Merged revisions 84655 via svnmerge from

svn+ssh://pythondev@svn.python.org/python/branches/py3k

........
  r84655 | antoine.pitrou | 2010-09-09 22:30:23 +0200 (jeu., 09 sept. 2010) | 6 lines

  Issue #9804: ascii() now always represents unicode surrogate pairs as
  a single `\UXXXXXXXX`, regardless of whether the character is printable
  or not.  Also, the "backslashreplace" error handler now joins surrogate
  pairs into a single character on UCS-2 builds.
........
This commit is contained in:
Antoine Pitrou 2010-09-09 20:33:43 +00:00
parent 8e0bb6a1e2
commit c9a8df24cc
4 changed files with 72 additions and 17 deletions

View file

@ -174,6 +174,28 @@ class BuiltinTest(unittest.TestCase):
a = {}
a[0] = a
self.assertEqual(ascii(a), '{0: {...}}')
# Advanced checks for unicode strings
def _check_uni(s):
self.assertEqual(ascii(s), repr(s))
_check_uni("'")
_check_uni('"')
_check_uni('"\'')
_check_uni('\0')
_check_uni('\r\n\t .')
# Unprintable non-ASCII characters
_check_uni('\x85')
_check_uni('\u1fff')
_check_uni('\U00012fff')
# Lone surrogates
_check_uni('\ud800')
_check_uni('\udfff')
# Issue #9804: surrogates should be joined even for printable
# wide characters (UCS-2 builds).
self.assertEqual(ascii('\U0001d121'), "'\\U0001d121'")
# All together
s = "'\0\"\n\r\t abcd\x85é\U00012fff\uD800\U0001D121xxx."
self.assertEqual(ascii(s),
r"""'\'\x00"\n\r\t abcd\x85\xe9\U00012fff\ud800\U0001d121xxx.'""")
def test_neg(self):
x = -sys.maxsize-1