def decompress(compressedfile, output, treefile='treefile.json'):
"""
File now written to with bitstring
"""
with open(compressedfile, 'rb') as f:
b = BitArray(f.read())
tree = json.load(open(treefile))
"""
Binary values returned here as string
need to create a loop to go through and scan until a match it found
start with 1 bit, loop until ended
write to file when bit is found
"""
newfile = open(output, "a", encoding="utf-8")
while len(b.bin) >= 0:
x = 0
i = 1
while x == 0:
tosearch = b.bin[0:i]
try:
towrite = list(tree.keys())[list(
tree.values()).index(tosearch)]
newfile.write(towrite)
newfile.close()
newfile = open(output, "a", encoding="utf-8")
b.bin = b.bin[i:]
x = 1
except:
i += 1
def compress(text):
tree = get_tree(text)
decompression_tree = get_decompression_tree(tree)
compressed_text = ""
for c in text:
compressed_text += get_code(c, tree)
# Add filler to get to a round nb of bytes
compressed_text += '0' * (8 - (len(compressed_text) % 8))
binary = BitArray()
binary.bin = compressed_text
return binary, decompression_tree
def get_weights(firstGeneration=False, DNA=''):
if firstGeneration:
#Generate random DNA for the first generation
DNAlength = (inputCount * neuronsPerLayer + neuronsPerLayer +
((neuronsPerLayer * neuronsPerLayer) * numHiddenLayers) +
neuronsPerLayer * outputNeurons +
neuronsPerLayer * numHiddenLayers + outputNeurons)
DNAbin = []
for i in range(0, DNAlength * 12):
DNAbin.append(str(random.randint(0, 1)))
DNAbin = ''.join(DNAbin)
b = BitArray(int=0, length=12)
DNA = []
for i in range(0, len(DNAbin), 12):
b.bin = DNAbin[i:i + 12]
DNA.append(float(b.int))
else:
DNAbin = DNA
DNA = []
b = BitArray(int=0, length=12)
for i in range(0, len(DNAbin), 12):
b.bin = DNAbin[i:i + 12]
DNA.append(float(b.int))
#Convert DNA to list of weights
weights = []
count = 0
#Normalise our converted DNA to floats between -0.5 and 0.5
DNA = 2 * (DNA - np.float32(-2048)) / (np.float32(2047) -
np.float32(-2048)) - 1
#Convert list of weights into shaped arrays for each layer
#Input weights: inputCount x neuronsPerLayer matrix
l = np.array(DNA[count:count + inputCount * neuronsPerLayer])
count += inputCount * neuronsPerLayer
l = l.reshape(inputCount, neuronsPerLayer)
weights.append(l)
#Input activation weights: neuronsPerLayer array
l = np.array(DNA[count:count + neuronsPerLayer])
count += neuronsPerLayer
l = l.reshape(neuronsPerLayer)
weights.append(l)
for i in range(numHiddenLayers):
#Hidden layer weights: neuronsPerLayer x neuronsPerLayer matrix
l = np.array(DNA[count:count + neuronsPerLayer * neuronsPerLayer])
count += neuronsPerLayer * neuronsPerLayer
l = l.reshape(neuronsPerLayer, neuronsPerLayer)
weights.append(l)
#Hidden activation weights: neuronsPerLayer array
l = np.array(DNA[count:count + neuronsPerLayer])
count += neuronsPerLayer
l = l.reshape(neuronsPerLayer)
weights.append(l)
#Output layer weights: neuronsPerLayer x outputNeruons matrix
l = np.array(DNA[count:count + outputNeurons * neuronsPerLayer])
l = l.reshape(neuronsPerLayer, outputNeurons)
weights.append(l)
count += neuronsPerLayer * outputNeurons
#Hidden activation weights: neuronsPerLayer array
l = np.array(DNA[count:count + neuronsPerLayer])
count += neuronsPerLayer
l = l.reshape(outputNeurons)
weights.append(l)
return weights, DNAbin
def BinaryValue_to_scapy(binvalue):
""" take binaryvalue return Ether scapy packet
"""
k = BitArray()
k.bin = binvalue.binstr
return Ether(k.bytes)
def bitstoint(bits):
a = BitArray();
a.bin = bits;
return a.int;
def bitstofloat(bits):
a = BitArray();
a.bin = bits;
return a.float;
