def __init__(self, modhandler, url, password):
self.cwd_vector = None
self.path = None
self.proxy = None
self.modhandler = modhandler
self.post_data = {}
self.current_mode = None
self.use_current_path = True
self.available_modes = self.params.get_parameter_choices('mode')
mode = self.params.get_parameter_value('mode')
if mode:
self.modes = [mode]
else:
self.modes = self.available_modes
proxy = self.params.get_parameter_value('proxy')
if proxy:
self.mprint(
'[!] Proxies can break weevely requests, if possibile use proxychains'
)
self.proxy = {'http': proxy}
Module.__init__(self, modhandler, url, password)
def __init__(self, **kwargs):
self._name = 'Stacked_'+kwargs['ae_type']
kwargs['dvc'] = torch.device('cpu')
Module.__init__(self, **kwargs)
self._feature, self._output = self.Sequential(out_number = 2)
self.opt()
self.Stacked()
def __init__(self, modhandler, url, password):
self.cwd_vector = None
self.path = None
self.proxy = None
self.modhandler = modhandler
self.post_data = {}
self.current_mode = None
self.use_current_path = True
self.available_modes = self.params.get_parameter_choices('mode')
mode = self.params.get_parameter_value('mode')
if mode:
self.modes = [ mode ]
else:
self.modes = self.available_modes
proxy = self.params.get_parameter_value('proxy')
if proxy:
self.mprint('[!] Proxies can break weevely requests, if possibile use proxychains')
self.proxy = { 'http' : proxy }
Module.__init__(self, modhandler, url, password)
def __init__(self, **kwargs):
if 'name' in kwargs.keys():
kwargs['_name'] = kwargs['name']
del kwargs['name']
if '_name' not in kwargs.keys(): kwargs['_name'] = 'DBN'
# 检测是否设置单元类型 - 用于预训练
if 'v_type' not in kwargs.keys():
kwargs['v_type'] = ['Gaussian']
if 'h_type' not in kwargs.keys():
kwargs['h_type'] = ['Gaussian']
Module.__init__(self, **kwargs)
if type(self.h_type) != list:
self.h_type = [self.h_type]
# 如果未定义 hidden_func 则按单元类型给定 - 用于微调
if hasattr(self, 'hidden_func') == False:
self.hidden_func = []
for tp in self.h_type:
if tp in ['Gaussian', 'g']: self.hidden_func.append('a')
elif tp in ['Binary', 'b']: self.hidden_func.append('s')
else: raise Exception("Unknown h_type!")
self._feature, self._output = self.Sequential(out_number=2)
self.opt()
self.Stacked()
def __init__(self, modhandler, url, password):
self.rand_post_addr = ''.join([choice('abcdefghijklmnopqrstuvwxyz') for i in xrange(4)])
self.rand_post_port = ''.join([choice('abcdefghijklmnopqrstuvwxyz') for i in xrange(4)])
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
Module.__init__(self, modhandler, url, password)
self.file_content = None
self.rand_post_name = ''.join(
[choice('abcdefghijklmnopqrstuvwxyz') for i in xrange(4)])
def __init__(self, **kwargs):
self._name = 'CNN'
Module.__init__(self, **kwargs)
Conv_Module.__init__(self, **kwargs)
self.layers = self.Convolutional()
self.fc = self.Sequential()
self.opt()
def __init__( self, modhandler , url, password):
self.probe_filename = ''.join(choice(letters) for i in xrange(4)) + '.html'
self.found_url = None
self.found_dir = None
Module.__init__(self, modhandler, url, password)
def __init__( self, modhandler , url, password):
self.probe_filename = ''.join(choice(letters) for i in xrange(4)) + '.html'
self.found_url = None
self.found_dir = None
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.rand_post_addr = ''.join(
[choice('abcdefghijklmnopqrstuvwxyz') for i in xrange(4)])
self.rand_post_port = ''.join(
[choice('abcdefghijklmnopqrstuvwxyz') for i in xrange(4)])
Module.__init__(self, modhandler, url, password)
def __init__(self,
cfg=None,
batch_norm=True,
use_bias=False,
load_pre=None,
init_weights=True,
**kwargs):
if type(cfg) == str: arch = cfgs[cfg]
else: arch = cfg
default = {
'img_size': [3, 224, 224],
'conv_struct': arch,
'conv_func': 'ReLU(True)',
'res_func': 'ReLU(True)',
'struct': [-1, 1000],
'dropout': 0,
'hidden_func': 'ReLU(True)'
}
for key in default.keys():
if key not in kwargs.keys():
kwargs[key] = default[key]
self.batch_norm = batch_norm
self.use_bias = use_bias
if type(cfg) == str:
self._name = cfg.upper()
else:
self._name = 'ResNet'
Module.__init__(self, **kwargs)
Conv_Module.__init__(self, **kwargs)
if type(cfg) == str:
blocks = self.Convolutional()
index = block_id[cfg]
self.conv1, self.bn1, self.relu, self.maxpool = \
blocks[0].conv_layers[0], blocks[0].conv_layers[1], blocks[0].act_layer, blocks[0].pool_layer
self.layer1 = nn.Sequential(*blocks[index[0]:index[1]])
self.layer2 = nn.Sequential(*blocks[index[1]:index[2]])
self.layer3 = nn.Sequential(*blocks[index[2]:index[3]])
self.layer4 = nn.Sequential(*blocks[index[3]:index[4]])
self.layers = blocks.children()
elif type(cfg) == list:
self.conv_struct = cfg
self.layers = self.Convolutional()
self.fc = self.Sequential()
self.opt()
if init_weights:
self._initialize_weights()
if load_pre == True or type(load_pre) == str:
self.load_pre(cfg, batch_norm, load_pre)
def __init__(self, **kwargs):
if 'name' in kwargs.keys():
kwargs['_name'] = kwargs['name']
del kwargs['name']
if '_name' not in kwargs.keys(): kwargs['_name'] = 'Stacked_'+kwargs['ae_type']
if 'decoder_func' not in kwargs.keys(): kwargs['decoder_func'] = 'a'
Module.__init__(self, **kwargs)
self._feature, self._output = self.Sequential(out_number = 2)
self.opt()
self.Stacked()
def __init__(self, modhandler, url, password):
self.encoder_callable = False
self.md5_callable = False
self.payload = None
self.vector = None
self.interpreter = None
self.transfer_dir = None
self.transfer_url_dir = None
self.lastreadfile = ''
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.encoder_callable = False
self.md5_callable = False
self.payload = None
self.vector = None
self.interpreter = None
self.transfer_dir = None
self.transfer_url_dir = None
self.lastreadfile = ''
Module.__init__(self, modhandler, url, password)
def __init__(self, **kwargs):
default = {
'struct2': None, # 解码部分的结构，默认为编码部分反向
'hidden_func2': None,
'n_category': None,
'dropout': 0.0,
'exec_dropout': [False, True],
'lr': 1e-3
}
for key in default.keys():
if key not in kwargs:
kwargs[key] = default[key]
kwargs['dvc'] = torch.device('cpu')
self._name = 'MMDGM_VAE'
Module.__init__(self, **kwargs)
# q(z|x)
self.qx = self.Sequential(struct=self.struct[:2],
hidden_func=self.hidden_func)
self.qy = self.Sequential(struct=[self.n_category, self.struct[1]],
hidden_func=self.hidden_func)
self.Q = self.Sequential(struct=self.struct[1:-1],
hidden_func=self.hidden_func[1:])
self.z_mu = self.Sequential(struct=self.struct[-2:], hidden_func='a')
self.z_logvar = self.Sequential(struct=self.struct[-2:],
hidden_func='a')
# p(x|z)
if self.struct2 is None:
self.struct2 = self.struct.copy()
self.struct2.reverse()
if self.hidden_func2 is None:
if type(self.hidden_func) == str:
self.hidden_func = [self.hidden_func]
self.hidden_func2 = self.hidden_func.copy()
self.hidden_func2.reverse()
self.pz = self.Sequential(struct=self.struct2[:2],
hidden_func=self.hidden_func2)
self.py = self.Sequential(struct=[self.n_category, self.struct2[1]],
hidden_func=self.hidden_func2)
self.P = self.Sequential(struct=self.struct2[1:],
hidden_func=self.hidden_func2[1:])
self.opt()
def __init__(self, module_list, **kwargs):
'''
_loss: 附加损失
loss: 全部损失
'''
name = ''
for i in range(len(module_list)):
if hasattr(module_list[i], 'name'):
name += module_list[i].name
if i < len(module_list) - 1:
name += '-'
self.name = name
Module.__init__(self, **kwargs)
self.modules = nn.Sequential(*module_list)
self.module_list = module_list
self.opt(False)
def __init__(self, cfg = None,
batch_norm = False, use_bias = True,
load_pre = None, init_weights=True, **kwargs):
if type(cfg) == str: arch = cfgs[cfg]
else: arch = cfg
default = {'img_size': [3, 224, 224],
'conv_struct': arch,
'conv_func': 'ReLU(True)',
'struct': [-1, 4096, 4096, 1000],
'dropout': [0, 0.5, 0.5],
'hidden_func': 'ReLU(True)'
}
for key in default.keys():
if key not in kwargs.keys():
kwargs[key] = default[key]
self.batch_norm = batch_norm
self.use_bias = use_bias
if type(cfg) == str:
self._name = cfg.upper()
elif type(cfg) == list:
self._name = 'VGG'
self.conv_struct = cfg
Module.__init__(self,**kwargs)
Conv_Module.__init__(self,**kwargs)
self.features = self.Convolutional('layers', auto_name = False)
self.classifier = self.Sequential()
self.opt()
if init_weights:
self._initialize_weights()
if load_pre == True or type(load_pre) == str:
self.load_pre(cfg, batch_norm, load_pre)
def __init__(self, modhandler, url, password):
self.structure = {}
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.chunksize = 5000
self.substitutive_wl = []
Module.__init__(self, modhandler, url, password)
def __init__( self, modhandler , url, password):
self.chunksize = 5000
self.substitutive_wl = []
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.ifaces = {}
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
Module.__init__(self, modhandler, url, password)
self.file_content = None
self.rand_post_name = ''.join([choice('abcdefghijklmnopqrstuvwxyz') for i in xrange(4)])
def __init__(self, modhandler, url, password):
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
Module.__init__(self, modhandler, url, password)
self.usersinfo = {}
def __init__(self, modhandler, url, password):
self.pathdict = {}
Module.__init__(self, modhandler, url, password)
def __init__( self, modhandler , url, password):
Module.__init__(self, modhandler, url, password)
self.usersfiles = {}
def __init__(self, modhandler, url, password):
self.list = []
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.reqlist = RequestList(modhandler)
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.list = []
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.last_vector = None
self.done = False
Module.__init__(self, modhandler, url, password)
def __init__(self, modhandler, url, password):
self.pathdict = {}
Module.__init__(self, modhandler, url, password)
def __init__( self, modhandler , url, password):
self.structure = {}
Module.__init__(self, modhandler, url, password)
def __init__( self, modhandler , url, password):
self.last_vector = None
self.done = False
Module.__init__(self, modhandler, url, password)
P(arg='cmd', help='Shell command', required=True, pos=0),
P(arg='stderr', help='Print standard error', default=True, type=bool)
)
def __init__( self, modhandler , url, password):
self.payload = None
self.cwd_vector = None
try:
modhandler.load('shell.php')
except ModuleException, e:
raise
else:
Module.__init__(self, modhandler, url, password)
def __execute_probe(self, vector):
try:
rand = random.randint( 11111, 99999 )
response = self.run_module( "echo %d" % rand, True, vector.payloads[0] )
if response == str(rand):
self.params.set_and_check_parameters({'vector' : vector.name})
return True
except:
#pass
def __init__(self, **kwargs):
self._name = 'DNN'
Module.__init__(self, **kwargs)
self._feature, self._output = self.Sequential(out_number = 2)
self.opt()
