048e25b8fa6123b8cd5c9a0 (diff) | |
| download | cpython-ae3b32ad6bd6326e14b4b4316af6edea8dfc9173.zip cpython-ae3b32ad6bd6326e14b4b4316af6edea8dfc9173.tar.gz cpython-ae3b32ad6bd6326e14b4b4316af6edea8dfc9173.tar.bz2 | |
Issue #11461: Fix the incremental UTF-16 decoder. Original patch by
Amaury Forgeot d'Arc. Added tests for partial decoding of non-BMP
characters.
Diffstat (limited to 'Lib/test/test_codecs.py')
| -rw-r--r-- | Lib/test/test_codecs.py | 48 |
1 files changed, 40 insertions, 8 deletions
diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py index f2a1ae3..15cd687 100644 --- a/Lib/test/test_codecs.py +++ b/Lib/test/test_codecs.py @@ -330,7 +330,7 @@ class UTF32Test(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", [ "", # first byte of BOM read "", # second byte of BOM read @@ -352,6 +352,10 @@ class UTF32Test(ReadTest): "\x00\xff\u0100", "\x00\xff\u0100", "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", ] ) @@ -386,7 +390,7 @@ class UTF32LETest(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", [ "", "", @@ -404,6 +408,10 @@ class UTF32LETest(ReadTest): "\x00\xff\u0100", "\x00\xff\u0100", "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", ] ) @@ -426,7 +434,7 @@ class UTF32BETest(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", [ "", "", @@ -444,6 +452,10 @@ class UTF32BETest(ReadTest): "\x00\xff\u0100", "\x00\xff\u0100", "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", ] ) @@ -494,7 +506,7 @@ class UTF16Test(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", [ "", # first byte of BOM read "", # second byte of BOM read => byteorder known @@ -506,6 +518,10 @@ class UTF16Test(ReadTest): "\x00\xff\u0100", "\x00\xff\u0100", "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", ] ) @@ -543,7 +559,7 @@ class UTF16LETest(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", [ "", "\x00", @@ -553,6 +569,10 @@ class UTF16LETest(ReadTest): "\x00\xff\u0100", "\x00\xff\u0100", "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", ] ) @@ -582,7 +602,7 @@ class UTF16BETest(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", [ "", "\x00", @@ -592,6 +612,10 @@ class UTF16BETest(ReadTest): "\x00\xff\u0100", "\x00\xff\u0100", "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff", + "\x00\xff\u0100\uffff\U00010000", ] ) @@ -621,7 +645,7 @@ class UTF8Test(ReadTest): def test_partial(self): self.check_partial( - "\x00\xff\u07ff\u0800\uffff", + "\x00\xff\u07ff\u0800\uffff\U00010000", [ "\x00", "\x00", @@ -634,6 +658,10 @@ class UTF8Test(ReadTest): "\x00\xff\u07ff\u0800", "\x00\xff\u07ff\u0800", "\x00\xff\u07ff\u0800\uffff", + "\x00\xff\u07ff\u0800\uffff", + "\x00\xff\u07ff\u0800\uffff", + "\x00\xff\u07ff\u0800\uffff", + "\x00\xff\u07ff\u0800\uffff\U00010000", ] ) @@ -816,7 +844,7 @@ class UTF8SigTest(ReadTest): def test_partial(self): self.check_partial( - "\ufeff\x00\xff\u07ff\u0800\uffff", + "\ufeff\x00\xff\u07ff\u0800\uffff\U00010000", [ "", "", @@ -835,6 +863,10 @@ class UTF8SigTest(ReadTest): "\ufeff\x00\xff\u07ff\u0800", "\ufeff\x00\xff\u07ff\u0800", "\ufeff\x00\xff\u07ff\u0800\uffff", + "\ufeff\x00\xff\u07ff\u0800\uffff", + "\ufeff\x00\xff\u07ff\u0800\uffff", + "\ufeff\x00\xff\u07ff\u0800\uffff", + "\ufeff\x00\xff\u07ff\u0800\uffff\U00010000", ] ) |