def datagramReceived(self, data, addr, outip):
if outip not in self.interfaces:
if self.INADDR_ANY not in self.interfaces:
return
req_line, data = data.split('\r\n', 1)
method, path, version = req_line.split(None, 3)
# check method
if method != 'M-SEARCH' or path != '*':
return
# parse header
headers = HTTPMessage(StringIO(data))
mx = int(headers.getheader('MX'))
# send M-SEARCH response
for udn in self.devices:
device = self.devices[udn]
delay = random() * mx
for packet in device.make_msearch_response(headers,
(outip, self.port),
addr):
buff = build_packet('HTTP/1.1 200 OK', packet)
self.reactor.callLater(delay, self._send_packet, self.ssdp,
buff, addr)
delay += self.SSDP_INTERVAL
def __init__(self, replay_response, method=None):
self.reason = replay_response['status']['message']
self.status = replay_response['status']['code']
self.version = None
if 'body_text' in replay_response:
# JSON decoder returns unicode, not str, so this needs to be
# encoded to properly reproduce content off the wire.
self._content = replay_response['body_text'].encode('utf8')
elif 'body_quoted_printable' in replay_response:
# quopri.decodestring returns str, which is correct for content off
# the wire.
self._content = quopri.decodestring(replay_response['body_quoted_printable'])
else:
# .decode('base64') returns str, which is correct for content off
# the wire.
self._content = replay_response['body'].decode('base64')
self.fp = StringIO(self._content)
msg_fp = StringIO('\r\n'.join('{}: {}'.format(h, v)
for h, v in replay_response['headers'].iteritems()))
self.msg = HTTPMessage(msg_fp)
self.msg.fp = None # httplib does this, okay?
length = self.msg.getheader('content-length')
self.length = int(length) if length else None
# Save method to handle HEAD specially as httplib does
self._method = method
class VCRHTTPResponse(object):
"""
Stub reponse class that gets returned instead of a HTTPResponse
"""
def __init__(self, recorded_response):
self.recorded_response = recorded_response
self.reason = recorded_response['status']['message']
self.status = recorded_response['status']['code']
self.version = None
self._content = StringIO(self.recorded_response['body']['string'])
self.msg = HTTPMessage(StringIO(''))
for k, v in self.recorded_response['headers'].iteritems():
self.msg.addheader(k, v)
self.length = self.msg.getheader('content-length') or None
def read(self, chunked=False):
# Note: I'm pretty much ignoring any chunking stuff because
# I don't really understand what it is or how it works.
return self._content.read()
def isclosed(self):
# Urllib3 seems to call this because it actually uses
# the weird chunking support in httplib
return True
def getheaders(self):
return self.recorded_response['headers'].iteritems()
def update(self, lock):
""" See if a http or rss page changed. """
self.error = False
self.specto.mark_watch_busy(True, self.id)
self.specto.logger.log(
_("Updating watch: \"%s\"") % self.name, "info", self.__class__)
# Create a unique name for each url.
digest = md5.new(self.url_).digest()
cacheFileName = "".join(["%02x" % (ord(c), ) for c in digest])
self.cacheFullPath_ = os.path.join(cacheSubDir__, cacheFileName)
request = urllib2.Request(self.url_, None, {"Accept-encoding": "gzip"})
if (self.cached == 1) or (os.path.exists(self.cacheFullPath_)):
self.cached = 1
f = file(self.cacheFullPath_, "r") # Load up the cached version
self.infoB_ = HTTPMessage(f)
if self.infoB_.has_key('last-modified'):
request.add_header("If-Modified-Since",
self.infoB_['last-modified'])
if self.infoB_.has_key('ETag'):
request.add_header("If-None-Match", self.infoB_['ETag'])
try:
response = urllib2.urlopen(request)
except (urllib2.URLError, BadStatusLine), e:
self.error = True
self.specto.logger.log(
_("Watch: \"%s\" has error: ") % self.name + str(e), "error",
self.__class__)
def parse_headers(header_list):
if isinstance(header_list, dict):
return parse_headers_backwards_compat(header_list)
headers = "".join(header_list) + "\r\n"
msg = HTTPMessage(StringIO(headers))
msg.fp.seek(0)
msg.readheaders()
return msg
def parse_headers(header_list):
if isinstance(header_list, dict):
return parse_headers_backwards_compat(header_list)
headers = "".join(header_list) + "\r\n"
msg = HTTPMessage(StringIO(headers))
msg.fp.seek(0)
msg.readheaders()
return msg
def __init__(self, data):
BytesIO.__init__(self, data)
self.status = 200
self.version = 'HTTP/1.1'
self.reason = 'OK'
if PY2:
self.msg = HTTPMessage(BytesIO(b'Content-Type: application/x-compressed\r\n'))
else:
self.msg = HTTPMessage()
self.msg.add_header('Content-Type', 'application/x-compressed')
def __init__(self, data):
BytesIO.__init__(self, data)
self.status = 200
self.version = 'HTTP/1.1'
self.reason = 'OK'
if PY2:
self.msg = HTTPMessage(BytesIO('Content-Type: application/x-compressed\r\n'))
else:
self.msg = HTTPMessage()
self.msg.add_header('Content-Type', 'application/x-compressed')
def parse_headers_backwards_compat(header_dict):
"""
In vcr 0.6.0, I changed the cassettes to store
headers as a list instead of a dict. This method
parses the old dictionary-style headers for
backwards-compatability reasons.
"""
msg = HTTPMessage(StringIO(""))
for key, val in header_dict.iteritems():
msg.addheader(key, val)
msg.headers.append("{0}:{1}".format(key, val))
return msg
def parse_headers_backwards_compat(header_dict):
"""
In vcr 0.6.0, I changed the cassettes to store
headers as a list instead of a dict. This method
parses the old dictionary-style headers for
backwards-compatability reasons.
"""
msg = HTTPMessage(StringIO(""))
for key, val in header_dict.iteritems():
msg.addheader(key, val)
msg.headers.append("{0}:{1}".format(key, val))
return msg
class VCRHTTPResponse(object):
def __init__(self, recorded_response):
self.recorded_response = recorded_response
self.reason = recorded_response["status"]["message"]
self.status = recorded_response["status"]["code"]
self._content = StringIO(self.recorded_response["body"]["string"])
self.msg = HTTPMessage(StringIO(""))
for k, v in self.recorded_response["headers"].iteritems():
self.msg.addheader(k, v)
def read(self, chunked=False):
return self._content.read()
def __init__(self, content):
if PY2x:
fp = StringIO(content)
self._info = HTTPMessage(fp)
else:
self._info = HTTPMessage()
# Adjust to testdata.
la = content.split(':')
if len(la) > 1:
# Get the type by just
# using the data at the end.
t = la[-1].strip()
self._info.set_type(t)
def __init__(self, data):
req_line, headers = data.split('\r\n', 1)
# HTTPMessage has no proper __repr__, so let's use the dictionary
dict = HTTPMessage(StringIO(headers)).dict.copy()
# all header names in the UPnP specs are uppercase
self.headers = {k.upper(): v for k, v in dict.items()}
method = req_line.split(' ')[0]
self._notify = method == 'NOTIFY'
# Unique Service Name => Unique Device Name + Type
usn = self.headers.get('USN')
self._udn, self._type = split_usn(usn) if usn else (None, None)
def update(self, lock):
""" See if a http or rss page changed. """
self.error = False
self.specto.mark_watch_busy(True, self.id)
self.specto.logger.log(_("Updating watch: \"%s\"") % self.name, "info", self.__class__)
# Create a unique name for each url.
digest = md5.new(self.url_).digest()
cacheFileName = "".join(["%02x" % (ord(c),) for c in digest])
self.cacheFullPath_ = os.path.join(cacheSubDir__, cacheFileName)
request = urllib2.Request(self.url_, None, {"Accept-encoding" : "gzip"})
if (self.cached == 1) or (os.path.exists(self.cacheFullPath_)):
self.cached = 1
f = file(self.cacheFullPath_, "r")# Load up the cached version
self.infoB_ = HTTPMessage(f)
if self.infoB_.has_key('last-modified'):
request.add_header("If-Modified-Since", self.infoB_['last-modified'])
if self.infoB_.has_key('ETag'):
request.add_header("If-None-Match", self.infoB_['ETag'])
try:
response = urllib2.urlopen(request)
except (urllib2.URLError, BadStatusLine), e:
self.error = True
self.specto.logger.log(_("Watch: \"%s\" has error: ") % self.name + str(e), "error", self.__class__)
class MockHttpLibResponse(BytesIO):
def __init__(self, data):
BytesIO.__init__(self, data)
self.status = 200
self.version = 'HTTP/1.1'
self.reason = 'OK'
if PY2:
self.msg = HTTPMessage(BytesIO('Content-Type: application/x-compressed\r\n'))
else:
self.msg = HTTPMessage()
self.msg.add_header('Content-Type', 'application/x-compressed')
def getheaders(self):
return list(self.msg.items())
def isclosed(self):
return self.closed
class MockHttpLibResponse(BytesIO):
def __init__(self, data):
BytesIO.__init__(self, data)
self.status = 200
self.version = 'HTTP/1.1'
self.reason = 'OK'
if PY2:
self.msg = HTTPMessage(BytesIO(b'Content-Type: application/x-compressed\r\n'))
else:
self.msg = HTTPMessage()
self.msg.add_header('Content-Type', 'application/x-compressed')
def getheaders(self):
return list(self.msg.items())
def isclosed(self):
return self.closed
def __init__(self, recorded_response):
self.recorded_response = recorded_response
self.reason = recorded_response["status"]["message"]
self.status = recorded_response["status"]["code"]
self._content = StringIO(self.recorded_response["body"]["string"])
self.msg = HTTPMessage(StringIO(""))
for k, v in self.recorded_response["headers"].iteritems():
self.msg.addheader(k, v)
def __init__(self, recorded_response):
self.recorded_response = recorded_response
self.reason = recorded_response['status']['message']
self.status = recorded_response['status']['code']
self.version = None
self._content = StringIO(self.recorded_response['body']['string'])
# We are skipping the header parsing (they have already been parsed
# at this point) and directly adding the headers to the header
# container, so just pass an empty StringIO.
self.msg = HTTPMessage(StringIO(''))
for key, val in self.recorded_response['headers'].iteritems():
self.msg.addheader(key, val)
# msg.addheaders adds the headers to msg.dict, but not to
# the msg.headers list representation of headers, so
# I have to add it to both.
self.msg.headers.append("{0}:{1}".format(key, val))
self.length = self.msg.getheader('content-length') or None
def test_from_httplib(self):
if six.PY3:
raise SkipTest()
from httplib import HTTPMessage
from StringIO import StringIO
msg = """
Server: nginx
Content-Type: text/html; charset=windows-1251
Connection: keep-alive
Set-Cookie: bb_lastvisit=1348253375; expires=Sat, 21-Sep-2013 18:49:35 GMT; path=/
Set-Cookie: bb_lastactivity=0; expires=Sat, 21-Sep-2013 18:49:35 GMT; path=/
"""
msg = HTTPMessage(StringIO(msg.lstrip().replace('\n', '\r\n')))
d = HTTPHeaderDict.from_httplib(msg)
self.assertEqual(d['server'], 'nginx')
cookies = d.getlist('set-cookie')
self.assertEqual(len(cookies), 2)
self.assertTrue(cookies[0].startswith("bb_lastvisit"))
self.assertTrue(cookies[1].startswith("bb_lastactivity"))
def test_from_httplib(self):
if six.PY3:
raise SkipTest()
from httplib import HTTPMessage
from StringIO import StringIO
msg = """
Server: nginx
Content-Type: text/html; charset=windows-1251
Connection: keep-alive
Set-Cookie: bb_lastvisit=1348253375; expires=Sat, 21-Sep-2013 18:49:35 GMT; path=/
Set-Cookie: bb_lastactivity=0; expires=Sat, 21-Sep-2013 18:49:35 GMT; path=/
"""
msg = HTTPMessage(StringIO(msg.lstrip().replace('\n', '\r\n')))
d = HTTPHeaderDict.from_httplib(msg)
self.assertEqual(d['server'], 'nginx')
cookies = d.getlist('set-cookie')
self.assertEqual(len(cookies), 2)
self.assertTrue(cookies[0].startswith("bb_lastvisit"))
self.assertTrue(cookies[1].startswith("bb_lastactivity"))
def __init__(self, recorded_response):
self.recorded_response = recorded_response
self.reason = recorded_response['status']['message']
self.status = recorded_response['status']['code']
self.version = None
self._content = StringIO(self.recorded_response['body']['string'])
self.msg = HTTPMessage(StringIO(''))
for k, v in self.recorded_response['headers'].iteritems():
self.msg.addheader(k, v)
self.length = self.msg.getheader('content-length') or None
class VCRHTTPResponse(object):
"""
Stub reponse class that gets returned instead of a HTTPResponse
"""
def __init__(self, recorded_response):
self.recorded_response = recorded_response
self.reason = recorded_response['status']['message']
self.status = recorded_response['status']['code']
self.version = None
self._content = StringIO(self.recorded_response['body']['string'])
# We are skipping the header parsing (they have already been parsed
# at this point) and directly adding the headers to the header
# container, so just pass an empty StringIO.
self.msg = HTTPMessage(StringIO(''))
for key, val in self.recorded_response['headers'].iteritems():
self.msg.addheader(key, val)
# msg.addheaders adds the headers to msg.dict, but not to
# the msg.headers list representation of headers, so
# I have to add it to both.
self.msg.headers.append("{0}:{1}".format(key, val))
self.length = self.msg.getheader('content-length') or None
def read(self, *args, **kwargs):
# Note: I'm pretty much ignoring any chunking stuff because
# I don't really understand what it is or how it works.
return self._content.read(*args, **kwargs)
def close(self):
return True
def isclosed(self):
# Urllib3 seems to call this because it actually uses
# the weird chunking support in httplib
return True
def getheaders(self):
return self.recorded_response['headers'].iteritems()
def __init__(self, replay_response, method=None):
self.reason = replay_response['status']['message']
self.status = replay_response['status']['code']
self.version = None
if 'body_quoted_printable' in replay_response:
self._content = quopri.decodestring(
replay_response['body_quoted_printable'])
else:
self._content = replay_response['body'].decode('base64')
self.fp = StringIO(self._content)
msg_fp = StringIO('\r\n'.join(
'{}: {}'.format(h, v)
for h, v in replay_response['headers'].iteritems()))
self.msg = HTTPMessage(msg_fp)
self.msg.fp = None # httplib does this, okay?
length = self.msg.getheader('content-length')
self.length = int(length) if length else None
# Save method to handle HEAD specially as httplib does
self._method = method
def __init__(self, conn):
HPPResponse.__init__(self,
sock=conn.sock,
debuglevel=conn.debuglevel,
strict=conn.strict,
method=conn._method)
self.chunked = False
self.will_close = False
self.length = 0
self.status = 200
self.reason = 'OK'
self.msg = HTTPMessage(StringIO(), seekable=0)
self.msg.fp = None
def inner(*args, **kwargs):
with patch('sanction.urlopen') as mock_urlopen:
bdata = type(data) is basestring and data.encode() or data
sheaders = ''
if headers is not None:
sheaders = '\r\n'.join(
['{}: {}'.format(k, v) for k, v in headers.items()])
bheaders = (sheaders or '').encode()
mock_urlopen.return_value = addinfourl(BytesIO(bdata),
HTTPMessage(
BytesIO(bheaders)),
'',
code=code)
fn(*args, **kwargs)
def expect_response(self):
if self.fp:
self.fp.close()
self.fp = None
self.fp = self.sock.makefile('rb', 0)
version, status, reason = self._read_status()
if status != CONTINUE:
self._read_status = lambda: (version, status, reason)
self.begin()
else:
self.status = status
self.reason = reason.strip()
self.version = 11
self.msg = HTTPMessage(self.fp, 0)
self.msg.fp = None
def datagramReceived(self, data, addr, outip):
if outip not in self.interfaces:
if self.INADDR_ANY not in self.interfaces:
return
req_line, data = data.split('\r\n', 1)
method, path, version = req_line.split(None, 3)
# check method
if method != 'M-SEARCH' or path != '*':
return
# parse header
headers = HTTPMessage(StringIO(data))
mx = int(headers.getheader('MX'))
# send M-SEARCH response
for udn in self.devices:
device = self.devices[udn]
delay = random() * mx
for packet in device.make_msearch_response(headers, (outip, self.port), addr):
buff = build_packet('HTTP/1.1 200 OK', packet)
self.reactor.callLater(delay, self._send_packet, self.ssdp, buff, addr)
delay += self.SSDP_INTERVAL
def find_user_password(self, realm, authuri):
""" Limit number of queries per request.
Note that retries needs to be reset in the calling code.
"""
# allow sending the username:password 5 times before failing!
if self.retries > 5:
from httplib import HTTPMessage
from StringIO import StringIO
raise urllib2.HTTPError(authuri, 401,
"basic auth failed for realm %r" % realm,
HTTPMessage(StringIO("")), None)
self.retries += 1
return urllib2.HTTPPasswordMgrWithDefaultRealm.find_user_password(
self, realm, authuri)
def from_dict(cls, data):
"""Create object from dict."""
obj = cls()
for k in cls.attrs:
if k in data:
setattr(obj, k, data[k])
obj.fp = cls.create_file_descriptor(obj.content)
obj.msg = HTTPMessage(io.StringIO(unicode()), 0)
for k, v in obj.headers.iteritems():
obj.msg.addheader(k, v)
obj.msg.headers = data["raw_headers"]
return obj
def __call__(self, opener, method, url, data, headers):
if url in self._exceptions:
raise self._exceptions[url]
if url in self._cookie_responses:
plain_headers = self._cookie_responses[url]
headers_text = '\n'.join([
header_name + ': ' + plain_headers[header_name]
for header_name in plain_headers
])
response_headers = HTTPMessage(StringIO(headers_text))
return self.build_response(url, 200, headers['Cookie'], 'OK',
response_headers)
elif url in self._responses:
return self.build_response(*self._responses[url])
else:
return self.build_response(url, 404, '', 'Not Found')
def __init__(self, replay_response, method=None):
self.reason = replay_response['status']['message']
self.status = replay_response['status']['code']
self.version = None
if 'body_quoted_printable' in replay_response:
self._content = quopri.decodestring(replay_response['body_quoted_printable'])
else:
self._content = replay_response['body'].decode('base64')
self.fp = StringIO(self._content)
msg_fp = StringIO('\r\n'.join('{}: {}'.format(h, v)
for h, v in replay_response['headers'].iteritems()))
self.msg = HTTPMessage(msg_fp)
self.msg.fp = None # httplib does this, okay?
length = self.msg.getheader('content-length')
self.length = int(length) if length else None
# Save method to handle HEAD specially as httplib does
self._method = method
def readbodychunked(self, ifile, ofile, dechunk=False, deflength=4096):
totaldata = 0
while True:
# Chunk length (discard trailing ';....')
clength = ifile.readline()
if not dechunk:
ofile.write(clength)
clength = clength.rstrip('\r\n').split(';', 1)[0]
clength = int(clength, 16)
if not clength:
break
# Chunk data
while clength:
data = ifile.read(min(clength, deflength))
ldata = len(data)
clength -= ldata
totaldata += ldata
ofile.write(data)
# Chunk end - trailing newline
chunkend = ifile.readline()
if not dechunk:
ofile.write(chunkend)
ofile.flush()
# Chunk trailers
if dechunk:
ofile = StringIO()
while True:
line = ifile.readline()
ofile.write(line)
if line == '\r\n' or line == '\r':
break
ofile.flush()
footers = None if not dechunk else HTTPMessage(ofile, seekable=0)
return (totaldata, footers)
def from_dict(cls, data):
"""Create object from dict."""
# Hack to ensure backwards compatibility with older versions of the
# that did not have the length and version attributes.
data.setdefault('length', len(data['content']))
data.setdefault('version', 10)
obj = cls()
for k in cls.attrs:
setattr(obj, k, data[k])
obj.fp = cls.create_file_descriptor(obj.content)
obj.msg = HTTPMessage(io.StringIO(unicode()), 0)
for k, v in obj.headers.iteritems():
obj.msg.addheader(k, v)
obj.msg.headers = data["raw_headers"]
return obj
def response(self, url, bypass_content): # bypass_content: { url1: [content, headers],
try:
data = json.loads(bypass_content)
if url in data:
d = data[url]
fp = io.BytesIO(d[0].encode('utf-8'))
headers = HTTPMessage(io.StringIO(unicode(d[1])), 0)
# for head in d[1].split("\n"):
# xy = head.split(":")
# headers.addheader(xy[0].strip(), xy[1].strip())
url = url
code = 200
msg = 'OK'
res = addinfourl(fp, headers, url, code)
youtubedl_logger.info('Request Bypassed for url -> '+ url)
return res
else:
youtubedl_logger.info('Request Skipped for url -> ' + url)
return None
except Exception as e:
youtubedl_logger.info('Request Exception for url -> ' + url)
traceback.print_exc()
return None
class Web_watch(Watch):
"""
Watch class that will check if http or rss pages are changed.
"""
url_ = ""
info_ = None
content_ = None
lastModified_ = None
digest_ = None
refresh_ = None
infoB_ = None
cached = 0
url2_ = ""
updated = False
actually_updated = False
type = 0
def __init__(self, specto, name, refresh, url, id, error_margin):
Watch.__init__(self, specto) #init superclass
self.refresh = refresh
self.id = id
self.url_ = url
if self.url_ == "":
self.specto.logger.log(
_("Watch: \"%s\" has error: empty url") % self.error, "error",
self.__class__)
self.name = name
self.error_margin = error_margin #the amount in percent (as a float) of what the filesize must change to consider the page changed
self.error = False
def dict_values(self):
return {
'name': self.name,
'refresh': self.refresh,
'uri': self.url_,
'error_margin': self.error_margin,
'type': 0
}
def start_watch(self):
""" Start the watch. """
self.thread_update()
def _real_update(self):
self.specto.notifier.connected_message(
True) #hide the network error message
lock = thread.allocate_lock()
lock.acquire()
t = thread.start_new_thread(self.update, (lock, ))
while lock.locked():
while gtk.events_pending():
gtk.main_iteration()
time.sleep(0.05)
while gtk.events_pending():
gtk.main_iteration()
def thread_update(self):
if not self.specto.connection_manager.connected():
self.specto.logger.log(_("No network connection detected"), "info",
self.__class__)
self.specto.notifier.connected_message(
False) #show the network error message
self.specto.connection_manager.add_callback(self._real_update)
self.specto.mark_watch_busy(False, self.id)
else:
self._real_update()
def update(self, lock):
""" See if a http or rss page changed. """
self.error = False
self.specto.mark_watch_busy(True, self.id)
self.specto.logger.log(
_("Updating watch: \"%s\"") % self.name, "info", self.__class__)
# Create a unique name for each url.
digest = md5.new(self.url_).digest()
cacheFileName = "".join(["%02x" % (ord(c), ) for c in digest])
self.cacheFullPath_ = os.path.join(cacheSubDir__, cacheFileName)
request = urllib2.Request(self.url_, None, {"Accept-encoding": "gzip"})
if (self.cached == 1) or (os.path.exists(self.cacheFullPath_)):
self.cached = 1
f = file(self.cacheFullPath_, "r") # Load up the cached version
self.infoB_ = HTTPMessage(f)
if self.infoB_.has_key('last-modified'):
request.add_header("If-Modified-Since",
self.infoB_['last-modified'])
if self.infoB_.has_key('ETag'):
request.add_header("If-None-Match", self.infoB_['ETag'])
try:
response = urllib2.urlopen(request)
except (urllib2.URLError, BadStatusLine), e:
self.error = True
self.specto.logger.log(
_("Watch: \"%s\" has error: ") % self.name + str(e), "error",
self.__class__)
else:
def clone(self):
cloneobj = HTTPMessage(StringIO(), seekable=0)
cloneobj.dict = self.dict.copy()
cloneobj.headers = copy(self.headers)
return cloneobj
class ReplayHTTPResponse(object):
"""
A replay response object, with just enough functionality to make
the various HTTP/URL libraries out there happy.
"""
__text_content_types = (
'text/',
'application/json',
)
def __init__(self, replay_response, method=None):
self.reason = replay_response['status']['message']
self.status = replay_response['status']['code']
self.version = None
if 'body_quoted_printable' in replay_response:
self._content = quopri.decodestring(replay_response['body_quoted_printable'])
else:
self._content = replay_response['body'].decode('base64')
self.fp = StringIO(self._content)
msg_fp = StringIO('\r\n'.join('{}: {}'.format(h, v)
for h, v in replay_response['headers'].iteritems()))
self.msg = HTTPMessage(msg_fp)
self.msg.fp = None # httplib does this, okay?
length = self.msg.getheader('content-length')
self.length = int(length) if length else None
# Save method to handle HEAD specially as httplib does
self._method = method
@classmethod
def make_replay_response(cls, response):
"""
Converts real response to replay_response dict which can be saved
and/or used to initialize a ReplayHTTPResponse.
"""
replay_response = {}
body = response.read() # undecoded byte string
# Add body to replay_response, either as quoted printable for
# text responses or base64 for binary responses.
if response.getheader('content-type', '') \
.startswith(cls.__text_content_types):
if response.getheader('content-encoding') in ['gzip', 'deflate']:
# http://stackoverflow.com/questions/2695152
body = zlib.decompress(body, 16 + zlib.MAX_WBITS)
del response.msg['content-encoding']
# decompression changes the length
if 'content-length' in response.msg:
response.msg['content-length'] = str(len(body))
replay_response['body_quoted_printable'] = quopri.encodestring(body)
else:
replay_response['body'] = body.encode('base64')
replay_response.update(dict(
status=dict(code=response.status, message=response.reason),
headers=dict(response.getheaders())))
return replay_response
def close(self):
self.fp = None
def isclosed(self):
return self.fp is None
def read(self, amt=None):
"""
The important parts of HTTPResponse.read()
"""
if self.fp is None:
return ''
if self._method == 'HEAD':
self.close()
return ''
if self.length is not None:
amt = min(amt, self.length)
# StringIO doesn't like read(None)
s = self.fp.read() if amt is None else self.fp.read(amt)
if not s:
self.close()
if self.length is not None:
self.length -= len(s)
if not self.length:
self.close()
return s
def getheader(self, name, default=None):
return self.msg.getheader(name, default)
def getheaders(self):
return self.msg.items()
class Web_watch(Watch):
"""
Watch class that will check if http or rss pages are changed.
"""
url_ = ""
info_ = None
content_ = None
lastModified_ = None
digest_ = None
refresh_ = None
infoB_ = None
cached = 0
url2_ = ""
updated = False
actually_updated = False
type = 0
def __init__(self, specto, name, refresh, url, id, error_margin):
Watch.__init__(self, specto) #init superclass
self.refresh = refresh
self.id = id
self.url_ = url
if self.url_ == "":
self.specto.logger.log(_("Watch: \"%s\" has error: empty url") % self.error, "error", self.__class__)
self.name = name
self.error_margin = error_margin#the amount in percent (as a float) of what the filesize must change to consider the page changed
self.error = False
def dict_values(self):
return { 'name': self.name, 'refresh': self.refresh, 'uri': self.url_, 'error_margin':self.error_margin, 'type':0 }
def start_watch(self):
""" Start the watch. """
self.thread_update()
def _real_update(self):
self.specto.notifier.connected_message(True)#hide the network error message
lock = thread.allocate_lock()
lock.acquire()
t=thread.start_new_thread(self.update,(lock,))
while lock.locked():
while gtk.events_pending():
gtk.main_iteration()
time.sleep(0.05)
while gtk.events_pending():
gtk.main_iteration()
def thread_update(self):
if not self.specto.connection_manager.connected():
self.specto.logger.log(_("No network connection detected"),
"info", self.__class__)
self.specto.notifier.connected_message(False) #show the network error message
self.specto.connection_manager.add_callback(self._real_update)
self.specto.mark_watch_busy(False, self.id)
else :
self._real_update()
def update(self, lock):
""" See if a http or rss page changed. """
self.error = False
self.specto.mark_watch_busy(True, self.id)
self.specto.logger.log(_("Updating watch: \"%s\"") % self.name, "info", self.__class__)
# Create a unique name for each url.
digest = md5.new(self.url_).digest()
cacheFileName = "".join(["%02x" % (ord(c),) for c in digest])
self.cacheFullPath_ = os.path.join(cacheSubDir__, cacheFileName)
request = urllib2.Request(self.url_, None, {"Accept-encoding" : "gzip"})
if (self.cached == 1) or (os.path.exists(self.cacheFullPath_)):
self.cached = 1
f = file(self.cacheFullPath_, "r")# Load up the cached version
self.infoB_ = HTTPMessage(f)
if self.infoB_.has_key('last-modified'):
request.add_header("If-Modified-Since", self.infoB_['last-modified'])
if self.infoB_.has_key('ETag'):
request.add_header("If-None-Match", self.infoB_['ETag'])
try:
response = urllib2.urlopen(request)
except (urllib2.URLError, BadStatusLine), e:
self.error = True
self.specto.logger.log(_("Watch: \"%s\" has error: ") % self.name + str(e), "error", self.__class__)
else:
def __init__(self):
HTTPMessage.__init__(self, BytesIO())
def parse_flow(IP):
styleID = ''.join(choice(string.ascii_lowercase + string.digits) for x in range(randint(8, 12)))
ssize = len(styleID)
p = sub.Popen(['tcpflow', '-T %T--%A-%B', '-cJB', '-r', (os.getenv('PROCDOTPLUGIN_WindumpFilePcap'))], stdout=sub.PIPE, stderr=sub.PIPE)
stdout, stderr = p.communicate()
stdout = stdout.replace('\r\n', '\n')
if IP not in stdout:
e = str("No tcp flows found for ")+IP
open(out, 'ab').write(e)
else:
if os.getenv('PROCDOTPLUGIN_PluginEngineVersion') is not None:
open(out,'ab').write('{{{style-id:default;color:blue;style-id:'+styleID+';color:red}}}')
m = re.findall ( '\x1b\[0;31m(.*?)\x1b\[0m|\x1b\[0;34m(.*?)\x1b\[0m', stdout, re.DOTALL)
m = iter(m)
for b, r in m:
if b == '':
if IP in r:
r = r[56:]
r = re.sub( '[^!\"#\$%&\'\(\)\*\+,-\./0-9:;<=>\?@A-Z\[\]\^_`a-z\{\|\}\\\~\t\n\r ]','.', r)
if os.stat(out).st_size <= 53 + ssize:
if os.getenv('PROCDOTPLUGIN_PluginEngineVersion') is not None:
open(out,'ab').write('<'+styleID+'>'+r+'</'+styleID+'>')
else:
open(out,'ab').write(r)
else:
if os.getenv('PROCDOTPLUGIN_PluginEngineVersion') is not None:
open(out,'ab').write('\n\n'+'<'+styleID+'>'+r+'</'+styleID+'>')
else:
open(out,'ab').write('\n\n'+r)
else:
if IP in b:
b = b[56:]
match = re.match( '^HTTP.*', b)
try:
if match:
length = 1
num = 0
while length != num:
d = zlib.decompressobj(16+zlib.MAX_WBITS)
output = StringIO.StringIO(b)
status_line = output.readline()
msg = HTTPMessage(output, 0)
isLength = msg.get('Content-Length')
isGZipped = msg.get('content-encoding', '').find('gzip') >= 0
isChunked = msg.get('Transfer-Encoding', '').find('chunked') >= 0
if isGZipped and isChunked:
offset = msg.fp.readline()
body = msg.fp.read()
num = int(offset, 16)
encdata = ''
newdata = ''
encdata =body[:num]
length = len(encdata)
if length != num:
c = next(m)
d, e = c
b = b + d[56:]
else:
newdata = d.decompress(encdata)
header = str(msg)
open(out,'ab').write(status_line)
open(out,'ab').write(header)
open(out,'ab').write('\n')
open(out,'ab').write(newdata)
elif isGZipped:
length = int(isLength)
body = msg.fp.read()
num = len(body)
if length != num:
c = next(m)
d, e = c
if IP in d:
b = b + d[56:]
else:
data = d.decompress(body)
header = str(msg)
open(out,'ab').write(status_line)
open(out,'ab').write(header)
open(out,'ab').write('\n')
open(out,'ab').write(data)
else:
length = 1
num = 1
body = msg.fp.read()
body = re.sub( '[^!\"#\$%&\'\(\)\*\+,-\./0-9:;<=>\?@A-Z\[\]\^_`a-z\{\|\}\\\~\t\n\r ]','.', body)
header = str(msg)
open(out,'ab').write(status_line)
open(out,'ab').write(header)
open(out,'ab').write('\n')
open(out,'ab').write(body)
else:
b = re.sub( '[^!\"#\$%&\'\(\)\*\+,-\./0-9:;<=>\?@A-Z\[\]\^_`a-z\{\|\}\\\~\t\n\r ]','.', b)
open(out,'ab').write(b)
except:
open(out,'ab').write('DECOMPRESSION ERROR')
open(out,'ab').write('\n\n')
b = re.sub( '[^!\"#\$%&\'\(\)\*\+,-\./0-9:;<=>\?@A-Z\[\]\^_`a-z\{\|\}\\\~\t\n\r ]','.', b)
open(out,'ab').write(b)
def clone(self):
cloneobj = HTTPMessage(StringIO(), seekable=0)
cloneobj.dict = self.dict.copy()
cloneobj.headers = copy(self.headers)
return cloneobj
def __init__(self):
self.msg = HTTPMessage(StringIO())
self.msg['content-type'] = 'text/plain'
self.status = 400
self.reason = 'Bad Request'
def __init__(self, method, path, version='HTTP/1.1'):
self.method = method.upper()
self.path = path
self.version = version
self.msg = HTTPMessage(StringIO())
self.body = ''
def make_headers(fp): return HTTPMessage(fp) if Compatibility.PY2 else parse_headers(fp)
def __init__(self):
HTTPMessage.__init__(self, BytesIO())
def parse_headers(fp):
return HTTPMessage(fp, 0)
def make_headers(fp):
return HTTPMessage(fp) if PY2 else parse_headers(fp)
def parse_flow(IP):
p = sub.Popen(['tcpflow', '-T %T--%A-%B', '-cJB', '-r', (os.getenv('PROCDOTPLUGIN_WindumpFilePcap'))], stdout=sub.PIPE, stderr=sub.PIPE)
stdout, stderr = p.communicate()
stdout = stdout.replace('\r\n', '\n')
if IP not in stdout:
e = str("No tcp flows found for ")+IP
open(out, 'ab').write(e)
else:
m = re.findall ( '\x1b\[0;3[1|4]m(.*?)\x1b\[0m', stdout, re.DOTALL)
m = iter(m)
for line in m:
if IP in line:
line = line[56:]
match = re.match( '^HTTP.*', line)
try:
if match:
length = 1
num = 0
while length != num:
d = zlib.decompressobj(16+zlib.MAX_WBITS)
output = StringIO.StringIO(line)
status_line = output.readline()
msg = HTTPMessage(output, 0)
isGZipped = msg.get('content-encoding', '').find('gzip') >= 0
isChunked = msg.get('Transfer-Encoding', '').find('chunked') >= 0
if isGZipped and isChunked:
offset = msg.fp.readline()
body = msg.fp.read()
num = int(offset, 16)
encdata = ''
newdata = ''
encdata =body[:num]
length = len(encdata)
if length != num:
line = line + next(m)[56:]
else:
newdata = d.decompress(encdata)
header = str(msg)
open(out,'ab').write(status_line)
open(out,'ab').write(header)
open(out,'ab').write('\n')
open(out,'ab').write(newdata)
elif isGZipped:
length = 1
num = 1
body = msg.fp.read()
data = d.decompress(body)
header = str(msg)
open(out,'ab').write(status_line)
open(out,'ab').write(header)
open(out,'ab').write('\n')
open(out,'ab').write(data)
else:
break
else:
line = re.sub( '[^!\"#\$%&\'\(\)\*\+,-\./0-9:;<=>\?@A-Z\[\]\^_`a-z\{\|\}\\\~\t\n\r ]','.', line)
open(out,'ab').write(line)
except:
open(out,'ab').write('DECOMPRESSION ERROR')
open(out,'ab').write('\n\n')
open(out,'ab').write(line)
class ReplayHTTPResponse(object):
"""
A replay response object, with just enough functionality to make
the various HTTP/URL libraries out there happy.
"""
__text_content_types = (
'text/',
'application/json',
)
def __init__(self, replay_response, method=None):
self.reason = replay_response['status']['message']
self.status = replay_response['status']['code']
self.version = None
if 'body_text' in replay_response:
# JSON decoder returns unicode, not str, so this needs to be
# encoded to properly reproduce content off the wire.
self._content = replay_response['body_text'].encode('utf8')
elif 'body_quoted_printable' in replay_response:
# quopri.decodestring returns str, which is correct for content off
# the wire.
self._content = quopri.decodestring(replay_response['body_quoted_printable'])
else:
# .decode('base64') returns str, which is correct for content off
# the wire.
self._content = replay_response['body'].decode('base64')
self.fp = StringIO(self._content)
msg_fp = StringIO('\r\n'.join('{}: {}'.format(h, v)
for h, v in replay_response['headers'].iteritems()))
self.msg = HTTPMessage(msg_fp)
self.msg.fp = None # httplib does this, okay?
length = self.msg.getheader('content-length')
self.length = int(length) if length else None
# Save method to handle HEAD specially as httplib does
self._method = method
@classmethod
def make_replay_response(cls, response):
"""
Converts real response to replay_response dict which can be saved
and/or used to initialize a ReplayHTTPResponse.
"""
replay_response = {}
body = response.read() # undecoded byte string
# Add body to replay_response. Try to use simple text, falling back to
# quoted printable or base64 as required for binary responses.
if response.getheader('content-type', '') \
.startswith(cls.__text_content_types):
if response.getheader('content-encoding') in ['gzip', 'deflate']:
# http://stackoverflow.com/questions/2695152
body = zlib.decompress(body, 16 + zlib.MAX_WBITS)
del response.msg['content-encoding']
# decompression changes the length
if 'content-length' in response.msg:
response.msg['content-length'] = str(len(body))
try:
# Store body directly as text if it will decode properly.
body.decode('utf8')
replay_response['body_text'] = body
except UnicodeDecodeError:
# Store body as quoted printable.
# Remove unneccessary =\n pairs which make searching hard.
# These exist for line-wrapping in email, which is entirely
# pointless here.
body_quoted_printable = quopri.encodestring(body)
body_quoted_printable = body_quoted_printable.replace('=\n', '')
replay_response['body_quoted_printable'] = body_quoted_printable
else:
replay_response['body'] = body.encode('base64')
replay_response.update(dict(
status=dict(code=response.status, message=response.reason),
headers=dict(response.getheaders())))
return replay_response
def close(self):
self.fp = None
def isclosed(self):
return self.fp is None
def read(self, amt=None):
"""
The important parts of HTTPResponse.read()
"""
if self.fp is None:
return ''
if self._method == 'HEAD':
self.close()
return ''
if self.length is not None:
amt = min(amt, self.length)
# StringIO doesn't like read(None)
s = self.fp.read() if amt is None else self.fp.read(amt)
if not s:
self.close()
if self.length is not None:
self.length -= len(s)
if not self.length:
self.close()
return s
def getheader(self, name, default=None):
return self.msg.getheader(name, default)
def getheaders(self):
return self.msg.items()
