Use more obvious non-ascii characters in unicode test

IPython/lib/tests/test_security.py

@@ -1,26 +1,26 b''
 # coding: utf-8
 from IPython.lib import passwd
 from IPython.lib.security import passwd_check, salt_len
 import nose.tools as nt
 
 def test_passwd_structure():
     p = passwd('passphrase')
     algorithm, salt, hashed = p.split(':')
     nt.assert_equal(algorithm, 'sha1')
     nt.assert_equal(len(salt), salt_len)
     nt.assert_equal(len(hashed), 40)
 
 def test_roundtrip():
     p = passwd('passphrase')
     nt.assert_equal(passwd_check(p, 'passphrase'), True)
 
 def test_bad():
     p = passwd('passphrase')
     nt.assert_equal(passwd_check(p, p), False)
     nt.assert_equal(passwd_check(p, 'a:b:c:d'), False)
     nt.assert_equal(passwd_check(p, 'a:b'), False)
 
 def test_passwd_check_unicode():
     # GH issue #4524
-    phash = u'sha1:9dc18846ca26:6bb62badc41fde529c258a8a7fbe259a91313df8'
-    assert passwd_check(phash, u'mypassword³')
  No newline at end of file
+    phash = u'sha1:23862bc21dd3:7a415a95ae4580582e314072143d9c382c491e4f'
+    assert passwd_check(phash, u"łe¶ŧ←↓→")
  No newline at end of file