示例#1
0
class CookieTransport(xmlrpclib.Transport):
    '''A subclass of xmlrpclib.Transport that supports cookies.'''
    cookiejar = None
    scheme = 'http'

    # Cribbed from xmlrpclib.Transport.send_user_agent
    def send_cookies(self, connection, cookie_request):
        if self.cookiejar is None:
            self.cookiejar = CookieJar()
        elif self.cookiejar:
            # Let the cookiejar figure out what cookies are appropriate
            self.cookiejar.add_cookie_header(cookie_request)
            # Pull the cookie headers out of the request object...
            cookielist=list()
            for h,v in cookie_request.header_items():
                if h.startswith('Cookie'):
                    cookielist.append([h,v])
            # ...and put them over the connection
            for h,v in cookielist:
                connection.putheader(h,v)

    # This is just python 2.7's xmlrpclib.Transport.single_request, with
    # send additions noted below to send cookies along with the request
    def single_request_with_cookies(self, host, handler, request_body, verbose=0):
        # ADDED: construct the URL and Request object for proper cookie handling
        request_url = "%s://%s%s" % (self.scheme,host,handler)
        #log.debug("request_url is %s" % request_url)
        cookie_request  = urllib2.Request(request_url)

        try:
            if six.PY2:
                h = self.make_connection(host)
                if verbose:
                    h.set_debuglevel(1)
                self.send_request(h, handler, request_body)
                self.send_host(h, host)
                self.send_cookies(h, cookie_request)  # ADDED. creates cookiejar if None.
                self.send_user_agent(h)
                self.send_content(h, request_body)
            else:
                # Python 3 xmlrpc.client.Transport makes its own connection
                h = self.send_request(host, handler, request_body, verbose)

            response = h.getresponse()

            # ADDED: parse headers and get cookies here
            cookie_response = CookieResponse(response.msg)
            # Okay, extract the cookies from the headers
            self.cookiejar.extract_cookies(cookie_response,cookie_request)
            #log.debug("cookiejar now contains: %s" % self.cookiejar._cookies)
            # And write back any changes
            if hasattr(self.cookiejar,'save'):
                try:
                    self.cookiejar.save(self.cookiejar.filename)
                except Exception as e:
                    raise
                    #log.error("Couldn't write cookiefile %s: %s" % \
                    #        (self.cookiejar.filename,str(e)))

            if response.status == 200:
                self.verbose = verbose
                return self.parse_response(response)

            if (response.getheader("content-length", 0)):
                response.read()
            raise xmlrpclib.ProtocolError(
                host + handler,
                response.status, response.reason,
                response.msg,
                )
        except xmlrpclib.Fault:
            raise
        finally:
            try:
                h.close()
            except NameError:  # h not initialized yet
                pass

    # Override the appropriate request method
    single_request = single_request_with_cookies # python 2.7+
示例#2
0
文件: cookie.py 项目: lorien/grab
class CookieManager(object):
    """
    Each Grab instance has `cookies` attribute that is instance of
    `CookieManager` class.

    That class contains helpful methods to create, load, save cookies from/to
    different places.
    """

    __slots__ = ('cookiejar',)

    def __init__(self, cookiejar=None):
        if cookiejar is not None:
            self.cookiejar = cookiejar
        else:
            self.cookiejar = CookieJar()
        # self.disable_cookiejar_lock(self.cookiejar)

    # def disable_cookiejar_lock(self, cj):
        # cj._cookies_lock = dummy_threading.RLock()

    def set(self, name, value, domain, **kwargs):
        """Add new cookie or replace existing cookie with same parameters.

        :param name: name of cookie
        :param value: value of cookie
        :param kwargs: extra attributes of cookie
        """

        if domain == 'localhost':
            domain = ''

        self.cookiejar.set_cookie(create_cookie(name, value, domain, **kwargs))

    def update(self, cookies):
        if isinstance(cookies, CookieJar):
            for cookie in cookies:
                self.cookiejar.set_cookie(cookie)
        elif isinstance(cookies, CookieManager):
            for cookie in cookies.cookiejar:
                self.cookiejar.set_cookie(cookie)
        else:
            raise GrabMisuseError('Unknown type of cookies argument: %s'
                                  % type(cookies))

    @classmethod
    def from_cookie_list(cls, clist):
        jar = CookieJar()
        for cookie in clist:
            jar.set_cookie(cookie)
        return cls(jar)

    def clear(self):
        self.cookiejar = CookieJar()

    def __getstate__(self):
        state = {}
        for cls in type(self).mro():
            cls_slots = getattr(cls, '__slots__', ())
            for slot in cls_slots:
                if slot != '__weakref__':
                    if hasattr(self, slot):
                        state[slot] = getattr(self, slot)

        state['_cookiejar_cookies'] = list(self.cookiejar)
        del state['cookiejar']

        return state

    def __setstate__(self, state):
        state['cookiejar'] = CookieJar()
        for cookie in state['_cookiejar_cookies']:
            state['cookiejar'].set_cookie(cookie)
        del state['_cookiejar_cookies']

        for slot, value in state.items():
            setattr(self, slot, value)

    def __getitem__(self, key):
        for cookie in self.cookiejar:
            if cookie.name == key:
                return cookie.value
        raise KeyError

    def items(self):
        res = []
        for cookie in self.cookiejar:
            res.append((cookie.name, cookie.value))
        return res

    def load_from_file(self, path):
        """
        Load cookies from the file.

        Content of file should be a JSON-serialized list of dicts.
        """

        with open(path) as inf:
            data = inf.read()
            if data:
                items = json.loads(data)
            else:
                items = {}
        for item in items:
            extra = dict((x, y) for x, y in item.items()
                         if x not in ['name', 'value', 'domain'])
            self.set(item['name'], item['value'], item['domain'], **extra)

    def get_dict(self):
        res = []
        for cookie in self.cookiejar:
            res.append(dict((x, getattr(cookie, x)) for x in COOKIE_ATTRS))
        return res

    def save_to_file(self, path):
        """
        Dump all cookies to file.

        Cookies are dumped as JSON-serialized dict of keys and values.
        """

        with open(path, 'w') as out:
            out.write(json.dumps(self.get_dict()))

    def get_cookie_header(self, req):
        """
        :param req: object with httplib.Request interface
            Actually, it have to have `url` and `headers` attributes
        """
        mocked_req = MockRequest(req)
        self.cookiejar.add_cookie_header(mocked_req)
        return mocked_req.get_new_headers().get('Cookie')
示例#3
0
class CookieManager(object):
    """
    Each Grab instance has `cookies` attribute that is instance of
    `CookieManager` class.

    That class contains helpful methods to create, load, save cookies from/to
    different places.
    """

    __slots__ = ('cookiejar', )

    def __init__(self, cookiejar=None):
        if cookiejar is not None:
            self.cookiejar = cookiejar
        else:
            self.cookiejar = CookieJar()
        # self.disable_cookiejar_lock(self.cookiejar)

    # def disable_cookiejar_lock(self, cj):
    # cj._cookies_lock = dummy_threading.RLock()

    def set(self, name, value, domain, **kwargs):
        """Add new cookie or replace existing cookie with same parameters.

        :param name: name of cookie
        :param value: value of cookie
        :param kwargs: extra attributes of cookie
        """

        if domain == 'localhost':
            domain = ''

        self.cookiejar.set_cookie(create_cookie(name, value, domain, **kwargs))

    def update(self, cookies):
        if isinstance(cookies, CookieJar):
            for cookie in cookies:
                self.cookiejar.set_cookie(cookie)
        elif isinstance(cookies, CookieManager):
            for cookie in cookies.cookiejar:
                self.cookiejar.set_cookie(cookie)
        else:
            raise GrabMisuseError('Unknown type of cookies argument: %s' %
                                  type(cookies))

    @classmethod
    def from_cookie_list(cls, clist):
        jar = CookieJar()
        for cookie in clist:
            jar.set_cookie(cookie)
        return cls(jar)

    def clear(self):
        self.cookiejar = CookieJar()

    def __getstate__(self):
        state = {}
        for cls in type(self).mro():
            cls_slots = getattr(cls, '__slots__', ())
            for slot in cls_slots:
                if slot != '__weakref__':
                    if hasattr(self, slot):
                        state[slot] = getattr(self, slot)

        state['_cookiejar_cookies'] = list(self.cookiejar)
        del state['cookiejar']

        return state

    def __setstate__(self, state):
        state['cookiejar'] = CookieJar()
        for cookie in state['_cookiejar_cookies']:
            state['cookiejar'].set_cookie(cookie)
        del state['_cookiejar_cookies']

        for slot, value in state.items():
            setattr(self, slot, value)

    def __getitem__(self, key):
        for cookie in self.cookiejar:
            if cookie.name == key:
                return cookie.value
        raise KeyError

    def items(self):
        res = []
        for cookie in self.cookiejar:
            res.append((cookie.name, cookie.value))
        return res

    def load_from_file(self, path):
        """
        Load cookies from the file.

        Content of file should be a JSON-serialized list of dicts.
        """

        with open(path) as inf:
            data = inf.read()
            if data:
                items = json.loads(data)
            else:
                items = {}
        for item in items:
            extra = dict((x, y) for x, y in item.items()
                         if x not in ['name', 'value', 'domain'])
            self.set(item['name'], item['value'], item['domain'], **extra)

    def get_dict(self):
        res = []
        for cookie in self.cookiejar:
            res.append(dict((x, getattr(cookie, x)) for x in COOKIE_ATTRS))
        return res

    def save_to_file(self, path):
        """
        Dump all cookies to file.

        Cookies are dumped as JSON-serialized dict of keys and values.
        """

        with open(path, 'w') as out:
            out.write(json.dumps(self.get_dict()))

    def get_cookie_header(self, req):
        """
        :param req: object with httplib.Request interface
            Actually, it have to have `url` and `headers` attributes
        """
        mocked_req = MockRequest(req)
        self.cookiejar.add_cookie_header(mocked_req)
        return mocked_req.get_new_headers().get('Cookie')
示例#4
0
class HttpTransport(Transport):
    """
    HTTP transport using urllib2.  Provided basic http transport
    that provides for cookies, proxies but no authentication.
    """

    def __init__(self, **kwargs):
        """
        @param kwargs: Keyword arguments.
            - B{proxy} - An http proxy to be specified on requests.
                 The proxy is defined as {protocol:proxy,}
                    - type: I{dict}
                    - default: {}
            - B{timeout} - Set the url open timeout (seconds).
                    - type: I{float}
                    - default: 90
        """
        Transport.__init__(self)
        Unskin(self.options).update(kwargs)
        self.cookiejar = CookieJar()
        self.proxy = {}
        self.urlopener = None

    def open(self, request):
        try:
            url = request.url
            log.debug('opening (%s)', url)
            u2request = urllib.request.Request(url)
            self.proxy = self.options.proxy
            return self.u2open(u2request)
        except urllib.error.HTTPError as e:
            raise TransportError(str(e), e.code, e.fp)

    def send(self, request):
        result = None
        url = request.url
        msg = request.message
        headers = request.headers
        try:
            u2request = urllib.request.Request(url, msg, headers)
            self.addcookies(u2request)
            self.proxy = self.options.proxy
            request.headers.update(u2request.headers)
            log.debug('sending:\n%s', request)
            fp = self.u2open(u2request)
            self.getcookies(fp, u2request)
            result = Reply(200, fp.headers.__dict__, fp.read())
            log.debug('received:\n%s', result)
        except urllib.error.HTTPError as e:
            if e.code in (202, 204):
                result = None
            else:
                raise TransportError(e.msg, e.code, e.fp)
        return result

    def addcookies(self, u2request):
        """
        Add cookies in the cookiejar to the request.
        @param u2request: A urllib2 request.
        @rtype: u2request: urllib2.Requet.
        """
        self.cookiejar.add_cookie_header(u2request)

    def getcookies(self, fp, u2request):
        """
        Add cookies in the request to the cookiejar.
        @param u2request: A urllib2 request.
        @rtype: u2request: urllib2.Requet.
        """
        self.cookiejar.extract_cookies(fp, u2request)

    def u2open(self, u2request):
        """
        Open a connection.
        @param u2request: A urllib2 request.
        @type u2request: urllib2.Requet.
        @return: The opened file-like urllib2 object.
        @rtype: fp
        """
        tm = self.options.timeout
        url = self.u2opener()
        return url.open(u2request, timeout=tm)

    def u2opener(self):
        """
        Create a urllib opener.
        @return: An opener.
        @rtype: I{OpenerDirector}
        """
        if self.urlopener is None:
            return urllib.request.build_opener(*self.u2handlers())
        else:
            return self.urlopener

    def u2handlers(self):
        """
        Get a collection of urllib handlers.
        @return: A list of handlers to be installed in the opener.
        @rtype: [Handler,...]
        """
        handlers = []
        handlers.append(urllib.request.ProxyHandler(self.proxy))
        return handlers

    def __deepcopy__(self, memo={}):
        clone = self.__class__()
        p = Unskin(self.options)
        cp = Unskin(clone.options)
        cp.update(p)
        return clone