1 package edu.uci.iotproject.util;
3 import edu.uci.iotproject.DnsMap;
4 import edu.uci.iotproject.analysis.PcapPacketPair;
5 import org.apache.commons.math3.stat.clustering.Cluster;
7 import java.io.FileInputStream;
8 import java.io.FileOutputStream;
9 import java.io.ObjectInputStream;
10 import java.io.ObjectOutputStream;
11 import java.util.List;
12 import java.util.Optional;
14 import java.util.stream.Collectors;
16 import org.pcap4j.core.PcapPacket;
19 * Utility methods for generating (output) strings.
21 * @author Janus Varmarken {@literal <jvarmark@uci.edu>}
22 * @author Rahmadi Trimananda {@literal <rtrimana@uci.edu>}
24 public class PrintUtils {
27 * This is the path for writing the list of list of packet pairs {@code List<List<PcapPacket>>} into a file.
28 * The packet pairs are the pairs in one cluster, so the list represents a cluster that has been derived through
29 * the DBSCAN algorithm.
31 * E.g., this file could contain a list like the following:
33 * [[1109, 613],[1111, 613],[1115, 613],...]
35 * This list has lists of PcapPacket pairs as its members. We do not maintain the pairs in the form of
36 * {@code Cluster<PcapPacketPair>} objects because there might be a situation where we could combine multiple
37 * PcapPacketPair objects into a longer signature, i.e., a string of PcapPacket objects and not just a pair.
39 private static final String SERIALIZABLE_FILE_PATH = "./signature.sig";
41 private PrintUtils() { /* private constructor to prevent instantiation */ }
44 * Write the list of list of packet pairs {@code List<List<PcapPacket>>} into a file.
46 * After the DBSCAN algorithm derives the clusters from pairs, we save the signature in the form of list of
47 * packet pairs. We harvest the pairs and transform them back into a list of PcapPacket objects.
48 * We do not maintain the pairs in the form of {@code Cluster<PcapPacketPair>} objects because there might be
49 * a situation where we could combine multiple PcapPacketPair objects into a longer signature, i.e., a string of
50 * PcapPacket objects and not just a pair.
52 * @param fileName The path of the file in {@link String}. We could leave this one {@code null} if we wanted the
53 * default file name {@code SERIALIZABLE_FILE_PATH}.
54 * @param clusterPackets The {@link Cluster} objects in the form of list of {@code PcapPacket} objects.
56 public static void serializeClustersIntoFile(String fileName, List<List<PcapPacket>> clusterPackets) {
58 fileName = SERIALIZABLE_FILE_PATH;
59 try (ObjectOutputStream oos =
60 new ObjectOutputStream(new FileOutputStream(fileName))) {
61 oos.writeObject(clusterPackets);
62 } catch (Exception ex) {
68 * Read the list of list of packet pairs {@code List<List<PcapPacket>>} from a file.
70 * After the DBSCAN algorithm derives the clusters from pairs, we save the signature in the form of list of
71 * packet pairs. We harvest the pairs and transform them back into a list of PcapPacket objects.
72 * We do not maintain the pairs in the form of {@code Cluster<PcapPacketPair>} objects because there might be
73 * a situation where we could combine multiple PcapPacketPair objects into a longer signature, i.e., a string of
74 * PcapPacket objects and not just a pair.
76 * @param fileName The path of the file in {@link String}. We could leave this one {@code null} if we wanted the
77 * default file name {@code SERIALIZABLE_FILE_PATH}.
78 * @return The list of list of {@link Cluster} objects ({@code List<List<PcapPacket>>}) that is read from file.
80 public static List<List<PcapPacket>> serializeClustersFromFile(String fileName) {
82 fileName = SERIALIZABLE_FILE_PATH;
83 List<List<PcapPacket>> ppListOfList = null;
84 try (ObjectInputStream ois =
85 new ObjectInputStream(new FileInputStream(fileName))) {
86 ppListOfList = (List<List<PcapPacket>>) ois.readObject();
87 } catch (Exception ex) {
95 * Converts a {@code PcapPacketPair} into a CSV string containing the packet lengths of the two packets in the pair.
97 * For example, the resulting string will be "123, 456" if the first packet of the pair has a length of 123 and the
98 * second packet of the pair has a length of 456.
100 * <b>Note:</b> if the {@link PcapPacketPair} has no second element, 0 is printed as the length of the second packet
103 * @return a CSV string containing the packet lengths of the two packets of the given {@code PcapPacketPair}.
105 public static String toCsv(PcapPacketPair packetPair) {
106 return String.format("%d, %d", packetPair.getFirst().getOriginalLength(),
107 packetPair.getSecond().map(pp -> pp.getOriginalLength()).orElse(0));
111 * Converts a {@code PcapPacketPair} into a CSV string containing the packet lengths of the two packets in the pair
112 * followed by the source of each packet. The source will be a (set of) hostname(s) if the source IP can be resolved
113 * to a (set of) hostname(s) using the provided {@link DnsMap}.
115 * For example, the resulting string will be "123, 456, 192.168.1.42, domain.com" if the first packet of the pair
116 * has a length of 123, the second packet of the pair has a length of 456, the first packet of the pair the pair has
117 * a source IP of '192.168.1.42' that cannot be resolved to a hostname, and the second packet of the pair has an IP
118 * that resolves to 'domain.com'.
120 * <b>Note:</b> if the {@link PcapPacketPair} has no second element, 0 is printed as the length of the second packet
121 * in the pair, and null is printed for its source.
123 * @return a CSV string containing the packet lengths of the two packets of the given {@code PcapPacketPair} as well
124 * as their respective sources.
126 public static String toCsv(PcapPacketPair packetPair, DnsMap ipHostnameMappings) {
127 // First obtain source IPs
128 String firstSrc = PcapPacketUtils.getSourceIp(packetPair.getFirst());
129 // Note: use optional for second item in pair as there might not be one.
130 Optional<String> secondSrc = packetPair.getSecond().map(pkt -> PcapPacketUtils.getSourceIp(pkt));
132 // If possible, map source IPs to hostnames.
133 Set<String> firstHostnames = ipHostnameMappings.getHostnamesForIp(firstSrc);
134 Optional<Set<String>> secondHostnames = secondSrc.map(src -> ipHostnameMappings.getHostnamesForIp(src));
135 final String delimiter = " ";
136 if (firstHostnames != null) {
137 // If one IP maps to multiple hostnames, we concatenate the hostnames (separated by a delimiter)
138 firstSrc = firstHostnames.stream().collect(Collectors.joining(delimiter));
140 // If one IP maps to multiple hostnames, we concatenate the hostnames (separated by a delimiter)
141 Optional<String> hostnames = secondHostnames.map(hostnameSet -> hostnameSet.stream().collect(Collectors.joining(delimiter)));
142 // Fall back to IP if we couldn't second pair is present, but we couldn't map to (a) hostname(s).
143 secondSrc = hostnames.isPresent() ? hostnames : secondSrc;
145 // Check if the first source is C (client) or S (server)
146 String firstSrcCorS = packetPair.isFirstClient() ? "C" : "S";
147 String secondSrcCorS = packetPair.isSecondClient() ? "C" : "S";
149 return String.format("%d, %d, %s, %s, %s, %s", packetPair.getFirst().getOriginalLength(),
150 packetPair.getSecond().map(pp -> pp.getOriginalLength()).orElse(0),
152 secondSrc.orElse("null"),
158 * Generate a string that summarizes/describes {@code cluster}.
159 * @param cluster The {@link Cluster} to summarize/describe.
160 * @return A string that summarizes/describes {@code cluster}.
162 public static String toSummaryString(Cluster<PcapPacketPair> cluster) {
163 StringBuilder sb = new StringBuilder();
164 for (PcapPacketPair ppp : cluster.getPoints()) {
165 sb.append(toCsv(ppp, ppp.getDnsMap()) + System.lineSeparator());
167 return sb.toString();