def test_iri_to_uri_is_sane_for_complex_url(self):
# test without the added complexity of unicode characters just
# to sanity check that a complex URL is not mangled
unicode_string = (u"http://*****:*****@www.host.com/with/path/and"
u"?query=string&is=sane")
encoded_string = url_util.iri_to_uri(unicode_string)
self.assertIsInstance(encoded_string, str)
self.assertEqual(unicode_string, encoded_string)
def test_iri_to_uri_encodes_to_utf8(self):
original_unicode = u"/\u00fe/\u00f4?\u00ef=\u2021"
# verify correct encoding
encoded_string = url_util.iri_to_uri(original_unicode)
self.assertIsInstance(encoded_string, str)
self.assertEqual("/%C3%BE/%C3%B4?%C3%AF=%E2%80%A1", encoded_string)
# verify that we can roundtrip back to the original unicode
decoded_string = urllib.unquote(encoded_string).decode("utf-8")
self.assertIsInstance(decoded_string, unicode)
self.assertEqual(original_unicode, decoded_string)
def redirect(self, uri, *args, **kwargs):
"""Override to handle locations with non-ASCII unicode characters.
A URI containing non-ASCII characters is known as an IRI, or
Internationalized Resource Identifier. This override conveniently
converts IRIs to URIs for use in the HTTP Location header.
See http://www.ietf.org/rfc/rfc3987.txt
webapp2.RequestHandler has redirect() and redirect_to(). Since
redirect_to() calls redirect(), this override handles both cases.
"""
uri = url_util.iri_to_uri(uri)
super(RequestHandler, self).redirect(uri, *args, **kwargs)
def test_iri_to_uri_is_idempotent(self):
unicode_string = u"/\u00fe/\u00f4?\u00ef=\u2021"
self.assertEqual(
url_util.iri_to_uri(unicode_string),
url_util.iri_to_uri(url_util.iri_to_uri(unicode_string)))
