def http_error_302(self, req, fp, code, msg, headers):
# Code from mechanize._urllib2_fork.HTTPRedirectHandler:
if 'location' in headers:
newurl = headers.getheaders('location')[0]
elif 'uri' in headers:
newurl = headers.getheaders('uri')[0]
else:
return
newurl = _rfc3986.clean_url(newurl, "latin-1")
newurl = _rfc3986.urljoin(req.get_full_url(), newurl)
new = self.redirect_request(req, fp, code, msg, headers, newurl)
if new is None:
return
if hasattr(req, 'redirect_dict'):
visited = new.redirect_dict = req.redirect_dict
if (visited.get(newurl, 0) >= self.max_repeats or
len(visited) >= self.max_redirections):
raise urllib.error.HTTPError(req.get_full_url(), code,
self.inf_msg + msg, headers, fp)
else:
visited = new.redirect_dict = req.redirect_dict = {}
visited[newurl] = visited.get(newurl, 0) + 1
fp.read()
fp.close()
# If the redirected URL doesn't match
new_url = new.get_full_url()
if not re.search('^http(?:s)?\:\/\/.*www\.linkedin\.com', new_url):
return _response.make_response('', headers.items(), new_url, 200, 'OK')
else:
return self.parent.open(new)
def _test_link_encoding(self, factory):
import mechanize
from mechanize._rfc3986 import clean_url
url = "http://example.com/"
for encoding in ["UTF-8", "latin-1"]:
encoding_decl = "; charset=%s" % encoding
b = TestBrowser(factory=factory)
r = MockResponse(url, """\
<a href="http://example.com/foo/bar——.html"
name="name0——">blah——</a>
""", #"
{"content-type": "text/html%s" % encoding_decl})
b.add_handler(make_mock_handler()([("http_open", r)]))
r = b.open(url)
Link = mechanize.Link
try:
mdashx2 = u"\u2014".encode(encoding)*2
except UnicodeError:
mdashx2 = '——'
qmdashx2 = clean_url(mdashx2, encoding)
# base_url, url, text, tag, attrs
exp = Link(url, "http://example.com/foo/bar%s.html" % qmdashx2,
"blah"+mdashx2, "a",
[("href", "http://example.com/foo/bar%s.html" % mdashx2),
("name", "name0%s" % mdashx2)])
# nr
link = b.find_link()
## print
## print exp
## print link
self.assertEqual(link, exp)
def test_link_encoding(self):
import mechanize
from mechanize._rfc3986 import clean_url
url = "http://example.com/"
for encoding in ["UTF-8", "latin-1"]:
encoding_decl = "; charset=%s" % encoding
b = TestBrowser()
r = MockResponse(url, """\
<a href="http://example.com/foo/bar——.html"
name="name0——">blah——</a>
""", {"content-type": "text/html%s" % encoding_decl})
b.add_handler(make_mock_handler()([("http_open", r)]))
r = b.open(url)
Link = mechanize.Link
mdashx2 = em_dash.encode('utf-8') * 2
qmdashx2 = clean_url(mdashx2, encoding)
# base_url, url, text, tag, attrs
exp = Link(
url, "http://example.com/foo/bar%s.html" % qmdashx2,
unicode_type("blah") + (em_dash * 2), "a",
[("href", unicode_type("http://example.com/foo/bar%s.html") % (
em_dash*2)),
("name", unicode_type("name0") + em_dash + em_dash)])
# nr
link = b.find_link()
# print
# print exp
# print link
self.assertEqual(link, exp)
def _test_link_encoding(self, factory):
import mechanize
from mechanize._rfc3986 import clean_url
url = "http://example.com/"
for encoding in ["UTF-8", "latin-1"]:
encoding_decl = "; charset=%s" % encoding
b = TestBrowser(factory=factory)
r = MockResponse(
url,
"""\
<a href="http://example.com/foo/bar——.html"
name="name0——">blah——</a>
""", #"
{"content-type": "text/html%s" % encoding_decl})
b.add_handler(make_mock_handler()([("http_open", r)]))
r = b.open(url)
Link = mechanize.Link
try:
mdashx2 = u"\u2014".encode(encoding) * 2
except UnicodeError:
mdashx2 = '——'
qmdashx2 = clean_url(mdashx2, encoding)
# base_url, url, text, tag, attrs
exp = Link(
url, "http://example.com/foo/bar%s.html" % qmdashx2,
"blah" + mdashx2, "a",
[("href", "http://example.com/foo/bar%s.html" % mdashx2),
("name", "name0%s" % mdashx2)])
# nr
link = b.find_link()
## print
## print exp
## print link
self.assertEqual(link, exp)