def get_cache_timeout_response(self,
url,
http_timeout_in_seconds = 20,
max_cache_age_seconds = (1) * (24 * 60 * 60), # (number of days) * (number of seconds in a day),
header_addons = {}):
http_cached = httplib2.Http(self.CACHE_DIR, timeout=http_timeout_in_seconds)
header_dict = {'cache-control':'max-age='+str(max_cache_age_seconds)}
header_dict.update(header_addons)
cache_read = http_cached.cache.get(url)
if (cache_read):
(response, content) = cache_read.split("\r\n\r\n", 1)
else:
(response, content) = http_cached.request(url, headers=header_dict)
response['cache-control'] = "max-age=" + str(max_cache_age_seconds)
httplib2._updateCache(header_dict, response, content, http_cached.cache, url)
if response.fromcache:
self.status["count_got_response_from_cache"] += 1
else:
self.status["count_missed_cache"] += 1
self.status["count_cache_miss_details"] = str(self.status["count_cache_miss_details"]) + "; " + url
self.status["count_cache_miss_response"] = str(response)
self.status["count_api_requests"] += 1
if False:
self.status["count_request_exception"] = "EXCEPTION!"
self.status["count_uncached_call"] += 1
self.status["count_api_requests"] += 1
#(response, content) = http_cached.request(url, headers=header_dict.update({'cache-control':'no-cache'}))
req = urllib2.Request(url, headers=header_dict)
uh = urllib2.urlopen(req)
content = uh.read()
response = uh.info()
return(response, content)
def _follow_redirect(self, uri, method, body, headers, response, content, max_redirects):
""" Internal function to follow a redirect recieved by L{request} """
(scheme, authority, absolute_uri, defrag_uri) = httplib2.urlnorm(httplib2.iri2uri(uri))
if self.cache:
cachekey = defrag_uri
else:
cachekey = None
# Pick out the location header and basically start from the beginning
# remembering first to strip the ETag header and decrement our 'depth'
if not response.has_key('location') and response.status != 300:
raise httplib2.RedirectMissingLocation("Redirected but the response is missing a Location: header.", response, content)
# Fix-up relative redirects (which violate an RFC 2616 MUST)
if response.has_key('location'):
location = response['location']
(scheme, authority, path, query, fragment) = httplib2.parse_uri(location)
if authority == None:
response['location'] = httplib2.urlparse.urljoin(uri, location)
logging.debug('Relative redirect: changed [%s] to [%s]' % (location, response['location']))
if response.status == 301 and method in ["GET", "HEAD"]:
response['-x-permanent-redirect-url'] = response['location']
if not response.has_key('content-location'):
response['content-location'] = absolute_uri
httplib2._updateCache(headers, response, content, self.cache, cachekey)
headers.pop('if-none-match', None)
headers.pop('if-modified-since', None)
if response.has_key('location'):
location = response['location']
redirect_method = ((response.status == 303) and (method not in ["GET", "HEAD"])) and "GET" or method
return self.request(location, redirect_method, body=body, headers = headers, max_redirects = max_redirects - 1)
else:
raise httplib2.RedirectLimit("Redirected more times than redirection_limit allows.", response, content)
def _follow_redirect(self, uri, method, body, headers, response, content,
max_redirects):
"""Internal function to follow a redirect recieved by L{request}"""
(scheme, authority, absolute_uri,
defrag_uri) = httplib2.urlnorm(httplib2.iri2uri(uri))
if self.cache:
cachekey = defrag_uri
else:
cachekey = None
# Pick out the location header and basically start from the beginning
# remembering first to strip the ETag header and decrement our 'depth'
if "location" not in response and response.status != 300:
raise httplib2.RedirectMissingLocation(
"Redirected but the response is missing a Location: header.",
response, content)
# Fix-up relative redirects (which violate an RFC 2616 MUST)
if "location" in response:
location = response['location']
(scheme, authority, path, query,
fragment) = httplib2.parse_uri(location)
if authority is None:
response['location'] = httplib2.urlparse.urljoin(uri, location)
pywikibot.debug(
u"Relative redirect: changed [%s] to [%s]" %
(location, response['location']), _logger)
if response.status == 301 and method in ["GET", "HEAD"]:
response['-x-permanent-redirect-url'] = response['location']
if "content-location" not in response:
response['content-location'] = absolute_uri
httplib2._updateCache(headers, response, content, self.cache,
cachekey)
headers.pop('if-none-match', None)
headers.pop('if-modified-since', None)
if "location" in response:
location = response['location']
redirect_method = (
(response.status == 303) and
(method not in ["GET", "HEAD"])) and "GET" or method
return self.request(location,
redirect_method,
body=body,
headers=headers,
max_redirects=max_redirects - 1)
else:
raise httplib2.RedirectLimit(
"Redirected more times than redirection_limit allows.",
response, content)
def OverrideRequest(self, conn, host, absolute_uri, request_uri, method,
body, headers, redirections, cachekey):
"""Do the actual request using the connection object.
Also follow one level of redirects if necessary.
"""
auths = ([(auth.depth(request_uri), auth) for auth in self.authorizations
if auth.inscope(host, request_uri)])
auth = auths and sorted(auths)[0][1] or None
if auth:
auth.request(method, request_uri, headers, body)
(response, content) = self._conn_request(conn, request_uri, method, body,
headers)
if auth:
if auth.response(response, body):
auth.request(method, request_uri, headers, body)
(response, content) = self._conn_request(conn, request_uri, method,
body, headers)
response._stale_digest = 1
if response.status == 401:
for authorization in self._auth_from_challenge(
host, request_uri, headers, response, content):
authorization.request(method, request_uri, headers, body)
(response, content) = self._conn_request(conn, request_uri, method,
body, headers)
if response.status != 401:
self.authorizations.append(authorization)
authorization.response(response, body)
break
if (self.follow_all_redirects or (method in ["GET", "HEAD"])
or response.status == 303):
if self.follow_redirects and response.status in [300, 301, 302,
303, 307]:
# Pick out the location header and basically start from the beginning
# remembering first to strip the ETag header and decrement our 'depth'
if redirections:
if not response.has_key('location') and response.status != 300:
raise httplib2.RedirectMissingLocation(
"Redirected but the response is missing a Location: header.",
response, content)
# Fix-up relative redirects (which violate an RFC 2616 MUST)
if response.has_key('location'):
location = response['location']
(scheme, authority, path, query, fragment) = parse_uri(location)
if authority == None:
response['location'] = urlparse.urljoin(absolute_uri, location)
if response.status == 301 and method in ["GET", "HEAD"]:
response['-x-permanent-redirect-url'] = response['location']
if not response.has_key('content-location'):
response['content-location'] = absolute_uri
httplib2._updateCache(headers, response, content, self.cache,
cachekey)
if headers.has_key('if-none-match'):
del headers['if-none-match']
if headers.has_key('if-modified-since'):
del headers['if-modified-since']
if ('authorization' in headers and
not self.forward_authorization_headers):
del headers['authorization']
if response.has_key('location'):
location = response['location']
old_response = copy.deepcopy(response)
if not old_response.has_key('content-location'):
old_response['content-location'] = absolute_uri
redirect_method = method
if response.status in [302, 303]:
redirect_method = "GET"
body = None
(response, content) = self.request(
location, redirect_method, body=body, headers=headers,
redirections=redirections-1,
connection_type=conn.__class__)
response.previous = old_response
else:
raise httplib2.RedirectLimit(
"Redirected more times than redirection_limit allows.",
response, content)
elif response.status in [200, 203] and method in ["GET", "HEAD"]:
# Don't cache 206's since we aren't going to handle byte range
# requests
if not response.has_key('content-location'):
response['content-location'] = absolute_uri
httplib2._updateCache(headers, response, content, self.cache,
cachekey)
return (response, content)
def OverrideRequest(self, conn, host, absolute_uri, request_uri, method,
body, headers, redirections, cachekey):
"""Do the actual request using the connection object.
Also follow one level of redirects if necessary.
"""
auths = ([(auth.depth(request_uri), auth)
for auth in self.authorizations
if auth.inscope(host, request_uri)])
auth = auths and sorted(auths)[0][1] or None
if auth:
auth.request(method, request_uri, headers, body)
(response, content) = self._conn_request(conn, request_uri, method,
body, headers)
if auth:
if auth.response(response, body):
auth.request(method, request_uri, headers, body)
(response,
content) = self._conn_request(conn, request_uri, method, body,
headers)
response._stale_digest = 1
if response.status == 401:
for authorization in self._auth_from_challenge(
host, request_uri, headers, response, content):
authorization.request(method, request_uri, headers, body)
(response,
content) = self._conn_request(conn, request_uri, method, body,
headers)
if response.status != 401:
self.authorizations.append(authorization)
authorization.response(response, body)
break
if (self.follow_all_redirects or (method in ["GET", "HEAD"])
or response.status == 303):
if self.follow_redirects and response.status in [
300, 301, 302, 303, 307
]:
# Pick out the location header and basically start from the beginning
# remembering first to strip the ETag header and decrement our 'depth'
if redirections:
if not response.has_key(
'location') and response.status != 300:
raise httplib2.RedirectMissingLocation(
"Redirected but the response is missing a Location: header.",
response, content)
# Fix-up relative redirects (which violate an RFC 2616 MUST)
if response.has_key('location'):
location = response['location']
(scheme, authority, path, query,
fragment) = parse_uri(location)
if authority == None:
response['location'] = urlparse.urljoin(
absolute_uri, location)
if response.status == 301 and method in ["GET", "HEAD"]:
response['-x-permanent-redirect-url'] = response[
'location']
if not response.has_key('content-location'):
response['content-location'] = absolute_uri
httplib2._updateCache(headers, response, content,
self.cache, cachekey)
if headers.has_key('if-none-match'):
del headers['if-none-match']
if headers.has_key('if-modified-since'):
del headers['if-modified-since']
if ('authorization' in headers
and not self.forward_authorization_headers):
del headers['authorization']
if response.has_key('location'):
location = response['location']
old_response = copy.deepcopy(response)
if not old_response.has_key('content-location'):
old_response['content-location'] = absolute_uri
redirect_method = method
if response.status in [302, 303]:
redirect_method = "GET"
body = None
(response, content) = self.request(
location,
redirect_method,
body=body,
headers=headers,
redirections=redirections - 1,
connection_type=conn.__class__)
response.previous = old_response
else:
raise httplib2.RedirectLimit(
"Redirected more times than redirection_limit allows.",
response, content)
elif response.status in [200, 203] and method in ["GET", "HEAD"]:
# Don't cache 206's since we aren't going to handle byte range
# requests
if not response.has_key('content-location'):
response['content-location'] = absolute_uri
httplib2._updateCache(headers, response, content, self.cache,
cachekey)
return (response, content)
def request(self, uri, method="GET", body=None, headers=None, redirections=httplib2.DEFAULT_MAX_REDIRECTS, connection_type=None):
"""request handler with thread safety hacked in"""
try:
if headers is None:
headers = {}
else:
headers = httplib2._normalize_headers(headers)
if not headers.has_key('user-agent'):
headers['user-agent'] = "Python-httplib2/%s" % httplib2.__version__
uri = httplib2.iri2uri(uri)
(scheme, authority, request_uri, defrag_uri) = httplib2.urlnorm(uri)
domain_port = authority.split(":")[0:2]
if len(domain_port) == 2 and domain_port[1] == '443' and scheme == 'http':
scheme = 'https'
authority = domain_port[0]
conn_key = scheme+":"+authority
def get_conn(conn_key):
if conn_key in self.connections:
conn = self.connections[conn_key]
if type(conn) is list:
for c in conn:
if not getattr(c, 'busy', True):
return c
else: return c
if type(conn) is list:
return None
conn = get_conn(conn_key)
if conn is None:
if not connection_type:
connection_type = (scheme == 'https') and httplib2.HTTPSConnectionWithTimeout or httplib2.HTTPConnectionWithTimeout
certs = list(self.certificates.iter(authority))
if scheme == 'https' and certs:
conn = connection_type(authority, key_file=certs[0][0],
cert_file=certs[0][1], timeout=self.timeout, proxy_info=self.proxy_info)
self.connections.setdefault(conn_key, []).append(conn)
else:
conn = connection_type(authority, timeout=self.timeout, proxy_info=self.proxy_info)
self.connections.setdefault(conn_key, []).append(conn)
conn.set_debuglevel(httplib2.debuglevel)
conn.busy = True
if method in ["GET", "HEAD"] and 'range' not in headers and 'accept-encoding' not in headers:
headers['accept-encoding'] = 'deflate, gzip'
info = httplib2.email.Message.Message()
cached_value = None
if self.cache:
cachekey = defrag_uri
cached_value = self.cache.get(cachekey)
if cached_value:
try:
info, content = cached_value.split('\r\n\r\n', 1)
feedparser = httplib2.email.FeedParser.FeedParser()
feedparser.feed(info)
info = feedparser.close()
feedparser._parse = None
except IndexError:
self.cache.delete(cachekey)
cachekey = None
cached_value = None
else: cachekey = None
if method not in ["GET", "HEAD"] and self.cache and cachekey:
# RFC 2616 Section 13.10
self.cache.delete(cachekey)
if cached_value and method in ["GET", "HEAD"] and self.cache and 'range' not in headers:
if info.has_key('-x-permanent-redirect-url'):
(response, new_content) = self.request(info['-x-permanent-redirect-url'], "GET", headers = headers, redirections = redirections - 1)
response.previous = Response(info)
response.previous.fromcache = True
else:
entry_disposition = httplib2._entry_disposition(info, headers)
if entry_disposition == "FRESH":
if not cached_value:
info['status'] = '504'
content = ""
response = Response(info)
if cached_value:
response.fromcache = True
return (response, content)
if entry_disposition == "STALE":
if info.has_key('etag') and not self.ignore_etag and not 'if-none-match' in headers:
headers['if-none-match'] = info['etag']
if info.has_key('last-modified') and not 'last-modified' in headers:
headers['if-modified-since'] = info['last-modified']
elif entry_disposition == "TRANSPARENT": pass
(response, new_content) = self._request(conn, authority, uri, request_uri, method, body, headers, redirections, cachekey)
if response.status == 304 and method == "GET":
# Rewrite the cache entry with the new end-to-end headers
# Take all headers that are in response
# and overwrite their values in info.
# unless they are hop-by-hop, or are listed in the connection header.
for key in httplib2._get_end2end_headers(response):
info[key] = response[key]
merged_response = Response(info)
if hasattr(response, "_stale_digest"):
merged_response._stale_digest = response._stale_digest
httplib2._updateCache(headers, merged_response, content, self.cache, cachekey)
response = merged_response
response.status = 200
response.fromcache = True
elif response.status == 200:
content = new_content
else:
self.cache.delete(cachekey)
content = new_content
else:
cc = httplib2._parse_cache_control(headers)
if cc.has_key('only-if-cached'):
info['status'] = '504'
response = Response(info)
content = ""
else:
(response, content) = self._request(conn, authority, uri, request_uri, method, body, headers, redirections, cachekey)
except Exception, e:
if self.force_exception_to_status_code:
if isinstance(e, httplib2.HttpLib2ErrorWithResponse):
response = e.response
content = e.content
response.status = 500
response.reason = str(e)
elif isinstance(e, socket.timeout):
content = "Request Timeout"
response = Response( {
"content-type": "text/plain",
"status": "408",
"content-length": len(content)
})
response.reason = "Request Timeout"
else:
content = str(e)
response = Response( {
"content-type": "text/plain",
"status": "400",
"content-length": len(content)
})
response.reason = "Bad Request"
else: raise