class Attachment:
    def __init__(self, part):
        encoding = part.encoding or "utf-8"
        self.headers = CaseInsensitiveDict({
            k.decode(encoding): v.decode(encoding)
            for k, v in part.headers.items()
        })
        self.content_type = self.headers.get("Content-Type", None)
        self.content_id = self.headers.get("Content-ID", None)
        self.content_location = self.headers.get("Content-Location", None)
        self._part = part

    def __repr__(self):
        return "<Attachment(%r, %r)>" % (self.content_id, self.content_type)

    @cached_property
    def content(self):
        """Return the content of the attachment

        :rtype: bytes or str

        """
        encoding = self.headers.get("Content-Transfer-Encoding", None)
        content = self._part.content

        if encoding == "base64":
            return base64.b64decode(content)
        elif encoding == "binary":
            return content.strip(b"\r\n")
        else:
            return content
Exemple #2
0
    def string_to_sign(self, request):
        """
        Generates the string we need to sign on.

        Params:
            - request   The request object

        Returns
            String ready to be signed on

        """

        # We'll use case insensitive dict to store the headers
        h = CaseInsensitiveDict()
        # Add the hearders
        h.update(request.headers)

        # If we have an 'x-amz-date' header,
        # we'll try to use it instead of the date
        if b'x-amz-date' in h or 'x-amz-date' in h:
            date = ''
        else:
            # No x-amz-header, we'll generate a date
            date = h.get('Date') or self._get_date()

        # Set the date header
        request.headers['Date'] = date

        # A fix for the content type header extraction in python 3
        # This have to be done because requests will try to set
        # application/www-url-encoded header if we pass bytes as the content,
        # and the content-type is set with a key that is b'Content-Type' and
        # not 'Content-Type'
        content_type = ''
        if b'Content-Type' in request.headers:
            # Fix content type
            content_type = h.get(b'Content-Type')
            del request.headers[b'Content-Type']
            request.headers['Content-Type'] = content_type

        # The string we're about to generate
        # There's more information about it here:
        # http://docs.aws.amazon.com/AmazonS3/latest/dev/
        # RESTAuthentication.html#ConstructingTheAuthenticationHeader
        msg = [
            # HTTP Method
            request.method,
            # MD5 If provided
            h.get(b'Content-MD5', '') or h.get('Content-MD5', ''),
            # Content type if provided
            content_type or h.get('Content-Type', ''),
            # Date
            date,
            # Canonicalized special amazon headers and resource uri
            self._get_canonicalized_amz_headers(h) +
            self._get_canonicalized_resource(request)
        ]

        # join with a newline and return
        return '\n'.join(msg)
Exemple #3
0
    def response(self):
        if self._response is None and self.request is not None:
            request = self.request
            existing = None
            if self.use_cache:
                existing = self.context.get_tag(self.request_id)
            if existing is not None:
                headers = CaseInsensitiveDict(existing.get('headers'))
                last_modified = headers.get('last-modified')
                if last_modified:
                    request.headers['If-Modified-Since'] = last_modified

                etag = headers.get('etag')
                if etag:
                    request.headers['If-None-Match'] = etag

            self._rate_limit(request.url)

            session = self.http.session
            prepared = session.prepare_request(request)
            response = session.send(prepared,
                                    stream=True,
                                    verify=False,
                                    allow_redirects=self.allow_redirects)

            if existing is not None and response.status_code == 304:
                self.context.log.info("Using cached HTTP response: %s",
                                      response.url)
                self.apply_data(existing)
            else:
                self._response = response

            # update the serialised session with cookies etc.
            self.http.save()
        return self._response
Exemple #4
0
class Attachment(object):
    def __init__(self, part):
        encoding = part.encoding or 'utf-8'
        self.headers = CaseInsensitiveDict({
            k.decode(encoding): v.decode(encoding)
            for k, v in part.headers.items()
        })
        self.content_type = self.headers.get('Content-Type', None)
        self.content_id = self.headers.get('Content-ID', None)
        self.content_location = self.headers.get('Content-Location', None)
        self._part = part

    def __repr__(self):
        return '<Attachment(%r, %r)>' % (self.content_id, self.content_type)

    @cached_property
    def content(self):
        """Return the content of the attachment

        :rtype: bytes or str

        """
        encoding = self.headers.get('Content-Transfer-Encoding', None)
        content = self._part.content

        if encoding == 'base64':
            return base64.b64decode(content)
        elif encoding == 'binary':
            return content.strip(b'\r\n')
        else:
            return content
Exemple #5
0
class Attachment(object):
    def __init__(self, part):
        self.headers = CaseInsensitiveDict({
            k.decode(part.encoding): v.decode(part.encoding)
            for k, v in part.headers.items()
        })
        self.content_type = self.headers.get('Content-Type', None)
        self.content_id = self.headers.get('Content-ID', None)
        self.content_location = self.headers.get('Content-Location', None)
        self._part = part

    def __repr__(self):
        return '<Attachment(%r, %r)>' % (self.content_id, self.content_type)

    @cached_property
    def content(self):
        encoding = self.headers.get('Content-Transfer-Encoding', None)
        content = self._part.content

        if encoding == 'base64':
            return base64.b64decode(content)
        elif encoding == 'binary':
            return content
        else:
            return content
Exemple #6
0
    def string_to_sign(self, request):
        """
        Generates the string we need to sign on.

        Params:
            - request   The request object

        Returns
            String ready to be signed on

        """

        # We'll use case insensitive dict to store the headers
        h = CaseInsensitiveDict()
        # Add the hearders
        h.update(request.headers)

        # If we have an 'x-amz-date' header,
        # we'll try to use it instead of the date
        if b'x-amz-date' in h or 'x-amz-date' in h:
            date = ''
        else:
            # No x-amz-header, we'll generate a date
            date = h.get('Date') or self._get_date()

        # Set the date header
        request.headers['Date'] = date

        # A fix for the content type header extraction in python 3
        # This have to be done because requests will try to set
        # application/www-url-encoded header if we pass bytes as the content,
        # and the content-type is set with a key that is b'Content-Type' and
        # not 'Content-Type'
        content_type = ''
        if b'Content-Type' in request.headers:
            # Fix content type
            content_type = h.get(b'Content-Type')
            del request.headers[b'Content-Type']
            request.headers['Content-Type'] = content_type

        # The string we're about to generate
        # There's more information about it here:
        # http://docs.aws.amazon.com/AmazonS3/latest/dev/
        # RESTAuthentication.html#ConstructingTheAuthenticationHeader
        msg = [
            # HTTP Method
            request.method,
            # MD5 If provided
            h.get(b'Content-MD5', '') or h.get('Content-MD5', ''),
            # Content type if provided
            content_type or h.get('Content-Type', ''),
            # Date
            date,
            # Canonicalized special amazon headers and resource uri
            self._get_canonicalized_amz_headers(h) +
            self._get_canonicalized_resource(request)
        ]

        # join with a newline and return
        return '\n'.join(msg)
Exemple #7
0
    def string_to_sign(self, request):
        h = CaseInsensitiveDict()
        h.update(request.headers)

        # Try to use

        if b'x-amz-date' in h or 'x-amz-date' in h:
            date = ''
        else:
            date = h.get('Date') or self._get_date()
            request.headers['Date'] = date

        # Set the date header
        request.headers['Date'] = date

        # A fix for the content type header extraction in python 3
        # This have to be done because requests will try to set application/www-url-encoded herader
        # if we pass bytes as the content, and the content-type is set with a key that is b'Content-Type' and not
        # 'Content-Type'
        content_type = ''
        if b'Content-Type' in request.headers:
            # Fix content type
            content_type = h.get(b'Content-Type')
            del request.headers[b'Content-Type']
            request.headers['Content-Type'] = content_type

        msg = [
            request.method,
            h.get(b'Content-MD5', '') or h.get('Content-MD5', ''),
            content_type or h.get('Content-Type', ''),
            date,
            self._get_canonicalized_amz_headers(h) + self._get_canonicalized_resource(request)
        ]

        return '\n'.join(msg)
Exemple #8
0
 def test_get(self):
     cid = CaseInsensitiveDict()
     cid['spam'] = 'oneval'
     cid['SPAM'] = 'blueval'
     assert cid.get('spam') == 'blueval'
     assert cid.get('SPAM') == 'blueval'
     assert cid.get('sPam') == 'blueval'
     assert cid.get('notspam', 'default') == 'default'
 def test_get(self):
     cid = CaseInsensitiveDict()
     cid['spam'] = 'oneval'
     cid['SPAM'] = 'blueval'
     self.assertEqual(cid.get('spam'), 'blueval')
     self.assertEqual(cid.get('SPAM'), 'blueval')
     self.assertEqual(cid.get('sPam'), 'blueval')
     self.assertEqual(cid.get('notspam', 'default'), 'default')
 def test_get(self):
     cid = CaseInsensitiveDict()
     cid['spam'] = 'oneval'
     cid['SPAM'] = 'blueval'
     assert cid.get('spam') == 'blueval'
     assert cid.get('SPAM') == 'blueval'
     assert cid.get('sPam') == 'blueval'
     assert cid.get('notspam', 'default') == 'default'
Exemple #11
0
 def test_get(self):
     cid = CaseInsensitiveDict()
     cid["spam"] = "oneval"
     cid["SPAM"] = "blueval"
     assert cid.get("spam") == "blueval"
     assert cid.get("SPAM") == "blueval"
     assert cid.get("sPam") == "blueval"
     assert cid.get("notspam", "default") == "default"
Exemple #12
0
 def test_get(self):
     cid = CaseInsensitiveDict()
     cid["spam"] = "oneval"
     cid["SPAM"] = "blueval"
     self.assertEqual(cid.get("spam"), "blueval")
     self.assertEqual(cid.get("SPAM"), "blueval")
     self.assertEqual(cid.get("sPam"), "blueval")
     self.assertEqual(cid.get("notspam", "default"), "default")
 def test_get(self):
     cid = CaseInsensitiveDict()
     cid['spam'] = 'oneval'
     cid['SPAM'] = 'blueval'
     self.assertEqual(cid.get('spam'), 'blueval')
     self.assertEqual(cid.get('SPAM'), 'blueval')
     self.assertEqual(cid.get('sPam'), 'blueval')
     self.assertEqual(cid.get('notspam', 'default'), 'default')
def aws_invoke(app,gateway_input,server_name='localhost',server_port='5000',http_protocol='HTTP/1.1',TLS=True,block_headers=True):
   headers = CaseInsensitiveDict(gateway_input.get('headers',{}))
   requestContext = gateway_input.get('requestContext')
   queryStringParameters = gateway_input.get('queryStringParameters',{})
   clientIp = headers.get('x-forwarded-for')
   if clientIp is None:
      clientIp = requestContext.get('identity',{}).get('sourceIp') if requestContext is not None else ''
   else:
      clientIp = clientIp.split(',')[0]
   environ = {
      'REQUEST_METHOD': gateway_input.get('httpMethod','GET').upper(),
      'SCRIPT_NAME': '',
      'PATH_INFO': gateway_input.get('path','/'),
      'QUERY_STRING': urlencode(queryStringParameters) if queryStringParameters is not None else '',
      'SERVER_NAME': headers.get('host',server_name),
      'SERVER_PORT': headers.get('x-forwarded-port',server_port),
      'SERVER_PROTOCOL': http_protocol,
      'SERVER_SOFTWARE': 'flask-serverless',
      'REMOTE_ADDR': clientIp,
      'wsgi.version': (1, 0),
      'wsgi.url_scheme': headers.get('x-forwarded-proto','https' if TLS else 'http'),
      'wsgi.input': None,
      'wsgi.errors': sys.stderr,
      'wsgi.multiprocess': True,
      'wsgi.multithread': False,
      'wsgi.run_once': True
   }

   if environ['REQUEST_METHOD']=='POST' or environ['REQUEST_METHOD']=='PUT':
      contentType = headers.get('content-type','application/octet-stream')
      parsedContentType = parse_options_header(contentType)
      raw = gateway_input.get('body')
      if raw is None or gateway_input.get('isBase64Encoded',False):
         body = b64decode(raw) if raw is not None else None
      else:
         body = raw.encode(parsedContentType[1].get('charset','utf-8'))
      add_body(environ,body,contentType)

   add_headers(environ,headers,block_headers)

   response = Response.from_app(app.wsgi_app, environ)

   gateway_output = {
      'headers' : dict(response.headers),
      'statusCode' : response.status_code,
   }

   compressed = response.headers.get('Content-Encoding')=='gzip'

   responseType = parse_options_header(response.headers.get('Content-Type','application/octet-stream'))
   if not compressed and ('charset' in responseType[1] or responseType[0] in textTypes or responseType[0][0:5]=='text/'):
      gateway_output['body'] = response.data.decode(responseType[1].get('charset','utf-8'))
      gateway_output['isBase64Encoded'] = False
   else:
      gateway_output['body'] = b64encode(response.data).decode('utf-8')
      gateway_output['isBase64Encoded'] = True

   return gateway_output
Exemple #15
0
def _guess_mime_type(headers: CaseInsensitiveDict) -> Optional[str]:
    location = headers.get('location')
    if location:
        mime_type, _ = mimetypes.guess_type(location)
        if mime_type:
            return mime_type

    # Parse mime type from content-type header, e.g. 'image/jpeg;charset=US-ASCII' -> 'image/jpeg'
    mime_type, _ = cgi.parse_header(headers.get('content-type', ''))
    return mime_type or None
Exemple #16
0
    def __init__(self,
                 response=None,
                 status=None,
                 headers=None,
                 mimetype=None,
                 content_type=None,
                 direct_passthrough=False):
        headers = CaseInsensitiveDict(headers) if headers is not None else None
        if response is not None and isinstance(
                response, BaseResponse) and response.headers is not None:
            headers = CaseInsensitiveDict(response.headers)

        if headers is None:
            headers = CaseInsensitiveDict()

        h = headers
        h['Access-Control-Allow-Origin'] = headers.get(
            'Access-Control-Allow-Origin', '*')
        h['Access-Control-Allow-Methods'] = headers.get(
            'Access-Control-Allow-Methods',
            "GET, PUT, POST, HEAD, OPTIONS, DELETE")
        h['Access-Control-Max-Age'] = headers.get('Access-Control-Max-Age',
                                                  "21600")
        h['Cache-Control'] = headers.get(
            'Cache-Control', "no-cache, must-revalidate, no-store")
        if 'Access-Control-Allow-Headers' not in headers and len(
                headers.keys()) > 0:
            h['Access-Control-Allow-Headers'] = ', '.join(iterkeys(headers))

        data = None
        if response is not None and isinstance(response, string_types):
            data = response
            response = None

        if response is not None and isinstance(response, BaseResponse):
            new_response_headers = CaseInsensitiveDict(
                response.headers if response.headers is not None else {})
            new_response_headers.update(h)
            response.headers = new_response_headers
            headers = None
            data = response.get_data()

        else:
            headers.update(h)
            headers = dict(headers)

        super(IppResponse,
              self).__init__(response=response,
                             status=status,
                             headers=headers,
                             mimetype=mimetype,
                             content_type=content_type,
                             direct_passthrough=direct_passthrough)
        if data is not None:
            self.set_data(data)
Exemple #17
0
    def prepare_response(self, cached):
        """Verify our vary headers match and construct a real urllib3
        HTTPResponse object.
        """
        # Special case the '*' Vary value as it means we cannot actually
        # determine if the cached response is suitable for this request.
        if "*" in cached.get("vary", {}):
            return

        body_raw = cached["response"].pop("body")

        headers = CaseInsensitiveDict(data=cached['response']['headers'])
        if headers.get('transfer-encoding', '') == 'chunked':
            headers.pop('transfer-encoding')

        cached['response']['headers'] = headers

        try:
            body = io.BytesIO(body_raw)
        except TypeError:
            # This can happen if cachecontrol serialized to v1 format (pickle)
            # using Python 2. A Python 2 str(byte string) will be unpickled as
            # a Python 3 str (unicode string), which will cause the above to
            # fail with:
            #
            #     TypeError: 'str' does not support the buffer interface
            body = io.BytesIO(body_raw.encode('utf8'))

        return HTTPResponse(
            body=body,
            preload_content=False,
            **cached["response"]
        )
Exemple #18
0
    def prepare_response(self, request, cached):
        """Verify our vary headers match and construct a real urllib3
        HTTPResponse object.
        """
        # Special case the '*' Vary value as it means we cannot actually
        # determine if the cached response is suitable for this request.
        if "*" in cached.get("vary", {}):
            return

        # Ensure that the Vary headers for the cached response match our
        # request
        for header, value in cached.get("vary", {}).items():
            if request.headers.get(header, None) != value:
                return

        body_file = cached[u'response'].pop(u'body')
        body = open(os.path.join(self.cache.directory, body_file), 'rb')

        headers = CaseInsensitiveDict(data=cached['response']['headers'])
        if headers.get('transfer-encoding', '') == 'chunked':
            headers.pop('transfer-encoding')

        cached['response']['headers'] = headers

        return HTTPResponse(
            body=body,
            preload_content=False,
            **cached["response"]
        )
Exemple #19
0
        def call_processor_with_span_context(self, seqid, iprot, oprot):
            context = baseplate.make_context_object()

            # Allow case-insensitivity for THeader headers
            headers = CaseInsensitiveDict(data=iprot.get_headers())

            try:
                trace_info = _extract_trace_info(headers)
            except (KeyError, ValueError):
                trace_info = None

            edge_payload = headers.get(b"Edge-Request", None)
            if edge_context_factory:
                edge_context = edge_context_factory.from_upstream(edge_payload)
                edge_context.attach_context(context)
            else:
                # just attach the raw context so it gets passed on
                # downstream even if we don't know how to handle it.
                context.raw_request_context = edge_payload

            baseplate.make_server_span(context,
                                       name=fn_name,
                                       trace_info=trace_info)

            context.headers = headers

            handler = processor._handler
            context_aware_handler = _ContextAwareHandler(
                handler, context, logger)
            context_aware_processor = processor.__class__(
                context_aware_handler)
            return processor_fn(context_aware_processor, seqid, iprot, oprot)
Exemple #20
0
class DDWRT(Router):
    def __init__(self, conf, hostnames):
        self.hostnames = CaseInsensitiveDict()
        self.hostnames.update(hostnames)
        self.conf = conf
        self.auth = self.conf.auth()

    def clients(self):
        """ Receives all currently logged in users in a wifi network.

        :rtype : list
        :return: Returns a list of dicts, containing the following keys: mac, ipv4, seen, hostname
        """
        clients = self._get_clients_raw()

        clients_json = []
        for client in clients:
            client_hostname_from_router = client[0]
            client_ipv4 = client[1].strip()
            client_mac = client[2].strip().upper()
            client_hostname = self.hostnames.get(client_mac, client_hostname_from_router).strip()
            client_connections = int(client[3].strip())

            # Clients with less than 20 connections are considered offline
            if client_connections < 20:
                continue

            clients_json.append({
                'mac': client_mac,
                'ipv4': client_ipv4,
                'seen': int(time.time()),
                'hostname': client_hostname,
            })

        logger.debug('The router got us {} clients.'.format(len(clients_json)))
        logger.debug(str(clients_json))
        return clients_json

    def _get_clients_raw(self):
        info_page = self.conf.internal()
        response = requests.get(info_page, auth=self.auth)
        logger.info('Got response from router with code {}.'.format(response.status_code))
        return DDWRT._convert_to_clients(response.text) or []

    @staticmethod
    def _convert_to_clients(router_info_all):
        # Split router info in lines and filter empty info
        router_info_lines = filter(None, router_info_all.split("\n"))

        # Get key / value of router info
        router_info_items = dict()
        for item in router_info_lines:
            key, value = item[1:-1].split("::")  # Remove curly braces and split
            router_info_items[key.strip()] = value.strip()

        # Get client info as a list
        arp_table = utils.groupn(router_info_items['arp_table'].replace("'", "").split(","), 4)
        dhcp_leases = utils.groupn(router_info_items['dhcp_leases'].replace("'", "").split(","), 5)

        return arp_table if (len(arp_table) > 0) else []
Exemple #21
0
    def _index_into_idol(self, documents, query):
        index_data = ''
        for _d in documents:
            fields = _d.get('fields', [])
            content = _d.get('drecontent', '')
            DOCUMENTS = _d.get('content', {}).get('DOCUMENT', [])
            for DOC in DOCUMENTS:
                for key in DOC:
                    if key == 'DRECONTENT':
                        for value in DOC[key]:
                            content += value
                    else:
                        for value in DOC[key]:
                            fields.append((key, value))

            index_data += '\n'.join(
                [f"#DREREFERENCE {_d.get('reference')}"] +
                [f"#DREFIELD {_f[0]}=\"{_f[1]}\"" for _f in fields] +
                [f"#DRECONTENT", f"{content}", "#DREENDDOC\n\n"])
        # add to queue
        _query = CaseInsensitiveDict(query)
        if _query.get('priority', 0) >= 100:  # bypass queue
            self.post_index_data(_query,
                                 [(None, _query, index_data, len(documents))])
        else:
            self.add_into_batch_queue(_query, index_data, len(documents))
    def prepare_response(self, cached):
        """Verify our vary headers match and construct a real urllib3
        HTTPResponse object.
        """
        # Special case the '*' Vary value as it means we cannot actually
        # determine if the cached response is suitable for this request.
        if "*" in cached.get("vary", {}):
            return

        body_raw = cached["response"].pop("body")

        headers = CaseInsensitiveDict(data=cached['response']['headers'])
        if headers.get('transfer-encoding', '') == 'chunked':
            headers.pop('transfer-encoding')

        cached['response']['headers'] = headers

        try:
            body = io.BytesIO(body_raw)
        except TypeError:
            # This can happen if cachecontrol serialized to v1 format (pickle)
            # using Python 2. A Python 2 str(byte string) will be unpickled as
            # a Python 3 str (unicode string), which will cause the above to
            # fail with:
            #
            #     TypeError: 'str' does not support the buffer interface
            body = io.BytesIO(body_raw.encode('utf8'))

        return HTTPResponse(body=body,
                            preload_content=False,
                            **cached["response"])
Exemple #23
0
def parser_cookies_by_headers(headers: CaseInsensitiveDict):
    cookies = RequestsCookieJar()
    cookie_str = headers.get("Cookie")
    if cookie_str:
        for cookie in cookie_str.strip().split(";"):
            key, value = cookie.strip().split("=")
            cookies.set(key.strip(), value.strip())
    return cookies
Exemple #24
0
def get_file_name(headers: CaseInsensitiveDict) -> str:
    file_name = str(uuid.uuid4()) + ".csv"
    content_disposition = headers.get("Content-Disposition")
    if content_disposition is not None:
        if "attachment;" in content_disposition:
            fn_directive = content_disposition.split(" ")[1]
            file_name = fn_directive.split("=")[1]
    return file_name
Exemple #25
0
    def build_response(self, req, resp):
        response = requests.adapters.HTTPAdapter.build_response(
            self, req, resp)

        if response.headers.get("content-type") != None:
            if response.headers["content-type"].startswith(
                    "multipart/encrypted"):
                _, options_part = response.headers["content-type"].split(
                    ";", 1)

                options = CaseInsensitiveDict()

                for item in options_part.split(";"):
                    key, value = item.split("=")
                    if value[0] == '"' and value[-1] == '"':
                        value = value[1:-1]
                        value = value.replace(b'\\\\',
                                              b'\\').replace(b'\\"', b'"')

                    options[key] = value

                if options.get("protocol") is not None and options[
                        "protocol"] == "application/HTTP-Kerberos-session-encrypted":
                    boundary = options["boundary"]
                    encrypted_data = None
                    re_multipart = r'(?:--' + boundary + r'(?:(?:\r\n)|(?:--(?:\r\n)*)))'
                    for part in re.split(re_multipart, response.content):
                        if part == '':
                            continue
                        (header_raw, data) = part.split('\r\n', 1)
                        key, value = map(lambda x: x.strip(),
                                         header_raw.split(":"))
                        if key.lower(
                        ) == "content-type" and value == "application/HTTP-Kerberos-session-encrypted":
                            _, orginaltype = map(lambda x: x.strip(),
                                                 data.split(":"))
                            original_values = CaseInsensitiveDict()
                            for item in orginaltype.split(";"):
                                subkey, subvalue = item.split("=")
                                original_values[subkey] = subvalue

                        if key.lower(
                        ) == "content-type" and value == "application/octet-stream":
                            encrypted_data = data

                    con = self.get_connection(req.url, None)
                    decrypted = HTTPMSKerberosAdapter.krb_dict[
                        req.url].decrypt(encrypted_data)
                    response.headers["Content-Type"] = original_values[
                        "type"] + "; charset=" + original_values["charset"]
                    response.headers["Content-Length"] = len(decrypted)
                    response.encoding = requests.utils.get_encoding_from_headers(
                        response.headers)
                    response.raw = StringIO.StringIO(decrypted)
                    response._content_consumed = False
                    response._content = False

        return response
Exemple #26
0
def get_retry_after(response):
    """Get the value of Retry-After in seconds.

    :param response: The PipelineResponse object
    :type response: ~azure.core.pipeline.PipelineResponse
    :return: Value of Retry-After in seconds.
    :rtype: float or None
    """
    headers = CaseInsensitiveDict(response.http_response.headers)
    retry_after = headers.get("retry-after")
    if retry_after:
        return parse_retry_after(retry_after)
    for ms_header in ["retry-after-ms", "x-ms-retry-after-ms"]:
        retry_after = headers.get(ms_header)
        if retry_after:
            parsed_retry_after = parse_retry_after(retry_after)
            return parsed_retry_after / 1000.0
    return None
class GroupStore:
    """
      Simple in-memory store for group info. Attribute names are case-insensitive.
      Users can be retrieved by id or groupname
    """
    def __init__(self):
        self.groups = CaseInsensitiveDict()  # key = id
        self.names = CaseInsensitiveDict()  # key = groupname

    def add(self, info):
        self.groups[info.get('id')] = info
        self.names[info.get('displayname')] = info
        return id

    def get_by_id(self, id):
        return self.groups.get(id, None)

    def get_by_name(self, username):
        return self.names.get(username, None)
Exemple #28
0
class UserStore:
    """
      Simple in-memory store for user info. Attribute names are case-insensitive.
      Users can be retrieved by id or username
    """
    def __init__(self):
        self.users = CaseInsensitiveDict()  # key = id
        self.names = CaseInsensitiveDict()  # key = username

    def add(self, user_info):
        self._add_default_attributes(user_info)
        self.users[user_info.get('id')] = dict(user_info)
        self.names[user_info.get('username')] = dict(user_info)
        return id

    def get_by_id(self, id):
        return self.users.get(id, None)

    def get_by_name(self, username):
        return self.names.get(username, None)

    def update_scopes(self, username, scopes):
        self.names[username]['consented_scopes'] += ' ' + scopes

    def list(self):
        """
          Returns a list of dictionaries representing users.
          password and consented_scopes attributes are not returned
        """
        return [self._copy_user(u[1]) for u in self.users.items()]

    def _copy_user(self, user):
        d = copy.deepcopy(dict(user))
        self._del_default_attributes(d)
        return d

    def _add_default_attributes(self, user_info):
        if 'consented_scopes' not in user_info:
            user_info['consented_scopes'] = ''

    def _del_default_attributes(self, dictionary):
        del dictionary['consented_scopes']
        del dictionary['password']
Exemple #29
0
class Response:
    def __init__(self, request):
        """Response object corresponding to the request object
        """
        self.request = request
        self.status_code = None
        self.content = bytearray()
        self.headers = CaseInsensitiveDict()

    def read_status(self):
        return int(self.request.file.readline().decode('utf-8').split()[1])

    def read_headers(self):
        for line in self.request.file:
            if line == b'\r\n':  # end of headers
                break
            header = line.decode().strip(
            )  # remove leading and trailing white spaces
            key, value = header.split(':', maxsplit=1)
            self.headers[key] = value.strip()

    def read_content(self):
        file = self.request.file
        ## Content-length specified:
        if self.headers.get('Content-length',
                            None):  # Content-length header exists
            self.content = file.read(int(self.headers['Content-length']))
        ## Chunked transfer specified: chunk represented as hexa string + CRLF
        elif self.headers.get('Transfer-Encoding',
                              None) == 'chunked':  # chunked transfer
            while True:
                content_length = int(file.readline().decode('utf-8').strip(),
                                     16)
                if content_length == 0:
                    break
                chunk = file.read(content_length)
                self.content.extend(chunk)
                file.readline()
            file.readline()  # skip CRLF
        ## nothing specified:
        else:
            self.content = file.read()  # read until FIN arrival
    def _get_email_headers_from_part(self, part, charset=None):

        email_headers = list(part.items())

        # TODO: the next 2 ifs can be condensed to use 'or'
        if charset is None:
            charset = part.get_content_charset()

        if charset is None:
            charset = 'utf8'

        if not email_headers:
            return {}
        # Convert the header tuple into a dictionary
        headers = CaseInsensitiveDict()
        try:
            [
                headers.update({x[0]: self._get_string(x[1], charset)})
                for x in email_headers
            ]
        except Exception as e:
            error_code, error_msg = self._get_error_message_from_exception(e)
            err = "Error occurred while converting the header tuple into a dictionary"
            self._base_connector.debug_print("{}. {}. {}".format(
                err, error_code, error_msg))

        # Handle received separately
        try:
            received_headers = [
                self._get_string(x[1], charset) for x in email_headers
                if x[0].lower() == 'received'
            ]
        except Exception as e:
            error_code, error_msg = self._get_error_message_from_exception(e)
            err = "Error occurred while handling the received header tuple separately"
            self._base_connector.debug_print("{}. {}. {}".format(
                err, error_code, error_msg))

        if received_headers:
            headers['Received'] = received_headers

        # handle the subject string, if required add a new key
        subject = headers.get('Subject')

        if subject:
            try:
                headers['decodedSubject'] = str(
                    make_header(decode_header(subject)))
            except Exception:
                headers['decodedSubject'] = self._decode_uni_string(
                    subject, subject)
        return dict(headers)
    def _cache_ratelimit_headers(self,
                                 headers: CaseInsensitiveDict,
                                 resource: str = CORE_RESOURCE) -> dict:
        """Cache rate limit information from response headers.

        :param requests.structures.CaseInsensitiveDict headers:
            Headers from response.
        :param str resource:
            Name of resource to get rate limit for. Either CORE_RESOURCE,
            SEARCH_RESOURCE, or GRAPHQL_RESOURCE.
        :returns dict:
            Dictionary containing remaining rate limit, full rate limit, and
            reset time as POSIX timestamp.  For more information see
            https://developer.github.com/v3/rate_limit/
        """
        if not self._ratelimit_cache:
            self._ratelimit_cache = {}
        if self._has_ratelimit_headers(headers):
            self._ratelimit_cache[resource] = {
                'limit': headers.get(self.RATELIMIT_LIMIT_HEADER),
                'remaining': headers.get(self.RATELIMIT_REMAINING_HEADER),
                'reset': headers.get(self.RATELIMIT_RESET_HEADER)
            }
Exemple #32
0
class UserStore:
    """
      Simple in-memory store for user info. Attribute names are case-insensitive.
      Users can be retrieved by id or username
    """

    def __init__(self):
        self.users = CaseInsensitiveDict()
        self.names = CaseInsensitiveDict()

    def add(self, info):
        self.users[info.get('id')] = info
        self.names[info.get('username')] = info
        return id

    def get_by_id(self, id):
        return self.users.get(id, None)

    def get_by_name(self, username):
        return self.names.get(username, None)

    def update_scopes(self, username, scopes):
        self.users[username]['consented_scopes'] += ' ' + scopes
Exemple #33
0
def _get_conflicting_active_course_enrollments(requests_by_key,
                                               existing_course_enrollments,
                                               program_uuid, course_key):
    """
    Process the list of existing course enrollments together with
    the enrollment request list stored in 'requests_by_key'. Detect
    whether we have conflicting ACTIVE ProgramCourseEnrollment entries.
    When detected, log about it and return the conflicting entry with
    duplicated status.

    Arguments:
        requests_by_key (dict)
        existing_course_enrollments (queryset[ProgramCourseEnrollment]),
        program_uuid (UUID|str),
        course_key (str)

    Returns:
        results (dict) with detected conflict entry, or empty dict.
    """
    conflicted_by_user_key = CaseInsensitiveDict()

    requested_statuses_by_user_key = CaseInsensitiveDict({
        key: request.get('status')
        for key, request in requests_by_key.items()
    })

    for existing_enrollment in existing_course_enrollments:
        external_user_key = existing_enrollment.program_enrollment.external_user_key
        requested_status = requested_statuses_by_user_key.get(
            existing_enrollment.program_enrollment.external_user_key)
        if (requested_status
                and requested_status == ProgramCourseEnrollmentStatuses.ACTIVE
                and existing_enrollment.is_active
                and str(existing_enrollment.program_enrollment.program_uuid) !=
                str(program_uuid)):
            logger.error(
                'Detected conflicting active ProgramCourseEnrollment. This is happening on'
                ' The program_uuid [{}] with course_key [{}] for external_user_key [{}]'
                .format(program_uuid, course_key, external_user_key))
            conflicted_by_user_key[
                external_user_key] = ProgramCourseOpStatuses.CONFLICT
    return conflicted_by_user_key
Exemple #34
0
class Response:
    """The object that each response must be packed into before sending. Same
    reason as the Request object. """
    __slots__ = ['version', 'status', 'headers', 'body']

    def __init__(self):
        self.version = HTTP1_1
        self.status = 200
        self.headers = CID()
        self.body = bytes()

    def __repr__(self):
        return '<Response %s %s>' % (self.status, self.reason)

    @property
    def reason(self):
        """Third argument in response status line"""
        return responses[self.status]

    def unpack(self):
        """Preprocess http message (e.g decode)"""
        if self.headers.get('content-encoding', '') == 'gzip':
            self.body = gzip.decompress(self.body)

    def prepare(self, request: Request) -> bytes:
        """Prepare the response for socket transmission"""
        if 'gzip' in request.headers.get('accept-encoding', ''):
            self.headers['content-encoding'] = 'gzip'
            self.body = gzip.compress(self.body)

        self.headers['content-length'] = str(len(self.body))

        return '{version} {status} {reason}{headers}\r\n\r\n'.format(
            version=self.version,
            status=self.status,
            reason=self.reason,
            headers=''.join(
                '\r\n%s: %s' % (k, v)
                for k, v in self.headers.items())).encode() + self.body
Exemple #35
0
    def prepare_response(self, request, cached):
        """Verify our vary headers match and construct a real urllib3
        HTTPResponse object.
        """
        # Special case the '*' Vary value as it means we cannot actually
        # determine if the cached response is suitable for this request.
        # This case is also handled in the controller code when creating
        # a cache entry, but is left here for backwards compatibility.
        if "*" in cached.get("vary", {}):
            return

        # Ensure that the Vary headers for the cached response match our
        # request
        for header, value in cached.get("vary", {}).items():
            if request.headers.get(header, None) != value:
                return

        body_raw = cached["response"].pop("body")

        headers = CaseInsensitiveDict(data=cached["response"]["headers"])
        if headers.get("transfer-encoding", "") == "chunked":
            headers.pop("transfer-encoding")

        cached["response"]["headers"] = headers

        try:
            body = io.BytesIO(body_raw)
        except TypeError:
            # This can happen if cachecontrol serialized to v1 format (pickle)
            # using Python 2. A Python 2 str(byte string) will be unpickled as
            # a Python 3 str (unicode string), which will cause the above to
            # fail with:
            #
            #     TypeError: 'str' does not support the buffer interface
            body = io.BytesIO(body_raw.encode("utf8"))

        return HTTPResponse(body=body,
                            preload_content=False,
                            **cached["response"])
    def read_resp_record(self, resp_headers, payload):
        len_ = payload.tell()
        payload.seek(0)

        warc_headers = self.parser.parse(payload)
        warc_headers = CaseInsensitiveDict(warc_headers.headers)

        record_type = warc_headers.get('WARC-Type', 'response')

        if record_type == 'response':
            status_headers = self.parser.parse(payload)
        else:
            status_headers = None

        record = ArcWarcRecord('warc', record_type, warc_headers, payload,
                              status_headers, '', len_)

        if record_type == 'response':
            self._set_header_buff(record)

        self.ensure_digest(record)

        return record_type, record
Exemple #37
0
    def prepare_response(self, request, cached):
        """Verify our vary headers match and construct a real urllib3
        HTTPResponse object.
        """
        # Special case the '*' Vary value as it means we cannot actually
        # determine if the cached response is suitable for this request.
        # This case is also handled in the controller code when creating
        # a cache entry, but is left here for backwards compatibility.
        if "*" in cached.get("vary", {}):
            return

        # Ensure that the Vary headers for the cached response match our
        # request
        for header, value in cached.get("vary", {}).items():
            if request.headers.get(header, None) != value:
                return

        body_raw = cached["response"].pop("body")

        headers = CaseInsensitiveDict(data=cached["response"]["headers"])
        if headers.get("transfer-encoding", "") == "chunked":
            headers.pop("transfer-encoding")

        cached["response"]["headers"] = headers

        try:
            body = io.BytesIO(body_raw)
        except TypeError:
            # This can happen if cachecontrol serialized to v1 format (pickle)
            # using Python 2. A Python 2 str(byte string) will be unpickled as
            # a Python 3 str (unicode string), which will cause the above to
            # fail with:
            #
            #     TypeError: 'str' does not support the buffer interface
            body = io.BytesIO(body_raw.encode("utf8"))

        return HTTPResponse(body=body, preload_content=False, **cached["response"])
Exemple #38
0
    def forward(self, method):
        data = self.data_bytes
        path = self.path
        forward_headers = CaseInsensitiveDict(self.headers)

        # force close connection
        connection_header = forward_headers.get('Connection') or ''
        if connection_header.lower() not in ['keep-alive', '']:
            self.close_connection = 1

        client_address = self.client_address[0]
        server_address = ':'.join(map(str, self.server.server_address))

        try:
            # run the actual response forwarding
            response = modify_and_forward(
                method=method,
                path=path,
                data_bytes=data,
                headers=forward_headers,
                forward_base_url=self.proxy.forward_base_url,
                listeners=self._listeners(),
                request_handler=self,
                client_address=client_address,
                server_address=server_address)

            # copy headers and return response
            self.send_response(response.status_code)

            # set content for chunked encoding
            is_chunked = uses_chunked_encoding(response)
            if is_chunked:
                response._content = create_chunked_data(response._content)

            # send headers
            content_length_sent = False
            for header_key, header_value in iteritems(response.headers):
                # filter out certain headers that we don't want to transmit
                if header_key.lower() not in ('transfer-encoding', 'date',
                                              'server'):
                    self.send_header(header_key, header_value)
                    content_length_sent = content_length_sent or header_key.lower(
                    ) == 'content-length'

            # fix content-type header if needed
            if not content_length_sent and not is_chunked:
                self.send_header(
                    'Content-Length',
                    '%s' % len(response.content) if response.content else 0)

            if isinstance(response, LambdaResponse):
                self.send_multi_value_headers(response.multi_value_headers)

            self.end_headers()
            if response.content and len(response.content):
                self.wfile.write(to_bytes(response.content))
        except Exception as e:
            trace = str(traceback.format_exc())
            conn_errors = ('ConnectionRefusedError', 'NewConnectionError',
                           'Connection aborted', 'Unexpected EOF',
                           'Connection reset by peer',
                           'cannot read from timed out object')
            conn_error = any(e in trace for e in conn_errors)
            error_msg = 'Error forwarding request: %s %s' % (e, trace)
            if 'Broken pipe' in trace:
                LOG.warn(
                    'Connection prematurely closed by client (broken pipe).')
            elif not self.proxy.quiet or not conn_error:
                LOG.error(error_msg)
                if is_local_test_mode():
                    # During a test run, we also want to print error messages, because
                    # log messages are delayed until the entire test run is over, and
                    # hence we are missing messages if the test hangs for some reason.
                    print('ERROR: %s' % error_msg)
            self.send_response(502)  # bad gateway
            self.end_headers()
            # force close connection
            self.close_connection = 1
        finally:
            try:
                self.wfile.flush()
            except Exception as e:
                LOG.warning('Unable to flush write file: %s' % e)
Exemple #39
0
class Fossor(object):
    '''
    Engine for running automated investigations using three plugin types: variables, checks, reports. The engine:
    - Collects variables
    - Runs checks using applicable variables
    - Reports noteworthy findings
    '''
    def __init__(self):
        self.log = logging.getLogger(__name__)

        self.plugin_parent_classes = [
            fossor.plugin.Plugin, fossor.variables.variable.Variable,
            fossor.checks.check.Check, fossor.reports.report.Report
        ]

        # Variables
        self.variables = CaseInsensitiveDict()
        self.add_variable('timeout', 600)  # Default timeout

        # Imported Plugins
        self.variable_plugins = set()
        self.check_plugins = set()
        self.report_plugins = set()

        self.add_plugins(
        )  # Adds all plugins located within the fossor module recursively

    def _import_submodules_by_module(self, module) -> set:
        """
        Import all submodules from a module.
        Returns a set of modules
        """
        if '__path__' not in dir(module):
            return set()  # No other submodules to import, this is a file
        result = set()
        for loader, name, is_pkg in pkgutil.iter_modules(path=module.__path__):
            sub_module = importlib.import_module(module.__name__ + '.' + name)
            result.add(sub_module)
            result.update(self._import_submodules_by_module(sub_module))

        return result

    def _import_submodules_by_path(self, path: str) -> set:
        """
        Import all submodules from a path.
        Returns a set of modules
        """
        # Get a list of all the python files
        # Loop over list and group modules

        path = os.path.normpath(path)
        all_file_paths = []
        for root, dirs, files in os.walk(path):
            for file in files:
                all_file_paths.append(os.path.join(root, file))

        python_file_paths = [
            file for file in all_file_paths
            if '.py' == os.path.splitext(file)[1]
        ]

        result = set()
        for filepath in python_file_paths:
            # Create a module_name
            relative_path = filepath.split(path)[-1]
            module_name = os.path.splitext(relative_path)[
                0]  # Remove file extension
            module_name = module_name.lstrip(
                os.path.sep)  # Removing leading slash
            module_name = module_name.replace(os.path.sep,
                                              '.')  # Change / to .
            module_name = f'fossor.local.{module_name}'

            # Import with the new name
            spec = importlib.util.spec_from_file_location(
                module_name, filepath)
            module = importlib.util.module_from_spec(spec)
            spec.loader.exec_module(module)
            result.add(module)

        return result

    def add_plugins(self, source=fossor):
        '''
        Recursively return a dict of plugins (classes) that inherit from the given parent_class) from within that source.
        source accepts either a python module or a filesystem path as a string.
        '''
        if source is None:
            return

        if type(source) == str:
            modules = self._import_submodules_by_path(path=source)
        else:
            modules = self._import_submodules_by_module(source)

        for module in modules:
            for obj_name, obj in module.__dict__.items():
                # Get objects from each module that look like plugins for the Check abstract class
                if not inspect.isclass(
                        obj
                ):  # Must check if a class otherwise the next check with issubclass will get a TypeError
                    continue
                if obj in self.plugin_parent_classes:  # If this is one of the parent classes, skip, we only want the children
                    continue
                if issubclass(obj, fossor.variables.variable.Variable):
                    self.variable_plugins.add(obj)
                elif issubclass(obj, fossor.checks.check.Check):
                    self.check_plugins.add(obj)
                elif issubclass(obj, fossor.reports.report.Report):
                    self.report_plugins.add(obj)
                else:
                    continue
                self.log.debug(f"Adding Plugin ({obj_name}, {obj})")

    def clear_plugins(self):
        self.variable_plugins = set()
        self.check_plugins = set()
        self.report_plugins = set()

    def list_plugins(self):
        variables = list(self.variable_plugins)
        checks = list(self.check_plugins)
        reports = list(self.report_plugins)
        return variables + checks + reports

    def _terminate_process_group(self, process):
        '''Kill a process group leader and its process group'''
        pid = process.pid
        pgid = os.getpgid(pid)
        if pid != pgid:
            raise ValueError(
                f"Trying to terminate a pid that is not a pgid leader. pid: {pid}, pgid: {pgid}."
            )  # This should never happen

        # Try these signals to kill the process group
        for s in (signal.SIGTERM, signal.SIGKILL):
            try:
                children = psutil.Process().children(recursive=True)
            except psutil_exceptions:
                continue  # Process is already dead

            if len(children) > 0:
                try:
                    self.log.warning(
                        f"Issuing signal {s} to process group {pgid} because it has run for too long."
                    )
                    os.killpg(pgid, s)
                    time.sleep(1)
                except psutil_exceptions:
                    pass
                except PermissionError:  # Occurs if process has already exited on MacOS
                    pass
        # Terminate the plugin process (which is likely dead at this point)
        process.terminate()
        process.join()

    def _run_plugins_parallel(self, plugins):
        '''
        Accepts function to run, and a dictionary of plugins to run. Format is name:module
        Returns an output queue which outputs two values: plugin-name, output. Queue ends when plugin-name is EOF.

        '''
        def _run_plugins_parallel_helper(timeout):
            setproctitle.setproctitle('Plugin-Runner')
            processes = []
            queue_lock = mp.Lock()

            for Plugin in plugins:
                process = mp.Process(target=self.run_plugin,
                                     name=Plugin.get_name(),
                                     args=(Plugin, output_queue, queue_lock))
                process.start()
                processes.append(process)

            start_time = time.time()
            should_end = False
            while len(processes) > 0:

                time_spent = time.time() - start_time
                should_end = time_spent > timeout

                try:
                    os.kill(os.getppid(),
                            0)  # Check if the parent process is alive
                except ProcessLookupError:
                    should_end = True

                for process in processes:  # Using list so I can delete while iterating
                    process_is_dead = not process.is_alive(
                    ) and process.exitcode is not None
                    if process_is_dead:
                        process.join()
                        processes.remove(process)
                        continue
                    if should_end:
                        with queue_lock:
                            self.log.error(
                                f"Process {process} for plugin {process.name} ran longer than the timeout period: {timeout}"
                            )
                            if self.variables.get('verbose'):
                                output_queue.put((
                                    process.name,
                                    'Timed out (use --time-out to increase timeout)'
                                ))
                            self._terminate_process_group(process)
                            processes.remove(process)
                time.sleep(.1)

            with queue_lock:
                output_queue.put(('Stats', f'Ran {len(plugins)} plugins.'))
                output_queue.put(
                    ('EOF', 'EOF'))  # Indicate the queue is finished

            return output_queue

        timeout = int(self.variables.get('timeout'))

        # Run plugins in parallel
        # This child process will spawn child processes for each plugin, and then do the final termination and clean-up.
        # A separate child process for spawning plugin processes is needed because only a parent process can terminate its children.
        # This allows us to continue to the reporting phase while the checks finish.
        output_queue = mp.Queue()
        parallel_plugins_helper_process = mp.Process(
            target=_run_plugins_parallel_helper,
            name='parallel_plugins_helper_process',
            args=(timeout, ))
        parallel_plugins_helper_process.start()

        return output_queue

    def run_plugin(self, Plugin, output_queue, queue_lock):
        try:
            os.setsid(
            )  # Causes this plugin to become a process group leader, we can then kill that process group to kill all potential subprocesses
            p = Plugin()
            setproctitle.setproctitle(p.get_name())
            output = ''
            if p.should_run():
                output = p.run_helper(variables=self.variables)
                if output:
                    with queue_lock:  # Prevents the queue from becoming corrupt on process termination
                        output_queue.put((p.get_name(), output))
        except Exception as e:
            # This ensures that any plugins that fail to initialize show up properly in the report and don't output errors mid report.
            self.log.exception(f"Plugin {p} failed to initialize", e)

    def get_variables(self):
        '''
        Gather all possible variables that may be useful for plugins
            - Variables can have interdependencies, this will continue gathering missing variables
              until no new variables have appeared after a run
            - Can be run again if new variables are added to fill in any blanks
        '''
        while True:
            variable_count = len(self.variables)
            output_queue = self._run_plugins_parallel(self.variable_plugins)
            while True:
                name, output = output_queue.get()
                if 'EOF' in name:
                    break
                if name in self.variables:
                    self.log.debug(
                        "Skipping variable plugin because it has already been found: {vp}"
                        .format(vp=name))
                    continue
                self.add_variable(name, output)
            output_queue.close()

            if variable_count == len(self.variables):
                self.log.debug("Done finding variables: {variables}".format(
                    variables=self.variables))
                break

    def _convert_simple_type(self, value):
        '''Convert String to simple type if possible'''
        if type(value) == str:
            if 'false' == value.lower():
                return False
            if 'true' == value.lower():
                return True
            try:
                return int(value)
            except ValueError:
                pass
            try:
                return float(value)
            except ValueError:
                pass

        return value

    def add_variables(self, **kwargs):
        for name, value in kwargs.items():
            self.add_variable(name=name, value=value)
        return True

    def add_variable(self, name, value):
        '''Adds variable, converts numbers and booleans to their types'''
        if name in self.variables:
            old_value = self.variables[name]
            if value != old_value:
                self.log.warning(
                    f"Variable {name} with value of {old_value} is being replaced with {value}"
                )

        value = self._convert_simple_type(value)

        self.variables[name] = value
        self.log.debug(
            "Added variable {name} with value of {value} (type={type})".format(
                name=name, value=value, type=type(value)))
        return True

    def _process_whitelist(self, whitelist):
        if whitelist:
            whitelist = [name.casefold() for name in whitelist]
            self.variable_plugins = {
                plugin
                for plugin in self.variable_plugins
                if plugin.get_name().casefold() in whitelist
            }
            self.check_plugins = {
                plugin
                for plugin in self.check_plugins
                if plugin.get_name().casefold() in whitelist
            }

    def _process_blacklist(self, blacklist):
        if blacklist:
            blacklist = [name.casefold() for name in blacklist]
            self.variable_plugins = {
                plugin
                for plugin in self.variable_plugins
                if plugin.get_name().casefold() not in blacklist
            }
            self.check_plugins = {
                plugin
                for plugin in self.check_plugins
                if plugin.get_name().casefold() not in blacklist
            }

    def run(self, report='StdOut', **kwargs):
        '''Runs Fossor with the given report. Method returns a string of the report output.'''
        self.log.debug("Starting Fossor")

        self._process_whitelist(kwargs.get('whitelist'))
        self._process_blacklist(kwargs.get('blacklist'))

        # Add kwargs as variables
        self.add_variables(**kwargs)

        # Add directory plugins
        self.add_plugins(kwargs.get('plugin_dir'))

        # Gather Variables
        self.get_variables()

        # Run checks
        output_queue = self._run_plugins_parallel(self.check_plugins)

        # Run Report
        try:
            Report_Plugin = [
                plugin for plugin in self.report_plugins
                if report.lower() == plugin.get_name().lower()
            ].pop()
        except IndexError:
            message = f"Report_Plugin: {report}, was not found. Possibly reports are: {[plugin.get_name() for plugin in self.report_plugins]}"
            self.log.critical(message)
            raise ValueError(message)

        report_plugin = Report_Plugin()
        return report_plugin.run(variables=self.variables,
                                 report_input=output_queue)
    def forward(self, method):
        data = self.data_bytes
        forward_headers = CaseInsensitiveDict(self.headers)

        # force close connection
        if forward_headers.get('Connection', '').lower() != 'keep-alive':
            self.close_connection = 1

        path = self.path
        if '://' in path:
            path = '/' + path.split('://', 1)[1].split('/', 1)[1]
        forward_url = self.proxy.forward_url
        for listener in self._listeners():
            if listener:
                forward_url = listener.get_forward_url(
                    method, path, data, forward_headers) or forward_url

        proxy_url = '%s%s' % (forward_url, path)
        target_url = self.path
        if '://' not in target_url:
            target_url = '%s%s' % (forward_url, target_url)

        # update original "Host" header (moto s3 relies on this behavior)
        if not forward_headers.get('Host'):
            forward_headers['host'] = urlparse(target_url).netloc
        if 'localhost.atlassian.io' in forward_headers.get('Host'):
            forward_headers['host'] = 'localhost'
        forward_headers['X-Forwarded-For'] = self.build_x_forwarded_for(
            forward_headers)

        try:
            response = None
            modified_request = None
            # update listener (pre-invocation)
            for listener in self._listeners():
                if not listener:
                    continue
                listener_result = listener.forward_request(
                    method=method,
                    path=path,
                    data=data,
                    headers=forward_headers)
                if isinstance(listener_result, Response):
                    response = listener_result
                    break
                if isinstance(listener_result, dict):
                    response = Response()
                    response._content = json.dumps(listener_result)
                    response.status_code = 200
                    break
                elif isinstance(listener_result, Request):
                    modified_request = listener_result
                    data = modified_request.data
                    forward_headers = modified_request.headers
                    break
                elif listener_result is not True:
                    # get status code from response, or use Bad Gateway status code
                    code = listener_result if isinstance(listener_result,
                                                         int) else 503
                    self.send_response(code)
                    self.send_header('Content-Length', '0')
                    # allow pre-flight CORS headers by default
                    self._send_cors_headers()
                    self.end_headers()
                    return
            # perform the actual invocation of the backend service
            if response is None:
                forward_headers['Connection'] = forward_headers.get(
                    'Connection') or 'close'
                data_to_send = self.data_bytes
                request_url = proxy_url
                if modified_request:
                    if modified_request.url:
                        request_url = '%s%s' % (forward_url,
                                                modified_request.url)
                    data_to_send = modified_request.data

                response = self.method(request_url,
                                       data=data_to_send,
                                       headers=forward_headers,
                                       stream=True)

                # prevent requests from processing response body
                if not response._content_consumed and response.raw:
                    response._content = response.raw.read()
            # update listener (post-invocation)
            if self.proxy.update_listener:
                kwargs = {
                    'method': method,
                    'path': path,
                    'data': data,
                    'headers': forward_headers,
                    'response': response
                }
                if 'request_handler' in inspect.getargspec(
                        self.proxy.update_listener.return_response)[0]:
                    # some listeners (e.g., sqs_listener.py) require additional details like the original
                    # request port, hence we pass in a reference to this request handler as well.
                    kwargs['request_handler'] = self
                updated_response = self.proxy.update_listener.return_response(
                    **kwargs)
                if isinstance(updated_response, Response):
                    response = updated_response

            # copy headers and return response
            self.send_response(response.status_code)

            content_length_sent = False
            for header_key, header_value in iteritems(response.headers):
                # filter out certain headers that we don't want to transmit
                if header_key.lower() not in ('transfer-encoding', 'date',
                                              'server'):
                    self.send_header(header_key, header_value)
                    content_length_sent = content_length_sent or header_key.lower(
                    ) == 'content-length'
            if not content_length_sent:
                self.send_header(
                    'Content-Length',
                    '%s' % len(response.content) if response.content else 0)

            # allow pre-flight CORS headers by default
            self._send_cors_headers(response)

            self.end_headers()
            if response.content and len(response.content):
                self.wfile.write(to_bytes(response.content))
        except Exception as e:
            trace = str(traceback.format_exc())
            conn_errors = ('ConnectionRefusedError', 'NewConnectionError',
                           'Connection aborted', 'Unexpected EOF',
                           'Connection reset by peer')
            conn_error = any(e in trace for e in conn_errors)
            error_msg = 'Error forwarding request: %s %s' % (e, trace)
            if 'Broken pipe' in trace:
                LOG.warn(
                    'Connection prematurely closed by client (broken pipe).')
            elif not self.proxy.quiet or not conn_error:
                LOG.error(error_msg)
                if os.environ.get(ENV_INTERNAL_TEST_RUN):
                    # During a test run, we also want to print error messages, because
                    # log messages are delayed until the entire test run is over, and
                    # hence we are missing messages if the test hangs for some reason.
                    print('ERROR: %s' % error_msg)
            self.send_response(502)  # bad gateway
            self.end_headers()
            # force close connection
            self.close_connection = 1
        finally:
            try:
                self.wfile.flush()
            except Exception as e:
                LOG.warning('Unable to flush write file: %s' % e)
Exemple #41
0
    def post(self):
        form = RequestDataForm()
        data = form.data.data
        lines = data.splitlines(True)
        if len(lines) < 3:
            return 'data less 3 lines'

        origin_headers = []
        body = []
        body_start = False

        for index, line in enumerate(lines):
            if index == 0:
                method, path, _ = line.split(' ')
                continue

            if not line.split():
                body_start = True
                continue

            if body_start:
                body.append(line)
            else:
                line = line.strip()
                key, value = line.split(': ', 1)
                origin_headers.append((key, value))

        # for get header value
        header_dict = CaseInsensitiveDict(origin_headers)

        method = method.lower()
        body = ''.join(body)
        content_type = header_dict.get('Content-Type', '')

        # set headers
        headers = []
        origin_host = header_dict.get('Host')
        if form.host.data and origin_host and form.host.data != origin_host:
            headers.append(('Host', header_dict.get('Host')))
        user_agent = header_dict.get('User-Agent')
        referer = header_dict.get('Referer')
        if user_agent:
            headers.append(('User-Agent', user_agent))
        if referer:
            headers.append(('Referer', referer))

        # set cookie
        cookies = []
        cookie = header_dict.get('Cookie')
        C = SimpleCookie(cookie)
        for morsel in C.values():
            cookies.append((morsel.key, morsel.coded_value))

        host = form.host.data or header_dict.get('Host')
        p = urlsplit(path)
        url = urljoin('http://{}'.format(host), p.path)
        params = [(x, repr_value(y)) for x, y in parse_qsl(p.query)]

        if not content_type:
            pass
        elif 'x-www-form-urlencoded' in content_type:
            body = [(x, repr_value(y)) for x, y in parse_qsl(body)]
        elif 'json' in content_type:
            body = [(x, repr_value(y)) for x, y in json.loads(body).items()]
        else:
            headers.append(('Content-Type', content_type))

        code = render_template(
            'code.html',
            method=method,
            url=url,
            params=params,
            body=body,
            headers=headers,
            cookies=cookies,
            content_type=content_type
        )
        return render_template('code-page.html', code=code)
Exemple #42
0
    def match(self, request):
        """Does this URI match the request?"""
        # Match HTTP verb - GET, POST, PUT, DELETE, etc.
        if self.method is not None:
            if request.command.lower() != self.method.lower():
                return False

        # Match path
        if self.path != urlparse.urlparse(request.path).path:
            return False

        # Match querystring
        if request.querystring() != self.querystring:
            return False

        # Match headers
        if self.headers is not None:
            for header_var, header_value in self.headers.items():
                request_headers = CaseInsensitiveDict(request.headers)
                if request_headers.get(header_var) is None:
                    return False

                req_maintext, req_pdict = cgi.parse_header(request_headers.get(header_var))
                mock_maintext, mock_pdict = cgi.parse_header(header_value)

                if "boundary" in req_pdict and "boundary" in mock_pdict:
                    req_pdict['boundary'] = "xxx"
                    mock_pdict['boundary'] = "xxx"

                if req_maintext != mock_maintext:
                    return False

                if mock_pdict != {}:
                    if req_pdict != mock_pdict:
                        return False

        # Match processed request data
        if self.request_data is not None:
            # Check if exact match before parsing
            if request.body != self.request_data:
                if self.request_content_type.startswith("application/json"):
                    if json.loads(request.body) != json.loads(self.request_data):
                        return False
                elif self.request_content_type.startswith("application/x-www-form-urlencoded"):
                    if parse_qsl_as_dict(request.body) != parse_qsl_as_dict(self.request_data):
                        return False
                elif self.request_content_type.startswith('application/xml'):
                    actual_request_data_root = etree.fromstring(request.body)
                    mock_request_data_root = etree.fromstring(self.request_data)

                    if not xml_elements_equal(actual_request_data_root, mock_request_data_root):
                        return False
                elif self.request_content_type.startswith("multipart/form-data"):
                    ctype, pdict = cgi.parse_header(request.headers.get('Content-Type'))
                    req_multipart = cgi.parse_multipart(
                        io.BytesIO(request.body.encode('utf8')),
                        {x: y.encode('utf8') for x, y in pdict.items()}
                    )

                    ctype, pdict = cgi.parse_header(self.headers.get('Content-Type'))
                    mock_multipart = cgi.parse_multipart(
                        io.BytesIO(self.request_data.encode('utf8')),
                        {x: y.encode('utf8') for x, y in pdict.items()}
                    )
                    return mock_multipart == req_multipart
                else:
                    if request.body != self.request_data:
                        return False

        return True
Exemple #43
0
    def forward(self, method):
        path = self.path
        if '://' in path:
            path = '/' + path.split('://', 1)[1].split('/', 1)[1]
        proxy_url = '%s%s' % (self.proxy.forward_url, path)
        target_url = self.path
        if '://' not in target_url:
            target_url = '%s%s' % (self.proxy.forward_url, target_url)
        data = self.data_bytes

        forward_headers = CaseInsensitiveDict(self.headers)
        # update original "Host" header (moto s3 relies on this behavior)
        if not forward_headers.get('Host'):
            forward_headers['host'] = urlparse(target_url).netloc
        if 'localhost.atlassian.io' in forward_headers.get('Host'):
            forward_headers['host'] = 'localhost'

        try:
            response = None
            modified_request = None
            # update listener (pre-invocation)
            if self.proxy.update_listener:
                listener_result = self.proxy.update_listener.forward_request(method=method,
                    path=path, data=data, headers=forward_headers)
                if isinstance(listener_result, Response):
                    response = listener_result
                elif isinstance(listener_result, Request):
                    modified_request = listener_result
                    data = modified_request.data
                    forward_headers = modified_request.headers
                elif listener_result is not True:
                    # get status code from response, or use Bad Gateway status code
                    code = listener_result if isinstance(listener_result, int) else 503
                    self.send_response(code)
                    self.end_headers()
                    return
            # perform the actual invocation of the backend service
            if response is None:
                if modified_request:
                    response = self.method(proxy_url, data=modified_request.data,
                        headers=modified_request.headers)
                else:
                    response = self.method(proxy_url, data=self.data_bytes,
                        headers=forward_headers)
            # update listener (post-invocation)
            if self.proxy.update_listener:
                kwargs = {
                    'method': method,
                    'path': path,
                    'data': data,
                    'headers': forward_headers,
                    'response': response
                }
                if 'request_handler' in inspect.getargspec(self.proxy.update_listener.return_response)[0]:
                    # some listeners (e.g., sqs_listener.py) require additional details like the original
                    # request port, hence we pass in a reference to this request handler as well.
                    kwargs['request_handler'] = self
                updated_response = self.proxy.update_listener.return_response(**kwargs)
                if isinstance(updated_response, Response):
                    response = updated_response

            # copy headers and return response
            self.send_response(response.status_code)

            content_length_sent = False
            for header_key, header_value in iteritems(response.headers):
                # filter out certain headers that we don't want to transmit
                if header_key.lower() not in ('transfer-encoding', 'date', 'server'):
                    self.send_header(header_key, header_value)
                    content_length_sent = content_length_sent or header_key.lower() == 'content-length'
            if not content_length_sent:
                self.send_header('Content-Length', '%s' % len(response.content) if response.content else 0)

            # allow pre-flight CORS headers by default
            if 'Access-Control-Allow-Origin' not in response.headers:
                self.send_header('Access-Control-Allow-Origin', '*')
            if 'Access-Control-Allow-Methods' not in response.headers:
                self.send_header('Access-Control-Allow-Methods', ','.join(CORS_ALLOWED_METHODS))
            if 'Access-Control-Allow-Headers' not in response.headers:
                self.send_header('Access-Control-Allow-Headers', ','.join(CORS_ALLOWED_HEADERS))

            self.end_headers()
            if response.content and len(response.content):
                self.wfile.write(to_bytes(response.content))
            self.wfile.flush()
        except Exception as e:
            trace = str(traceback.format_exc())
            conn_errors = ('ConnectionRefusedError', 'NewConnectionError')
            conn_error = any(e in trace for e in conn_errors)
            error_msg = 'Error forwarding request: %s %s' % (e, trace)
            if 'Broken pipe' in trace:
                LOGGER.warn('Connection prematurely closed by client (broken pipe).')
            elif not self.proxy.quiet or not conn_error:
                LOGGER.error(error_msg)
                if os.environ.get(ENV_INTERNAL_TEST_RUN):
                    # During a test run, we also want to print error messages, because
                    # log messages are delayed until the entire test run is over, and
                    # hence we are missing messages if the test hangs for some reason.
                    print('ERROR: %s' % error_msg)
            self.send_response(502)  # bad gateway
            self.end_headers()
def proxy(identifier, in_method, in_headers, data):
    # find endpoint
    endpoint = pg.select1(
        'endpoints',
        what=['definition', 'method', 'pass_headers',
              'headers', 'url', 'url_dynamic'],
        where={'id': identifier}
    )
    if not endpoint:
        return 'endpoint not found, create it at ' \
               '<a href="/dashboard">dashboard</a>', 404, {}

    event = {
        'in': {
            'time': time.time(),
            'method': in_method,
            'headers': dict(in_headers),
            'body': data,
            'replay': True
        },
        'out': {
            'method': endpoint['method'],
            'url': endpoint['url'],
            'body': None,
            'headers': {}
        },
        'response': {
            'code': 0,
            'body': ''
        }
    }

    mutated, error = jq(endpoint['definition'], data=data)
    if not mutated or error:
        event['out']['error'] = error.decode('utf-8')
        publish(identifier, event)
        return 'transmutated into null and aborted', 201, {}

    h = CaseInsensitiveDict({'Content-Type': 'application/json'})
    if endpoint['pass_headers']:
        h.update(in_headers)
    h.update(endpoint['headers'])
    event['out']['headers'] = dict(h)

    # reformat the mutated data
    mutatedjson = json.loads(mutated.decode('utf-8'))
    if h.get('content-type') == 'application/x-www-form-urlencoded':
        # oops, not json
        mutated = urlencode(mutatedjson)
    else:
        mutated = json.dumps(mutatedjson)

    event['out']['body'] = mutated

    if endpoint['url_dynamic']:
        urlb, error = jq(endpoint['url'], data=data)
        print('URL', urlb, 'ERROR', error)
        if not urlb:
            event['out']['url_error'] = error.decode('utf-8')
            publish(identifier, event)
            return 'url building has failed', 200, {}
        url = urlb.decode('utf-8')
        event['out']['url'] = url
    else:
        url = endpoint['url']

    # event['out'] is completed at this point
    # and we all have everything needed to perform the request

    if url and is_valid_url(url):
        # we will only perform a request if there is an URL and it is valid
        try:
            s = requests.Session()
            req = requests.Request(endpoint['method'], url,
                                   data=mutated, headers=h).prepare()
            resp = s.send(req, timeout=4)

            if not resp.ok:
                print('FAILED TO POST', resp.text, identifier, mutated)

        except requests.exceptions.RequestException as e:
            print(identifier, 'FAILED TO POST', mutated, 'TO URL', url)
            print(e)
            publish(identifier, event)
            return "<request failed: '%s'>" % e, 503, {}

        event['response']['code'] = resp.status_code
        event['response']['body'] = resp.text
        publish(identifier, event)
        return resp.text, resp.status_code, dict(resp.headers)
    else:
        # no valid URL, just testing
        publish(identifier, event)
        return 'no URL to send this to', 201, {}
Exemple #45
0
    def cache_response(self, request, response, body=None, status_codes=None):
        """
        Algorithm for caching requests.

        This assumes a requests Response object.
        """
        # From httplib2: Don't cache 206's since we aren't going to
        #                handle byte range requests
        cacheable_status_codes = status_codes or self.cacheable_status_codes
        if response.status not in cacheable_status_codes:
            logger.debug(
                "Status code %s not in %s", response.status, cacheable_status_codes
            )
            return

        response_headers = CaseInsensitiveDict(response.headers)

        # If we've been given a body, our response has a Content-Length, that
        # Content-Length is valid then we can check to see if the body we've
        # been given matches the expected size, and if it doesn't we'll just
        # skip trying to cache it.
        if (
            body is not None
            and "content-length" in response_headers
            and response_headers["content-length"].isdigit()
            and int(response_headers["content-length"]) != len(body)
        ):
            return

        cc_req = self.parse_cache_control(request.headers)
        cc = self.parse_cache_control(response_headers)

        cache_url = self.cache_url(request.url)
        logger.debug('Updating cache with response from "%s"', cache_url)

        # Delete it from the cache if we happen to have it stored there
        no_store = False
        if "no-store" in cc:
            no_store = True
            logger.debug('Response header has "no-store"')
        if "no-store" in cc_req:
            no_store = True
            logger.debug('Request header has "no-store"')
        if no_store and self.cache.get(cache_url):
            logger.debug('Purging existing cache entry to honor "no-store"')
            self.cache.delete(cache_url)
        if no_store:
            return

        # https://tools.ietf.org/html/rfc7234#section-4.1:
        # A Vary header field-value of "*" always fails to match.
        # Storing such a response leads to a deserialization warning
        # during cache lookup and is not allowed to ever be served,
        # so storing it can be avoided.
        if "*" in response_headers.get("vary", ""):
            logger.debug('Response header has "Vary: *"')
            return

        # If we've been given an etag, then keep the response
        if self.cache_etags and "etag" in response_headers:
            logger.debug("Caching due to etag")
            self.cache.set(
                cache_url, self.serializer.dumps(request, response, body=body)
            )

        # Add to the cache any 301s. We do this before looking that
        # the Date headers.
        elif response.status == 301:
            logger.debug("Caching permanant redirect")
            self.cache.set(cache_url, self.serializer.dumps(request, response))

        # Add to the cache if the response headers demand it. If there
        # is no date header then we can't do anything about expiring
        # the cache.
        elif "date" in response_headers:
            # cache when there is a max-age > 0
            if "max-age" in cc and cc["max-age"] > 0:
                logger.debug("Caching b/c date exists and max-age > 0")
                self.cache.set(
                    cache_url, self.serializer.dumps(request, response, body=body)
                )

            # If the request can expire, it means we should cache it
            # in the meantime.
            elif "expires" in response_headers:
                if response_headers["expires"]:
                    logger.debug("Caching b/c of expires header")
                    self.cache.set(
                        cache_url, self.serializer.dumps(request, response, body=body)
                    )
Exemple #46
0
class MessageParser(object):
    def __init__(self, name, content, init_headers={}):
        self.name = name
        self.content = content

        self.headers = CaseInsensitiveDict(init_headers)
        self.html_stats = HTMLStats()
        self.email_stats = EmailStats()

        if '\n\n' in self.content:
            self.head, self.body = self.content.split('\n\n', 1)
        else:
            self.head, self.body = "", ""
        self._parse_headers()
        self._decode_body()
        self._count_links()

    @property
    def subject(self):
        return self.headers['Subject']

    @property
    def mime_type(self):
        if 'Content-Type' in self.headers:
            content_type = self.headers['Content-Type']
            mime_type = content_type.split(';', 1)[0]
            mime_type = mime_type.strip().lower()
            return mime_type

    @property
    def charset(self):
        if 'Content-Type' in self.headers:
            content_type = self.headers['Content-Type']
            rem = RE_CHARSET.search(content_type)
            if rem:
                charset = rem.group(1).lower()
                return charset

    @property
    def is_multipart(self):
        if self.mime_type and self.mime_type.startswith('multipart/'):
            return True
        else:
            return False

    def _get_boundary(self):
        if 'Content-Type' in self.headers:
            content_type = self.headers['Content-Type']
            rem = RE_HEADER_BOUNDARY.search(content_type)
            if rem:
                return rem.group(1)

    def _parse_headers(self):
        lines = self.head.split('\n')

        # Drop the "From..." line
        while lines and not RE_HEADER.match(lines[0]):
            del lines[0]
        while lines:
            line = self._get_header_line(lines)
            key, value = line.split(':', 1)
            key, value = key.strip(), value.strip()
            self.headers[key] = value

    def _get_header_line(self, lines):
        line_parts = [lines.pop(0)]
        while lines and (lines[0].startswith('\t') or
                         lines[0].startswith(' ')):
            line_parts.append(lines.pop(0))

        line = " ".join([part.strip() for part in line_parts])
        return line

    def _decode_body(self):
        if self.mime_type and (self.mime_type.startswith('image/') or
                               self.mime_type.startswith('application/')):
            LOG.info("Body marked as image, skipping body")
            self.email_stats['attached_images'] += 1
            self.body = ""
            return

        if self.is_multipart:
            LOG.debug("Detected multipart/* content-type")
            self._decode_multipart_body()
        else:
            self._decode_single_body()

    def _decode_single_body(self):
        self.body = self.body.strip()
        cte = self.headers.get('Content-Transfer-Encoding', '').lower()
        if 'quoted-printable' in cte:
            LOG.debug("Detected quoted-printable encoding, decoding")
            self.body = quopri.decodestring(self.body)
        if 'base64' in cte:
            LOG.debug("Detected base64 encoding, decoding")
            try:
                self.body = base64.decodestring(self.body)
            except base64.binascii.Error:
                LOG.info("base64 decoder failed, trying partial decoding")
                self.body = base64_partial_decode(self.body)

        LOG.debug("Detected charset: %s", self.charset)
        try:
            self.body = self.body.decode(
                validate_charset(self.charset) and self.charset or 'ascii',
                'strict'
            )
        except UnicodeDecodeError:
            LOG.info('Error during strict decoding')
            self.email_stats['charset_errors'] = 1
            self.body = self.body.decode(
                validate_charset(self.charset) and self.charset or 'ascii',
                'ignore'
            )

        if self._guess_html():
            LOG.debug("Message recognized as HTML")
            self._parse_html()
        else:
            LOG.debug("Message recognized as plaintext")

    def _decode_multipart_body(self):
        boundary = self._get_boundary()
        if not boundary:
            LOG.warn("Message detected as multipart but boundary "
                     "declaration was not found")
            return

        start_bnd = '\n' + '--' + boundary + '\n'
        end_bnd = '\n' + '--' + boundary + '--' + '\n'

        self.body = '\n' + self.body  # for string matching purpose

        try:
            start_index = self.body.index(start_bnd) + len(start_bnd)
        except ValueError:
            LOG.warn("Cannot find boundaries in body, "
                     "treating as single message")
            self._decode_single_body()
            return

        end_index = self.body.rfind(end_bnd)
        if end_index < 0:
            end_index = None

        content = self.body[start_index:end_index]

        parts = content.split(start_bnd)

        messages = [MessageParser(self.name, msg_content, self.headers)
                    for msg_content in parts]
        self.body = "\n".join([msg.body for msg in messages])
        for msg in messages:
            self.email_stats += msg.email_stats
            self.html_stats += msg.html_stats

    def _guess_html(self):
        if self.mime_type and self.mime_type == 'text/html':
            return True
        else:
            return False

    def _parse_html(self):
        parser = StatsHTMLParser()
        parser.feed(self.body)
        parser.close()
        self.body = parser.text
        self.html_stats = parser.stats

    def _count_links(self):
        mail_links = RE_MAIL_LINK.findall(self.body)
        self.body = RE_MAIL_LINK.sub('', self.body)
        self.email_stats['mail_links'] = len(mail_links)

        http_links = RE_HTTP_LINK.findall(self.body)
        self.body = RE_HTTP_LINK.sub('', self.body)
        self.email_stats['http_links'] = len(http_links)

        http_raw_links = [link for link in http_links
                          if RE_HTTP_RAW_LINK.match(link)]
        self.email_stats['http_links'] = len(http_raw_links)

    def as_series(self):
        body_length = len(self.body)
        s = pd.Series({'plain_body': self.body, 'body_length': body_length})

        rel_email_stats = self.email_stats.astype('float32')
        rel_email_stats.index = ['rel_' + idx for idx in
                                rel_email_stats.index]
        if body_length > 0:
            rel_email_stats /= body_length
        else:
            rel_email_stats[:] = 0

        rel_html_stats = self.html_stats.astype('float32')
        rel_html_stats.index = ['rel_' + idx for idx in
                                rel_html_stats.index]
        if body_length > 0:
            rel_html_stats /= body_length
        else:
            rel_html_stats[:] = 0

        s_headers = pd.Series(dict(self.headers))
        s_headers.index = s_headers.index.map(lambda h: 'h_' + h.lower())
        s = (s
             .append(s_headers, True)
             .append(self.email_stats, True)
             .append(self.html_stats, True)
             .append(rel_email_stats, True)
             .append(rel_html_stats, True)
             )
        s.name = self.name
        s = s.fillna(0)
        return s
Exemple #47
0
class MockRestURI(object):
    """Representation of a mock URI."""
    def __init__(self, uri_dict):
        self._uri_dict = uri_dict
        self.name = uri_dict.get('name', None)
        self.fullpath = uri_dict['request']['path']
        self._regexp = False

        self.path = urlparse.urlparse(self.fullpath).path
        self.querystring = urlparse.parse_qs(
            urlparse.urlparse(self.fullpath).query,
            keep_blank_values=True
        )

        self.method = uri_dict['request'].get('method', None)
        self.headers = CaseInsensitiveDict(uri_dict['request'].get('headers', {}))
        self.return_code = int(uri_dict['response'].get('code', '200'))
        self.request_content_type = self.headers.get('Content-Type', '')
        self.response_location = uri_dict['response'].get('location', None)
        self.response_content = uri_dict['response'].get('content', "")
        self.wait = float(uri_dict['response'].get('wait', 0.0))
        self.request_data = uri_dict['request'].get('data', None)
        self.encoding = uri_dict['request'].get("encoding", None)
        self.response_headers = uri_dict['response'].get("headers", {})

    def match(self, request):
        """Does this URI match the request?"""
        # Match HTTP verb - GET, POST, PUT, DELETE, etc.
        if self.method is not None:
            if request.command.lower() != self.method.lower():
                return False

        # Match path
        if self.path != urlparse.urlparse(request.path).path:
            return False

        # Match querystring
        if request.querystring() != self.querystring:
            return False

        # Match headers
        if self.headers is not None:
            for header_var, header_value in self.headers.items():
                request_headers = CaseInsensitiveDict(request.headers)
                if request_headers.get(header_var) is None:
                    return False

                req_maintext, req_pdict = cgi.parse_header(request_headers.get(header_var))
                mock_maintext, mock_pdict = cgi.parse_header(header_value)

                if "boundary" in req_pdict and "boundary" in mock_pdict:
                    req_pdict['boundary'] = "xxx"
                    mock_pdict['boundary'] = "xxx"

                if req_maintext != mock_maintext:
                    return False

                if mock_pdict != {}:
                    if req_pdict != mock_pdict:
                        return False

        # Match processed request data
        if self.request_data is not None:
            # Check if exact match before parsing
            if request.body != self.request_data:
                if self.request_content_type.startswith("application/json"):
                    if json.loads(request.body) != json.loads(self.request_data):
                        return False
                elif self.request_content_type.startswith("application/x-www-form-urlencoded"):
                    if parse_qsl_as_dict(request.body) != parse_qsl_as_dict(self.request_data):
                        return False
                elif self.request_content_type.startswith('application/xml'):
                    actual_request_data_root = etree.fromstring(request.body)
                    mock_request_data_root = etree.fromstring(self.request_data)

                    if not xml_elements_equal(actual_request_data_root, mock_request_data_root):
                        return False
                elif self.request_content_type.startswith("multipart/form-data"):
                    ctype, pdict = cgi.parse_header(request.headers.get('Content-Type'))
                    req_multipart = cgi.parse_multipart(
                        io.BytesIO(request.body.encode('utf8')),
                        {x: y.encode('utf8') for x, y in pdict.items()}
                    )

                    ctype, pdict = cgi.parse_header(self.headers.get('Content-Type'))
                    mock_multipart = cgi.parse_multipart(
                        io.BytesIO(self.request_data.encode('utf8')),
                        {x: y.encode('utf8') for x, y in pdict.items()}
                    )
                    return mock_multipart == req_multipart
                else:
                    if request.body != self.request_data:
                        return False

        return True

    def querystring_string(self):
        # TODO : Refactor this and docstring.
        query = ''
        for key in self.querystring.keys():
            for item in self.querystring[key]:
                query += str(key) + '=' + item + "&"
        query = query.rstrip("&")
        return "?" + query if query else ""

    def example_path(self):
        return xeger.xeger(self.path) if self._regexp else self.path + self.querystring_string()

    def return_code_description(self):
        return status_codes.CODES.get(self.return_code)[0]

    def request_data_values(self):
        if self.request_data is not None:
            return self.request_data.get('values', {}).iteritems()
        else:
            return []

    def request_data_type(self):
        if self.request_data is not None:
            return self.request_data.get('encoding')
        else:
            return None
class tizgmusicproxy(object):
    """A class for logging into a Google Play Music account and retrieving song
    URLs.

    """

    all_songs_album_title = "All Songs"
    thumbs_up_playlist_name = "Thumbs Up"

    # pylint: disable=too-many-instance-attributes,too-many-public-methods
    def __init__(self, email, password, device_id):
        self.__gmusic = Mobileclient()
        self.__email = email
        self.__device_id = device_id
        self.logged_in = False
        self.queue = list()
        self.queue_index = -1
        self.play_queue_order = list()
        self.play_modes = TizEnumeration(["NORMAL", "SHUFFLE"])
        self.current_play_mode = self.play_modes.NORMAL
        self.now_playing_song = None

        userdir = os.path.expanduser('~')
        tizconfig = os.path.join(userdir, ".config/tizonia/." + email + ".auth_token")
        auth_token = ""
        if os.path.isfile(tizconfig):
            with open(tizconfig, "r") as f:
                auth_token = pickle.load(f)
                if auth_token:
                    # 'Keep track of the auth token' workaround. See:
                    # https://github.com/diraimondo/gmusicproxy/issues/34#issuecomment-147359198
                    print_msg("[Google Play Music] [Authenticating] : " \
                              "'with cached auth token'")
                    self.__gmusic.android_id = device_id
                    self.__gmusic.session._authtoken = auth_token
                    self.__gmusic.session.is_authenticated = True
                    try:
                        self.__gmusic.get_registered_devices()
                    except CallFailure:
                        # The token has expired. Reset the client object
                        print_wrn("[Google Play Music] [Authenticating] : " \
                                  "'auth token expired'")
                        self.__gmusic = Mobileclient()
                        auth_token = ""

        if not auth_token:
            attempts = 0
            print_nfo("[Google Play Music] [Authenticating] : " \
                      "'with user credentials'")
            while not self.logged_in and attempts < 3:
                self.logged_in = self.__gmusic.login(email, password, device_id)
                attempts += 1

            with open(tizconfig, "a+") as f:
                f.truncate()
                pickle.dump(self.__gmusic.session._authtoken, f)

        self.library = CaseInsensitiveDict()
        self.song_map = CaseInsensitiveDict()
        self.playlists = CaseInsensitiveDict()
        self.stations = CaseInsensitiveDict()

    def logout(self):
        """ Reset the session to an unauthenticated, default state.

        """
        self.__gmusic.logout()

    def set_play_mode(self, mode):
        """ Set the playback mode.

        :param mode: curren tvalid values are "NORMAL" and "SHUFFLE"

        """
        self.current_play_mode = getattr(self.play_modes, mode)
        self.__update_play_queue_order()

    def current_song_title_and_artist(self):
        """ Retrieve the current track's title and artist name.

        """
        logging.info("current_song_title_and_artist")
        song = self.now_playing_song
        if song:
            title = to_ascii(song.get('title'))
            artist = to_ascii(song.get('artist'))
            logging.info("Now playing %s by %s", title, artist)
            return artist, title
        else:
            return '', ''

    def current_song_album_and_duration(self):
        """ Retrieve the current track's album and duration.

        """
        logging.info("current_song_album_and_duration")
        song = self.now_playing_song
        if song:
            album = to_ascii(song.get('album'))
            duration = to_ascii \
                       (song.get('durationMillis'))
            logging.info("album %s duration %s", album, duration)
            return album, int(duration)
        else:
            return '', 0

    def current_track_and_album_total(self):
        """Return the current track number and the total number of tracks in the
        album, if known.

        """
        logging.info("current_track_and_album_total")
        song = self.now_playing_song
        track = 0
        total = 0
        if song:
            try:
                track = song['trackNumber']
                total = song['totalTrackCount']
                logging.info("track number %s total tracks %s", track, total)
            except KeyError:
                logging.info("trackNumber or totalTrackCount : not found")
        else:
            logging.info("current_song_track_number_"
                         "and_total_tracks : not found")
        return track, total

    def current_song_year(self):
        """ Return the current track's year of publication.

        """
        logging.info("current_song_year")
        song = self.now_playing_song
        year = 0
        if song:
            try:
                year = song['year']
                logging.info("track year %s", year)
            except KeyError:
                logging.info("year : not found")
        else:
            logging.info("current_song_year : not found")
        return year

    def current_song_genre(self):
        """ Return the current track's genre.

        """
        logging.info("current_song_genre")
        song = self.now_playing_song
        if song:
            genre = to_ascii(song.get('genre'))
            logging.info("genre %s", genre)
            return genre
        else:
            return ''

    def current_song_album_art(self):
        """ Return the current track's album art image.

        """
        logging.info("current_song_art")
        song = self.now_playing_song
        if song:
            artref = song.get('albumArtRef')
            if artref and len(artref) > 0:
                url = to_ascii(artref[0].get('url'))
                logging.info("url %s", url)
                return url
        return ''

    def clear_queue(self):
        """ Clears the playback queue.

        """
        self.queue = list()
        self.queue_index = -1

    def enqueue_tracks(self, arg):
        """ Search the user's library for tracks and add
        them to the playback queue.

        :param arg: a track search term
        """
        try:
            songs = self.__gmusic.get_all_songs()

            track_hits = list ()
            for song in songs:
                song_title = song['title']
                if arg.lower() in song_title.lower():
                    track_hits.append(song)
                    print_nfo("[Google Play Music] [Track] '{0}'." \
                              .format(to_ascii(song_title)))

            if not len(track_hits):
                print_wrn("[Google Play Music] '{0}' not found. "\
                          "Feeling lucky?." \
                          .format(arg.encode('utf-8')))
                random.seed()
                track_hits = random.sample(songs, MAX_TRACKS)
                for hit in track_hits:
                    song_title = hit['title']
                    print_nfo("[Google Play Music] [Track] '{0}'." \
                              .format(to_ascii(song_title)))

            if not len(track_hits):
                raise KeyError

            tracks_added = self.__enqueue_tracks(track_hits)
            logging.info("Added %d tracks from %s to queue", \
                         tracks_added, arg)

            self.__update_play_queue_order()

        except KeyError:
            raise KeyError("Track not found : {0}".format(arg))

    def enqueue_artist(self, arg):
        """ Search the user's library for tracks from the given artist and add
        them to the playback queue.

        :param arg: an artist
        """
        try:
            self.__update_local_library()
            artist = None
            artist_dict = None
            if arg not in self.library.keys():
                for name, art in self.library.iteritems():
                    if arg.lower() in name.lower():
                        artist = name
                        artist_dict = art
                        if arg.lower() != name.lower():
                            print_wrn("[Google Play Music] '{0}' not found. " \
                                      "Playing '{1}' instead." \
                                      .format(arg.encode('utf-8'), \
                                              name.encode('utf-8')))
                        break
                if not artist:
                    # Play some random artist from the library
                    random.seed()
                    artist = random.choice(self.library.keys())
                    artist_dict = self.library[artist]
                    print_wrn("[Google Play Music] '{0}' not found. "\
                              "Feeling lucky?." \
                              .format(arg.encode('utf-8')))
            else:
                artist = arg
                artist_dict = self.library[arg]
            tracks_added = 0
            for album in artist_dict:
                tracks_added += self.__enqueue_tracks(artist_dict[album])
            print_wrn("[Google Play Music] Playing '{0}'." \
                      .format(to_ascii(artist)))
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Artist not found : {0}".format(arg))

    def enqueue_album(self, arg):
        """ Search the user's library for albums with a given name and add
        them to the playback queue.

        """
        try:
            self.__update_local_library()
            album = None
            artist = None
            tentative_album = None
            tentative_artist = None
            for library_artist in self.library:
                for artist_album in self.library[library_artist]:
                    print_nfo("[Google Play Music] [Album] '{0}'." \
                              .format(to_ascii(artist_album)))
                    if not album:
                        if arg.lower() == artist_album.lower():
                            album = artist_album
                            artist = library_artist
                            break
                    if not tentative_album:
                        if arg.lower() in artist_album.lower():
                            tentative_album = artist_album
                            tentative_artist = library_artist
                if album:
                    break

            if not album and tentative_album:
                album = tentative_album
                artist = tentative_artist
                print_wrn("[Google Play Music] '{0}' not found. " \
                          "Playing '{1}' instead." \
                          .format(arg.encode('utf-8'), \
                          album.encode('utf-8')))
            if not album:
                # Play some random album from the library
                random.seed()
                artist = random.choice(self.library.keys())
                album = random.choice(self.library[artist].keys())
                print_wrn("[Google Play Music] '{0}' not found. "\
                          "Feeling lucky?." \
                          .format(arg.encode('utf-8')))

            if not album:
                raise KeyError("Album not found : {0}".format(arg))

            self.__enqueue_tracks(self.library[artist][album])
            print_wrn("[Google Play Music] Playing '{0}'." \
                      .format(to_ascii(album)))
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Album not found : {0}".format(arg))

    def enqueue_playlist(self, arg):
        """Search the user's library for playlists with a given name
        and add the tracks of the first match to the playback queue.

        Requires Unlimited subscription.

        """
        try:
            self.__update_local_library()
            self.__update_playlists()
            self.__update_playlists_unlimited()
            playlist = None
            playlist_name = None
            for name, plist in self.playlists.items():
                print_nfo("[Google Play Music] [Playlist] '{0}'." \
                          .format(to_ascii(name)))
            if arg not in self.playlists.keys():
                for name, plist in self.playlists.iteritems():
                    if arg.lower() in name.lower():
                        playlist = plist
                        playlist_name = name
                        if arg.lower() != name.lower():
                            print_wrn("[Google Play Music] '{0}' not found. " \
                                      "Playing '{1}' instead." \
                                      .format(arg.encode('utf-8'), \
                                              to_ascii(name)))
                            break
            else:
                playlist_name = arg
                playlist = self.playlists[arg]

            random.seed()
            x = 0
            while (not playlist or not len(playlist)) and x < 3:
                x += 1
                # Play some random playlist from the library
                playlist_name = random.choice(self.playlists.keys())
                playlist = self.playlists[playlist_name]
                print_wrn("[Google Play Music] '{0}' not found or found empty. "\
                          "Feeling lucky?." \
                          .format(arg.encode('utf-8')))

            if not len(playlist):
                raise KeyError

            self.__enqueue_tracks(playlist)
            print_wrn("[Google Play Music] Playing '{0}'." \
                      .format(to_ascii(playlist_name)))
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Playlist not found or found empty : {0}".format(arg))

    def enqueue_podcast(self, arg):
        """Search Google Play Music for a podcast series and add its tracks to the
        playback queue ().

        Requires Unlimited subscription.

        """
        print_msg("[Google Play Music] [Retrieving podcasts] : '{0}'. " \
                  .format(self.__email))

        try:

            self.__enqueue_podcast(arg)

            if not len(self.queue):
                raise KeyError

            logging.info("Added %d episodes from '%s' to queue", \
                         len(self.queue), arg)
            self.__update_play_queue_order()

        except KeyError:
            raise KeyError("Podcast not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_station_unlimited(self, arg):
        """Search the user's library for a station with a given name
        and add its tracks to the playback queue.

        Requires Unlimited subscription.

        """
        try:
            # First try to find a suitable station in the user's library
            self.__enqueue_user_station_unlimited(arg)

            if not len(self.queue):
                # If no suitable station is found in the user's library, then
                # search google play unlimited for a potential match.
                self.__enqueue_station_unlimited(arg)

            if not len(self.queue):
                raise KeyError

        except KeyError:
            raise KeyError("Station not found : {0}".format(arg))

    def enqueue_genre_unlimited(self, arg):
        """Search Unlimited for a genre with a given name and add its
        tracks to the playback queue.

        Requires Unlimited subscription.

        """
        print_msg("[Google Play Music] [Retrieving genres] : '{0}'. " \
                  .format(self.__email))

        try:
            all_genres = list()
            root_genres = self.__gmusic.get_genres()
            second_tier_genres = list()
            for root_genre in root_genres:
                second_tier_genres += self.__gmusic.get_genres(root_genre['id'])
            all_genres += root_genres
            all_genres += second_tier_genres
            for genre in all_genres:
                print_nfo("[Google Play Music] [Genre] '{0}'." \
                          .format(to_ascii(genre['name'])))
            genre = dict()
            if arg not in all_genres:
                genre = next((g for g in all_genres \
                              if arg.lower() in to_ascii(g['name']).lower()), \
                             None)

            tracks_added = 0
            while not tracks_added:
                if not genre and len(all_genres):
                    # Play some random genre from the search results
                    random.seed()
                    genre = random.choice(all_genres)
                    print_wrn("[Google Play Music] '{0}' not found. "\
                              "Feeling lucky?." \
                              .format(arg.encode('utf-8')))

                genre_name = genre['name']
                genre_id = genre['id']
                station_id = self.__gmusic.create_station(genre_name, \
                                                          None, None, None, genre_id)
                num_tracks = MAX_TRACKS
                tracks = self.__gmusic.get_station_tracks(station_id, num_tracks)
                tracks_added = self.__enqueue_tracks(tracks)
                logging.info("Added %d tracks from %s to queue", tracks_added, genre_name)
                if not tracks_added:
                    # This will produce another iteration in the loop
                    genre = None

            print_wrn("[Google Play Music] Playing '{0}'." \
                      .format(to_ascii(genre['name'])))
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Genre not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_situation_unlimited(self, arg):
        """Search Unlimited for a situation with a given name and add its
        tracks to the playback queue.

        Requires Unlimited subscription.

        """
        print_msg("[Google Play Music] [Retrieving situations] : '{0}'. " \
                  .format(self.__email))

        try:

            self.__enqueue_situation_unlimited(arg)

            if not len(self.queue):
                raise KeyError

            logging.info("Added %d tracks from %s to queue", \
                         len(self.queue), arg)

        except KeyError:
            raise KeyError("Situation not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_artist_unlimited(self, arg):
        """Search Unlimited for an artist and add the artist's 200 top tracks to the
        playback queue.

        Requires Unlimited subscription.

        """
        try:
            artist = self.__gmusic_search(arg, 'artist')

            include_albums = False
            max_top_tracks = MAX_TRACKS
            max_rel_artist = 0
            artist_tracks = dict()
            if artist:
                artist_tracks = self.__gmusic.get_artist_info \
                                (artist['artist']['artistId'],
                                 include_albums, max_top_tracks,
                                 max_rel_artist)['topTracks']

            if not artist_tracks:
                raise KeyError

            for track in artist_tracks:
                song_title = track['title']
                print_nfo("[Google Play Music] [Track] '{0}'." \
                          .format(to_ascii(song_title)))

            tracks_added = self.__enqueue_tracks(artist_tracks)
            logging.info("Added %d tracks from %s to queue", \
                         tracks_added, arg)
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Artist not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_album_unlimited(self, arg):
        """Search Unlimited for an album and add its tracks to the
        playback queue.

        Requires Unlimited subscription.

        """
        try:
            album = self.__gmusic_search(arg, 'album')
            album_tracks = dict()
            if album:
                album_tracks = self.__gmusic.get_album_info \
                               (album['album']['albumId'])['tracks']
            if not album_tracks:
                raise KeyError

            print_wrn("[Google Play Music] Playing '{0}'." \
                      .format((album['album']['name']).encode('utf-8')))

            tracks_added = self.__enqueue_tracks(album_tracks)
            logging.info("Added %d tracks from %s to queue", \
                         tracks_added, arg)
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Album not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_tracks_unlimited(self, arg):
        """ Search Unlimited for a track name and add all the matching tracks
        to the playback queue.

        Requires Unlimited subscription.

        """
        print_msg("[Google Play Music] [Retrieving library] : '{0}'. " \
                  .format(self.__email))

        try:
            max_results = MAX_TRACKS
            track_hits = self.__gmusic.search(arg, max_results)['song_hits']
            if not len(track_hits):
                # Do another search with an empty string
                track_hits = self.__gmusic.search("", max_results)['song_hits']
                print_wrn("[Google Play Music] '{0}' not found. "\
                          "Feeling lucky?." \
                          .format(arg.encode('utf-8')))

            tracks = list()
            for hit in track_hits:
                tracks.append(hit['track'])
            tracks_added = self.__enqueue_tracks(tracks)
            logging.info("Added %d tracks from %s to queue", \
                         tracks_added, arg)
            self.__update_play_queue_order()
        except KeyError:
            raise KeyError("Playlist not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_playlist_unlimited(self, arg):
        """Search Unlimited for a playlist name and add all its tracks to the
        playback queue.

        Requires Unlimited subscription.

        """
        print_msg("[Google Play Music] [Retrieving playlists] : '{0}'. " \
                  .format(self.__email))

        try:
            playlist_tracks = list()

            playlist_hits = self.__gmusic_search(arg, 'playlist')
            if playlist_hits:
                playlist = playlist_hits['playlist']
                playlist_contents = self.__gmusic.get_shared_playlist_contents(playlist['shareToken'])
            else:
                raise KeyError

            print_nfo("[Google Play Music] [Playlist] '{}'." \
                      .format(playlist['name']).encode('utf-8'))

            for item in playlist_contents:
                print_nfo("[Google Play Music] [Playlist Track] '{} by {} (Album: {}, {})'." \
                          .format((item['track']['title']).encode('utf-8'),
                                  (item['track']['artist']).encode('utf-8'),
                                  (item['track']['album']).encode('utf-8'),
                                  (item['track']['year'])))
                track = item['track']
                playlist_tracks.append(track)

            if not playlist_tracks:
                raise KeyError

            tracks_added = self.__enqueue_tracks(playlist_tracks)
            logging.info("Added %d tracks from %s to queue", \
                         tracks_added, arg)
            self.__update_play_queue_order()

        except KeyError:
            raise KeyError("Playlist not found : {0}".format(arg))
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def enqueue_promoted_tracks_unlimited(self):
        """ Retrieve the url of the next track in the playback queue.

        """
        try:
            tracks = self.__gmusic.get_promoted_songs()
            count = 0
            for track in tracks:
                store_track = self.__gmusic.get_track_info(track['storeId'])
                if u'id' not in store_track.keys():
                    store_track[u'id'] = store_track['storeId']
                self.queue.append(store_track)
                count += 1
            if count == 0:
                print_wrn("[Google Play Music] Operation requires " \
                          "an Unlimited subscription.")
            logging.info("Added %d Unlimited promoted tracks to queue", \
                         count)
            self.__update_play_queue_order()
        except CallFailure:
            raise RuntimeError("Operation requires an Unlimited subscription.")

    def next_url(self):
        """ Retrieve the url of the next track in the playback queue.

        """
        if len(self.queue):
            self.queue_index += 1
            if (self.queue_index < len(self.queue)) \
               and (self.queue_index >= 0):
                next_song = self.queue[self.play_queue_order[self.queue_index]]
                return self.__retrieve_track_url(next_song)
            else:
                self.queue_index = -1
                return self.next_url()
        else:
            return ''

    def prev_url(self):
        """ Retrieve the url of the previous track in the playback queue.

        """
        if len(self.queue):
            self.queue_index -= 1
            if (self.queue_index < len(self.queue)) \
               and (self.queue_index >= 0):
                prev_song = self.queue[self.play_queue_order[self.queue_index]]
                return self.__retrieve_track_url(prev_song)
            else:
                self.queue_index = len(self.queue)
                return self.prev_url()
        else:
            return ''

    def __update_play_queue_order(self):
        """ Update the queue playback order.

        A sequential order is applied if the current play mode is "NORMAL" or a
        random order if current play mode is "SHUFFLE"

        """
        total_tracks = len(self.queue)
        if total_tracks:
            if not len(self.play_queue_order):
                # Create a sequential play order, if empty
                self.play_queue_order = range(total_tracks)
            if self.current_play_mode == self.play_modes.SHUFFLE:
                random.shuffle(self.play_queue_order)
            print_nfo("[Google Play Music] [Tracks in queue] '{0}'." \
                      .format(total_tracks))

    def __retrieve_track_url(self, song):
        """ Retrieve a song url

        """
        if song.get('episodeId'):
            song_url = self.__gmusic.get_podcast_episode_stream_url(song['episodeId'], self.__device_id)
        else:
            song_url = self.__gmusic.get_stream_url(song['id'], self.__device_id)

        try:
            self.now_playing_song = song
            return song_url
        except AttributeError:
            logging.info("Could not retrieve the song url!")
            raise

    def __update_local_library(self):
        """ Retrieve the songs and albums from the user's library

        """
        print_msg("[Google Play Music] [Retrieving library] : '{0}'. " \
                  .format(self.__email))

        songs = self.__gmusic.get_all_songs()
        self.playlists[self.thumbs_up_playlist_name] = list()

        # Retrieve the user's song library
        for song in songs:
            if "rating" in song and song['rating'] == "5":
                self.playlists[self.thumbs_up_playlist_name].append(song)

            song_id = song['id']
            song_artist = song['artist']
            song_album = song['album']

            self.song_map[song_id] = song

            if song_artist == "":
                song_artist = "Unknown Artist"

            if song_album == "":
                song_album = "Unknown Album"

            if song_artist not in self.library:
                self.library[song_artist] = CaseInsensitiveDict()
                self.library[song_artist][self.all_songs_album_title] = list()

            if song_album not in self.library[song_artist]:
                self.library[song_artist][song_album] = list()

            self.library[song_artist][song_album].append(song)
            self.library[song_artist][self.all_songs_album_title].append(song)

        # Sort albums by track number
        for artist in self.library.keys():
            logging.info("Artist : %s", to_ascii(artist))
            for album in self.library[artist].keys():
                logging.info("   Album : %s", to_ascii(album))
                if album == self.all_songs_album_title:
                    sorted_album = sorted(self.library[artist][album],
                                          key=lambda k: k['title'])
                else:
                    sorted_album = sorted(self.library[artist][album],
                                          key=lambda k: k.get('trackNumber',
                                                              0))
                self.library[artist][album] = sorted_album

    def __update_stations_unlimited(self):
        """ Retrieve stations (Unlimited)

        """
        self.stations.clear()
        stations = self.__gmusic.get_all_stations()
        self.stations[u"I'm Feeling Lucky"] = 'IFL'
        for station in stations:
            station_name = station['name']
            logging.info("station name : %s", to_ascii(station_name))
            self.stations[station_name] = station['id']

    def __enqueue_user_station_unlimited(self, arg):
        """ Enqueue a user station (Unlimited)

        """
        print_msg("[Google Play Music] [Station search "\
                  "in user's library] : '{0}'. " \
                  .format(self.__email))
        self.__update_stations_unlimited()
        station_name = arg
        station_id = None
        for name, st_id in self.stations.iteritems():
            print_nfo("[Google Play Music] [Station] '{0}'." \
                      .format(to_ascii(name)))
        if arg not in self.stations.keys():
            for name, st_id in self.stations.iteritems():
                if arg.lower() in name.lower():
                    station_id = st_id
                    station_name = name
                    break
        else:
            station_id = self.stations[arg]

        num_tracks = MAX_TRACKS
        tracks = list()
        if station_id:
            try:
                tracks = self.__gmusic.get_station_tracks(station_id, \
                                                          num_tracks)
            except KeyError:
                raise RuntimeError("Operation requires an "
                                   "Unlimited subscription.")
            tracks_added = self.__enqueue_tracks(tracks)
            if tracks_added:
                if arg.lower() != station_name.lower():
                    print_wrn("[Google Play Music] '{0}' not found. " \
                              "Playing '{1}' instead." \
                              .format(arg.encode('utf-8'), name.encode('utf-8')))
                logging.info("Added %d tracks from %s to queue", tracks_added, arg)
                self.__update_play_queue_order()
            else:
                print_wrn("[Google Play Music] '{0}' has no tracks. " \
                          .format(station_name))

        if not len(self.queue):
            print_wrn("[Google Play Music] '{0}' " \
                      "not found in the user's library. " \
                      .format(arg.encode('utf-8')))

    def __enqueue_station_unlimited(self, arg, max_results=MAX_TRACKS, quiet=False):
        """Search for a station and enqueue all of its tracks (Unlimited)

        """
        if not quiet:
            print_msg("[Google Play Music] [Station search in "\
                      "Google Play Music] : '{0}'. " \
                      .format(arg.encode('utf-8')))
        try:
            station_name = arg
            station_id = None
            station = self.__gmusic_search(arg, 'station', max_results, quiet)

            if station:
                station = station['station']
                station_name = station['name']
                seed = station['seed']
                seed_type = seed['seedType']
                track_id = seed['trackId'] if seed_type == u'2' else None
                artist_id = seed['artistId'] if seed_type == u'3' else None
                album_id = seed['albumId'] if seed_type == u'4' else None
                genre_id = seed['genreId'] if seed_type == u'5' else None
                playlist_token = seed['playlistShareToken'] if seed_type == u'8' else None
                curated_station_id = seed['curatedStationId'] if seed_type == u'9' else None
                num_tracks = max_results
                tracks = list()
                try:
                    station_id \
                        = self.__gmusic.create_station(station_name, \
                                                       track_id, \
                                                       artist_id, \
                                                       album_id, \
                                                       genre_id, \
                                                       playlist_token, \
                                                       curated_station_id)
                    tracks \
                        = self.__gmusic.get_station_tracks(station_id, \
                                                           num_tracks)
                except KeyError:
                    raise RuntimeError("Operation requires an "
                                       "Unlimited subscription.")
                tracks_added = self.__enqueue_tracks(tracks)
                if tracks_added:
                    if not quiet:
                        print_wrn("[Google Play Music] [Station] : '{0}'." \
                                  .format(station_name.encode('utf-8')))
                    logging.info("Added %d tracks from %s to queue", \
                                 tracks_added, arg.encode('utf-8'))
                    self.__update_play_queue_order()

        except KeyError:
            raise KeyError("Station not found : {0}".format(arg))

    def __enqueue_situation_unlimited(self, arg):
        """Search for a situation and enqueue all of its tracks (Unlimited)

        """
        print_msg("[Google Play Music] [Situation search in "\
                  "Google Play Music] : '{0}'. " \
                  .format(arg.encode('utf-8')))
        try:
            situation_hits = self.__gmusic.search(arg)['situation_hits']

            # If the search didn't return results, just do another search with
            # an empty string
            if not len(situation_hits):
                situation_hits = self.__gmusic.search("")['situation_hits']
                print_wrn("[Google Play Music] '{0}' not found. "\
                          "Feeling lucky?." \
                          .format(arg.encode('utf-8')))

            # Try to find a "best result", if one exists
            situation = next((hit for hit in situation_hits \
                              if 'best_result' in hit.keys() \
                              and hit['best_result'] == True), None)

            num_tracks = MAX_TRACKS

            # If there is no best result, then get a selection of tracks from
            # each situation. At least we'll play some music.
            if not situation and len(situation_hits):
                max_results = num_tracks / len(situation_hits)
                for hit in situation_hits:
                    situation = hit['situation']
                    print_nfo("[Google Play Music] [Situation] '{0} : {1}'." \
                              .format((hit['situation']['title']).encode('utf-8'),
                                      (hit['situation']['description']).encode('utf-8')))
                    self.__enqueue_station_unlimited(situation['title'], max_results, True)
            elif situation:
                # There is at list one sitution, enqueue its tracks.
                situation = situation['situation']
                max_results = num_tracks
                self.__enqueue_station_unlimited(situation['title'], max_results, True)

            if not situation:
                raise KeyError

        except KeyError:
           raise KeyError("Situation not found : {0}".format(arg))

    def __enqueue_podcast(self, arg):
        """Search for a podcast series and enqueue all of its tracks.

        """
        print_msg("[Google Play Music] [Podcast search in "\
                  "Google Play Music] : '{0}'. " \
                  .format(arg.encode('utf-8')))
        try:
            podcast_hits = self.__gmusic_search(arg, 'podcast', 10, quiet=False)

            if not podcast_hits:
                print_wrn("[Google Play Music] [Podcast] 'Search returned zero results'.")
                print_wrn("[Google Play Music] [Podcast] 'Are you in a supported region "
                          "(currently only US and Canada) ?'")

            # Use the first podcast retrieved. At least we'll play something.
            podcast = dict ()
            if podcast_hits and len(podcast_hits):
                podcast = podcast_hits['series']

            episodes_added = 0
            if podcast:
                # There is a podcast, enqueue its episodes.
                print_nfo("[Google Play Music] [Podcast] 'Playing '{0}' by {1}'." \
                          .format((podcast['title']).encode('utf-8'),
                                  (podcast['author']).encode('utf-8')))
                print_nfo("[Google Play Music] [Podcast] '{0}'." \
                          .format((podcast['description'][0:150]).encode('utf-8')))
                series = self.__gmusic.get_podcast_series_info(podcast['seriesId'])
                episodes = series['episodes']
                for episode in episodes:
                    print_nfo("[Google Play Music] [Podcast Episode] '{0} : {1}'." \
                              .format((episode['title']).encode('utf-8'),
                                      (episode['description'][0:80]).encode('utf-8')))
                episodes_added = self.__enqueue_tracks(episodes)

            if not podcast or not episodes_added:
                raise KeyError

        except KeyError:
            raise KeyError("Podcast not found or no episodes found: {0}".format(arg))

    def __enqueue_tracks(self, tracks):
        """ Add tracks to the playback queue

        """
        count = 0
        for track in tracks:
            if u'id' not in track.keys() and track.get('storeId'):
                track[u'id'] = track['storeId']
            self.queue.append(track)
            count += 1
        return count

    def __update_playlists(self):
        """ Retrieve the user's playlists

        """
        plists = self.__gmusic.get_all_user_playlist_contents()
        for plist in plists:
            plist_name = plist.get('name')
            tracks = plist.get('tracks')
            if plist_name and tracks:
                logging.info("playlist name : %s", to_ascii(plist_name))
                tracks.sort(key=itemgetter('creationTimestamp'))
                self.playlists[plist_name] = list()
                for track in tracks:
                    song_id = track.get('trackId')
                    if song_id:
                        song = self.song_map.get(song_id)
                        if song:
                            self.playlists[plist_name].append(song)

    def __update_playlists_unlimited(self):
        """ Retrieve shared playlists (Unlimited)

        """
        plists_subscribed_to = [p for p in self.__gmusic.get_all_playlists() \
                                if p.get('type') == 'SHARED']
        for plist in plists_subscribed_to:
            share_tok = plist['shareToken']
            playlist_items \
                = self.__gmusic.get_shared_playlist_contents(share_tok)
            plist_name = plist['name']
            logging.info("shared playlist name : %s", to_ascii(plist_name))
            self.playlists[plist_name] = list()
            for item in playlist_items:
                try:
                    song = item['track']
                    song['id'] = item['trackId']
                    self.playlists[plist_name].append(song)
                except IndexError:
                    pass

    def __gmusic_search(self, query, query_type, max_results=MAX_TRACKS, quiet=False):
        """ Search Google Play (Unlimited)

        """

        search_results = self.__gmusic.search(query, max_results)[query_type + '_hits']

        # This is a workaround. Some podcast results come without these two
        # keys in the dictionary
        if query_type == "podcast" and len(search_results) \
           and not search_results[0].get('navigational_result'):
            for res in search_results:
                res[u'best_result'] = False
                res[u'navigational_result'] = False
                res[query_type] = res['series']

        result = ''
        if query_type != "playlist":
            result = next((hit for hit in search_results \
                           if 'best_result' in hit.keys() \
                           and hit['best_result'] == True), None)

        if not result and len(search_results):
            secondary_hit = None
            for hit in search_results:
                name = ''
                if hit[query_type].get('name'):
                    name = hit[query_type].get('name')
                elif hit[query_type].get('title'):
                    name = hit[query_type].get('title')
                if not quiet:
                    print_nfo("[Google Play Music] [{0}] '{1}'." \
                              .format(query_type.capitalize(),
                                      (name).encode('utf-8')))
                if query.lower() == \
                   to_ascii(name).lower():
                    result = hit
                    break
                if query.lower() in \
                   to_ascii(name).lower():
                    secondary_hit = hit
            if not result and secondary_hit:
                result = secondary_hit

        if not result and not len(search_results):
            # Do another search with an empty string
            search_results = self.__gmusic.search("")[query_type + '_hits']

        if not result and len(search_results):
            # Play some random result from the search results
            random.seed()
            result = random.choice(search_results)
            if not quiet:
                print_wrn("[Google Play Music] '{0}' not found. "\
                          "Feeling lucky?." \
                          .format(query.encode('utf-8')))

        return result