1 | package de.ugoe.cs.eventbench.web;
|
---|
2 |
|
---|
3 | import java.io.FileNotFoundException;
|
---|
4 | import java.io.IOException;
|
---|
5 | import java.net.URI;
|
---|
6 | import java.net.URISyntaxException;
|
---|
7 | import java.text.ParseException;
|
---|
8 | import java.text.SimpleDateFormat;
|
---|
9 | import java.util.ArrayList;
|
---|
10 | import java.util.Collection;
|
---|
11 | import java.util.HashMap;
|
---|
12 | import java.util.LinkedList;
|
---|
13 | import java.util.List;
|
---|
14 | import java.util.Map;
|
---|
15 |
|
---|
16 | import de.ugoe.cs.eventbench.web.data.WebEvent;
|
---|
17 | import de.ugoe.cs.util.FileTools;
|
---|
18 | import de.ugoe.cs.util.console.Console;
|
---|
19 |
|
---|
20 | /**
|
---|
21 | * <p>
|
---|
22 | * Provides functionality to parse log files with web request.
|
---|
23 | * </p>
|
---|
24 | *
|
---|
25 | * @author Steffen Herbold
|
---|
26 | * @version 1.0
|
---|
27 | */
|
---|
28 | public class WeblogParser {
|
---|
29 |
|
---|
30 | /**
|
---|
31 | * <p>
|
---|
32 | * Timeout between two sessions in milliseconds.
|
---|
33 | * </p>
|
---|
34 | */
|
---|
35 | private long timeout;
|
---|
36 |
|
---|
37 | /**
|
---|
38 | * <p>
|
---|
39 | * Minimal length of a session. All shorter sessions will be pruned.<br>
|
---|
40 | * Default: 2
|
---|
41 | * </p>
|
---|
42 | */
|
---|
43 | private int minLength = 2;
|
---|
44 |
|
---|
45 | /**
|
---|
46 | * <p>
|
---|
47 | * Maximal length of a session. All longer sessions will be pruned.<br>
|
---|
48 | * Default: 100
|
---|
49 | * </p>
|
---|
50 | */
|
---|
51 | private int maxLength = 100;
|
---|
52 |
|
---|
53 | /**
|
---|
54 | * <p>
|
---|
55 | * URL of the server that generated the log that is currently parser; null
|
---|
56 | * of URL is not available.<br>
|
---|
57 | * Default: null
|
---|
58 | * </p>
|
---|
59 | */
|
---|
60 | private String url = null;
|
---|
61 |
|
---|
62 | /**
|
---|
63 | * <p>
|
---|
64 | * Collection of generated sequences.
|
---|
65 | * </p>
|
---|
66 | */
|
---|
67 | private List<List<WebEvent>> sequences;
|
---|
68 |
|
---|
69 | /**
|
---|
70 | * <p>
|
---|
71 | * Name and path of the robot filter.
|
---|
72 | * </p>
|
---|
73 | */
|
---|
74 | private static final String ROBOTFILTERFILE = "misc/robotfilter.txt";
|
---|
75 |
|
---|
76 | /**
|
---|
77 | * <p>
|
---|
78 | * Field that contains a regular expression that matches all robots
|
---|
79 | * contained in {@link #ROBOTFILTERFILE}.
|
---|
80 | * </p>
|
---|
81 | */
|
---|
82 | private String robotRegex = null;
|
---|
83 |
|
---|
84 | /**
|
---|
85 | * <p>
|
---|
86 | * Constructor. Creates a new WeblogParser with a default timeout of
|
---|
87 | * 3,600,000 milliseconds (1 hour).
|
---|
88 | * </p>
|
---|
89 | */
|
---|
90 | public WeblogParser() {
|
---|
91 | this(3600000);
|
---|
92 | }
|
---|
93 |
|
---|
94 | /**
|
---|
95 | * <p>
|
---|
96 | * Constructor. Creates a new WeblogParser.
|
---|
97 | * </p>
|
---|
98 | *
|
---|
99 | * @param timeout
|
---|
100 | * session timeout
|
---|
101 | */
|
---|
102 | public WeblogParser(long timeout) {
|
---|
103 | this.timeout = timeout;
|
---|
104 | }
|
---|
105 |
|
---|
106 | /**
|
---|
107 | * <p>
|
---|
108 | * Returns the generated event sequences.
|
---|
109 | * </p>
|
---|
110 | *
|
---|
111 | * @return generated event sequences
|
---|
112 | */
|
---|
113 | public Collection<List<WebEvent>> getSequences() {
|
---|
114 | return sequences;
|
---|
115 | }
|
---|
116 |
|
---|
117 | /**
|
---|
118 | * <p>
|
---|
119 | * Sets the session timeout.
|
---|
120 | * </p>
|
---|
121 | *
|
---|
122 | * @param timeout
|
---|
123 | * new session timeout
|
---|
124 | */
|
---|
125 | public void setTimeout(long timeout) {
|
---|
126 | this.timeout = timeout;
|
---|
127 | }
|
---|
128 |
|
---|
129 | /**
|
---|
130 | * <p>
|
---|
131 | * Sets the minimal length of a session. All sessions that contain less
|
---|
132 | * events will be pruned.
|
---|
133 | * </p>
|
---|
134 | *
|
---|
135 | * @param minLength
|
---|
136 | * new minimal length
|
---|
137 | */
|
---|
138 | public void setMinLength(int minLength) {
|
---|
139 | this.minLength = minLength;
|
---|
140 | }
|
---|
141 |
|
---|
142 | /**
|
---|
143 | * <p>
|
---|
144 | * Sets the maximal length of a session. All sessions that contain more
|
---|
145 | * events will be pruned.
|
---|
146 | * </p>
|
---|
147 | *
|
---|
148 | * @param maxLength
|
---|
149 | * new maximal length
|
---|
150 | */
|
---|
151 | public void setMaxLength(int maxLength) {
|
---|
152 | this.maxLength = maxLength;
|
---|
153 | }
|
---|
154 |
|
---|
155 | /**
|
---|
156 | * <p>
|
---|
157 | * Sets the URL of the server from which this log was generated. Often
|
---|
158 | * required for replay generation
|
---|
159 | * </p>
|
---|
160 | *
|
---|
161 | * @param url
|
---|
162 | * URL of the server
|
---|
163 | */
|
---|
164 | public void setUrl(String url) {
|
---|
165 | this.url = url;
|
---|
166 | }
|
---|
167 |
|
---|
168 | /**
|
---|
169 | * <p>
|
---|
170 | * Parses a web log file.
|
---|
171 | * </p>
|
---|
172 | *
|
---|
173 | * @param filename
|
---|
174 | * name and path of the log file
|
---|
175 | * @throws IOException
|
---|
176 | * thrown if there is a problem with reading the log file
|
---|
177 | * @throws FileNotFoundException
|
---|
178 | * thrown if the log file is not found
|
---|
179 | * @throws ParseException
|
---|
180 | * thrown the date format is invalid
|
---|
181 | */
|
---|
182 | public void parseFile(String filename) throws IOException,
|
---|
183 | FileNotFoundException, ParseException {
|
---|
184 | String[] lines = FileTools.getLinesFromFile(filename);
|
---|
185 |
|
---|
186 | Map<String, List<Integer>> cookieSessionMap = new HashMap<String, List<Integer>>();
|
---|
187 | int lastId = -1;
|
---|
188 |
|
---|
189 | SimpleDateFormat dateFormat = new SimpleDateFormat(
|
---|
190 | "yyyy-MM-dd HH:mm:ss");
|
---|
191 | loadRobotRegex();
|
---|
192 |
|
---|
193 | sequences = new ArrayList<List<WebEvent>>();
|
---|
194 |
|
---|
195 | int lineCounter = 0;
|
---|
196 | for (String line : lines) {
|
---|
197 | lineCounter++;
|
---|
198 | String[] values = line.substring(1, line.length() - 1).split(
|
---|
199 | "\" \"");
|
---|
200 |
|
---|
201 | // use cookie as session identifier
|
---|
202 | int cookieStart = values[0].lastIndexOf('.');
|
---|
203 | String cookie = values[0].substring(cookieStart + 1);
|
---|
204 | String dateString = values[1];
|
---|
205 | long timestamp = dateFormat.parse(dateString).getTime();
|
---|
206 | String uriString = values[2];
|
---|
207 | // String ref = values[3]; // referer is not yet used!
|
---|
208 | String agent;
|
---|
209 | if (values.length > 4) {
|
---|
210 | agent = values[4];
|
---|
211 | } else {
|
---|
212 | agent = "noagent";
|
---|
213 | }
|
---|
214 |
|
---|
215 | List<String> postedVars = new ArrayList<String>();
|
---|
216 | if (values.length == 6) { // post vars found
|
---|
217 | for (String postVar : values[5].trim().split(" ")) {
|
---|
218 | postedVars.add(postVar);
|
---|
219 | }
|
---|
220 | }
|
---|
221 | if (!isRobot(agent)) {
|
---|
222 | try {
|
---|
223 | URI uri = new URI(uriString);
|
---|
224 | String path = uri.getPath();
|
---|
225 | List<String> getVars = extractGetVarsFromUri(uri);
|
---|
226 |
|
---|
227 | WebEvent event = new WebEvent(url, path, timestamp,
|
---|
228 | postedVars, getVars);
|
---|
229 |
|
---|
230 | // find session and add event
|
---|
231 | List<Integer> sessionIds = cookieSessionMap.get(cookie);
|
---|
232 | if (sessionIds == null) {
|
---|
233 | sessionIds = new ArrayList<Integer>();
|
---|
234 | // start new session
|
---|
235 | sessionIds.add(++lastId);
|
---|
236 | cookieSessionMap.put(cookie, sessionIds);
|
---|
237 | sequences.add(new LinkedList<WebEvent>());
|
---|
238 | }
|
---|
239 | Integer lastSessionIndex = sessionIds
|
---|
240 | .get(sessionIds.size() - 1);
|
---|
241 | List<WebEvent> lastSession = sequences
|
---|
242 | .get(lastSessionIndex);
|
---|
243 | long lastEventTime = timestamp;
|
---|
244 | if (!lastSession.isEmpty()) {
|
---|
245 | lastEventTime = lastSession.get(lastSession.size() - 1)
|
---|
246 | .getTimestamp();
|
---|
247 | }
|
---|
248 | if (timestamp - lastEventTime > timeout) {
|
---|
249 | sessionIds.add(++lastId);
|
---|
250 | List<WebEvent> newSession = new LinkedList<WebEvent>();
|
---|
251 | newSession.add(event);
|
---|
252 | sequences.add(newSession);
|
---|
253 | } else {
|
---|
254 | lastSession.add(event);
|
---|
255 | }
|
---|
256 | } catch (URISyntaxException e) {
|
---|
257 | Console.traceln("Ignored line " + lineCounter + ": "
|
---|
258 | + e.getMessage());
|
---|
259 | }
|
---|
260 | }
|
---|
261 | }
|
---|
262 | pruneSequences();
|
---|
263 | }
|
---|
264 |
|
---|
265 | /**
|
---|
266 | * <p>
|
---|
267 | * Prunes sequences shorter than {@link #minLength}.
|
---|
268 | * </p>
|
---|
269 | */
|
---|
270 | private void pruneSequences() {
|
---|
271 | Console.traceln("" + sequences.size() + " user sequences found");
|
---|
272 | // prune sequences shorter than min-length and longer than maxLength
|
---|
273 | int i = 0;
|
---|
274 | while (i < sequences.size()) {
|
---|
275 | if ((sequences.get(i).size() < minLength)
|
---|
276 | || sequences.get(i).size() > maxLength) {
|
---|
277 | sequences.remove(i);
|
---|
278 | } else {
|
---|
279 | i++;
|
---|
280 | }
|
---|
281 | }
|
---|
282 | Console.traceln("" + sequences.size()
|
---|
283 | + " remaining after pruning of sequences shorter than "
|
---|
284 | + minLength);
|
---|
285 | }
|
---|
286 |
|
---|
287 | /**
|
---|
288 | * <p>
|
---|
289 | * Reads {@link #ROBOTFILTERFILE} and creates a regular expression that
|
---|
290 | * matches all the robots defined in the file. The regular expression is
|
---|
291 | * stored in the field {@link #robotRegex}.
|
---|
292 | * </p>
|
---|
293 | *
|
---|
294 | * @throws IOException
|
---|
295 | * thrown if there is a problem reading the robot filter
|
---|
296 | * @throws FileNotFoundException
|
---|
297 | * thrown if the robot filter is not found
|
---|
298 | */
|
---|
299 | private void loadRobotRegex() throws IOException, FileNotFoundException {
|
---|
300 | String[] lines = FileTools.getLinesFromFile(ROBOTFILTERFILE);
|
---|
301 | StringBuilder regex = new StringBuilder();
|
---|
302 | for (int i = 0; i < lines.length; i++) {
|
---|
303 | regex.append("(.*" + lines[i] + ".*)");
|
---|
304 | if (i != lines.length - 1) {
|
---|
305 | regex.append('|');
|
---|
306 | }
|
---|
307 | }
|
---|
308 | robotRegex = regex.toString();
|
---|
309 | }
|
---|
310 |
|
---|
311 | /**
|
---|
312 | * <p>
|
---|
313 | * Checks whether an agent is a robot.
|
---|
314 | * </p>
|
---|
315 | *
|
---|
316 | * @param agent
|
---|
317 | * agent that is checked
|
---|
318 | * @return true, if the agent is a robot; false otherwise
|
---|
319 | */
|
---|
320 | private boolean isRobot(String agent) {
|
---|
321 | return agent.matches(robotRegex);
|
---|
322 | }
|
---|
323 |
|
---|
324 | /**
|
---|
325 | * <p>
|
---|
326 | * Parses the URI and extracts the GET variables that have been passed.
|
---|
327 | * </p>
|
---|
328 | *
|
---|
329 | * @param uri
|
---|
330 | * URI that is parsed
|
---|
331 | * @return a list with all GET variables
|
---|
332 | */
|
---|
333 | private List<String> extractGetVarsFromUri(URI uri) {
|
---|
334 | List<String> getVars = new ArrayList<String>();
|
---|
335 | String query = uri.getQuery();
|
---|
336 | if (query != null) {
|
---|
337 | String[] paramPairs = query.split("&");
|
---|
338 | for (String paramPair : paramPairs) {
|
---|
339 | String[] paramSplit = paramPair.split("=");
|
---|
340 | getVars.add(paramSplit[0]);
|
---|
341 | }
|
---|
342 | }
|
---|
343 | return getVars;
|
---|
344 | }
|
---|
345 | }
|
---|