Modifying scripts for VPN detection.

[pingpong.git] / base_gexf_generator.py

diff --git a/base_gexf_generator.py b/base_gexf_generator.py
index a85184af02320347bf2138461fa62b3bb4b7672d..ad677eccdd2882fd3c9b1258674f02d16bf1b45c 100644 (file)
--- a/base_gexf_generator.py
+++ b/base_gexf_generator.py
@@ -106,7 +106,7 @@ def traverse_and_merge_nodes(G, dev_list_file):
     dev_list = create_device_list(DEVICE_MAC_LIST)
     # Traverse every node
     # Check that the node is not a smarthome device
-    for node in nodes:
+    for node in list(nodes):
         neighbors = G[node] #G.neighbors(node)
         #print "Neighbors: ", neighbors, "\n"
         # Skip if the node is a smarthome device
@@ -170,10 +170,19 @@ def place_in_graph(G, eth_src, eth_dst, device_dns_mappings, dev_list, layers,
     else:
         protocol = split_protocol[3] + ":" + split_protocol[4]
     #print "timestamp: ", timestamp, " - new protocol added: ", protocol, "\n"
+    # And source and destination IPs
+    ip_src = layers[JSON_KEY_IP][JSON_KEY_IP_SRC]
+    ip_dst = layers[JSON_KEY_IP][JSON_KEY_IP_DST]
+    # Categorize source and destination IP addresses: local vs. non-local
+    #ip_re = re.compile(r'\b192.168.[0-9.]+')
+    ip_re = re.compile(r'\b192.168.1.[0-9.]+')
+    src_is_local = ip_re.search(ip_src) 
+    dst_is_local = ip_re.search(ip_dst)
     # Store protocol into the set (source)
     protocols = None
     # Key to search in the dictionary is <src-mac-address>-<dst-mac_address>
-    dict_key = eth_src + "-" + eth_dst
+    dict_key = ip_src + "-" + ip_dst
+    #print "Key: ", dict_key
     if dict_key not in edge_to_prot:
         edge_to_prot[dict_key] = set()
     protocols = edge_to_prot[dict_key]
@@ -185,13 +194,6 @@ def place_in_graph(G, eth_src, eth_dst, device_dns_mappings, dev_list, layers,
         edge_to_vol[dict_key] = 0;
     edge_to_vol[dict_key] = edge_to_vol[dict_key] + packet_len
     volume = str(edge_to_vol[dict_key])
-    # And source and destination IPs
-    ip_src = layers[JSON_KEY_IP][JSON_KEY_IP_SRC]
-    ip_dst = layers[JSON_KEY_IP][JSON_KEY_IP_DST]
-    # Categorize source and destination IP addresses: local vs. non-local
-    ip_re = re.compile(r'\b192.168.[0-9.]+')
-    src_is_local = ip_re.search(ip_src) 
-    dst_is_local = ip_re.search(ip_dst)
 
     # Skip device to cloud communication if we are interested in the local graph.
     # TODO should this go before the protocol dict is changed?
@@ -252,7 +254,7 @@ def parse_json(file_path):
     # Create an exclusion list
     exc_list = create_device_list(EXCLUSION_MAC_LIST)
     # First parse the file once, constructing a map that contains information about individual devices' DNS resolutions.
-    device_dns_mappings = parser.parse_dns.parse_json_dns(file_path) # "./json/eth1.dump.json"
+    device_dns_mappings = parser.parse_dns.parse_json_dns(file_path)
     # Init empty graph
     G = nx.DiGraph()
     # Mapping from edge to a set of protocols
@@ -260,6 +262,7 @@ def parse_json(file_path):
     # Mapping from edge to traffic volume
     edge_to_vol = dict()
     # Parse file again, this time constructing a graph of device<->server and device<->device communication.
+    i = 0
     with open(file_path) as jf:
         # Read JSON; data becomes reference to root JSON object (or in our case json array)
         data = json.load(jf)