001    /**
002     *  Licensed to the Apache Software Foundation (ASF) under one or more
003     *  contributor license agreements.  See the NOTICE file distributed with
004     *  this work for additional information regarding copyright ownership.
005     *  The ASF licenses this file to You under the Apache License, Version 2.0
006     *  (the "License"); you may not use this file except in compliance with
007     *  the License.  You may obtain a copy of the License at
008     *
009     *     http://www.apache.org/licenses/LICENSE-2.0
010     *
011     *  Unless required by applicable law or agreed to in writing, software
012     *  distributed under the License is distributed on an "AS IS" BASIS,
013     *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014     *  See the License for the specific language governing permissions and
015     *  limitations under the License.
016     */
017    package org.apache.geronimo.tomcat;
018    
019    import org.apache.geronimo.gbean.GBeanInfo;
020    import org.apache.geronimo.gbean.GBeanInfoBuilder;
021    import org.apache.geronimo.system.serverinfo.ServerInfo;
022    import org.apache.commons.logging.Log;
023    import org.apache.commons.logging.LogFactory;
024    import org.apache.catalina.valves.AccessLogValve;
025    
026    import java.util.ArrayList;
027    import java.util.Collection;
028    import java.util.Date;
029    import java.util.Iterator;
030    import java.util.LinkedList;
031    import java.util.List;
032    import java.util.regex.Pattern;
033    import java.util.regex.Matcher;
034    import java.io.BufferedReader;
035    import java.io.File;
036    import java.io.FileInputStream;
037    import java.io.FilenameFilter;
038    import java.io.IOException;
039    import java.io.InputStreamReader;
040    import java.text.SimpleDateFormat;
041    import java.text.ParseException;
042    
043    /**
044     * Tomcat implementation of the WebAccessLog management interface.
045     *
046     * @version $Rev: 706640 $ $Date: 2008-10-21 14:44:05 +0000 (Tue, 21 Oct 2008) $
047     */
048    public class TomcatLogManagerImpl implements TomcatLogManager {
049        private final static Log log = LogFactory.getLog(TomcatLogManagerImpl.class);
050    
051        // Pattern that matches the date in the logfile name
052        private final static Pattern FILENAME_DATE_PATTERN = Pattern.compile("[-_ /.](((19|20)\\d\\d)[-_ /.](0[1-9]|1[012])[-_ /.](0[1-9]|[12][0-9]|3[01]))");
053        private final static int GROUP_FILENAME_FULL_DATE = 1;
054        private final static int GROUP_FILENAME_YEAR  = 2;
055        private final static int GROUP_FILENAME_MONTH = 4;
056        private final static int GROUP_FILENAME_DAY   = 5;
057        // NOTE:  The file separators are specified here rather than using something like File.separator because
058        //        they are hard coded in config plans and sometimes in java code itself rather than being dependent
059        //        upon the OS.  This should be fixed someday, but for now we will manually check for either format.
060        private final static String FILE_SEPARATOR_UNIX_STYLE = "/";
061        private final static String FILE_SEPARATOR_WIN_STYLE = "\\";
062    
063        // Pattern that matches a single line  (used to calculate line numbers)
064        private final static Pattern FULL_LINE_PATTERN = Pattern.compile("^.*", Pattern.MULTILINE);
065        private final static Pattern ACCESS_LOG_PATTERN = Pattern.compile("(\\S*) (\\S*) (\\S*) \\[(.*)\\] \\\"(\\S*) (\\S*).*?\\\" (\\S*) (\\S*).*");
066        private final static int GROUP_HOST = 1;
067        private final static int GROUP_USER = 3;
068        private final static int GROUP_DATE = 4;
069        private final static int GROUP_METHOD = 5;
070        private final static int GROUP_URI = 6;
071        private final static int GROUP_RESPONSE_CODE = 7;
072        private final static int GROUP_RESPONSE_LENGTH = 8;
073        private final static String ACCESS_LOG_DATE_FORMAT = "dd/MMM/yyyy:HH:mm:ss ZZZZ";
074        private final static String LOG_FILE_NAME_FORMAT = "yyyy-MM-dd";
075        private final Collection logGbeans;
076        private final ServerInfo serverInfo;
077    
078        public TomcatLogManagerImpl(ServerInfo serverInfo, Collection logGbeans) {
079            this.serverInfo = serverInfo;
080            this.logGbeans = logGbeans;
081        }
082    
083        /**
084         * Gets the name of all logs used by this system.  Typically there
085         * is only one, but specialized cases may use more.
086         *
087         * @return An array of all log names
088         *
089         */
090        public String[] getLogNames() {
091            List logNames = new ArrayList();
092            for (Iterator it = logGbeans.iterator(); it.hasNext();) {
093                ValveGBean logGBean = (ValveGBean) it.next();
094                AccessLogValve logFile = (AccessLogValve) logGBean.getInternalObject();
095                if(logFile != null) {
096                    logNames.add( "var/catalina/logs/"+logFile.getPrefix()+LOG_FILE_NAME_FORMAT+logFile.getSuffix());    
097                }
098            }
099            return (String[]) logNames.toArray(new String[logNames.size()]);
100        }
101    
102        /**
103         * Gets the names of all log files for this log name.  
104         *
105         * @param logName The name of the log for which to return the specific file names.
106         *
107         * @return An array of log file names
108         *
109         */
110        public String[] getLogFileNames(String logName) {
111            List names = new ArrayList();
112    
113            // Find all the files for this logName
114            File[] logFiles = getLogFiles(logName);
115    
116            if (logFiles !=null) {
117                for (int i = 0; i < logFiles.length; i++) {
118                    names.add(logFiles[i].getName());
119                }
120            }
121            return (String[]) names.toArray(new String[names.size()]);
122        }
123    
124        /**
125         * Gets the name of all log files used by this log.  Typically there
126         * is only one, but specialized cases may use more.
127         *
128         * @param logName The name of the log for which to return the specific files.
129         *
130         * @return An array of all log file names
131         *
132         */
133        private File[] getLogFiles(String logName) {
134            File[] logFiles = null;
135    
136            try {
137                String fileNamePattern = logName;
138                if (fileNamePattern.indexOf(FILE_SEPARATOR_UNIX_STYLE) > -1) {
139                    fileNamePattern = fileNamePattern.substring(fileNamePattern.lastIndexOf(FILE_SEPARATOR_UNIX_STYLE) + 1);
140                } else if (fileNamePattern.indexOf(FILE_SEPARATOR_WIN_STYLE) > -1) {
141                    fileNamePattern = fileNamePattern.substring(fileNamePattern.lastIndexOf(FILE_SEPARATOR_WIN_STYLE) + 1);
142                }
143    
144                String logFile = serverInfo.resolveServerPath(logName);
145    
146                File parent = new File(logFile).getParentFile();
147    
148                if (parent != null) {
149                    logFiles = parent.listFiles(new PatternFilenameFilter(fileNamePattern));
150                }
151            } catch (Exception e) {
152                log.error("Exception attempting to locate Tomcat log files", e);
153                logFiles = new File[0];
154            }
155            return logFiles;
156        }
157    
158        /**
159         * Searches the log for records matching the specified parameters.  The
160         * maximum results returned will be the lesser of 1000 and the
161         * provided maxResults argument.
162         *
163         * @see #MAX_SEARCH_RESULTS
164         */
165        public SearchResults getMatchingItems(String logName, String host, String user, String method, String uri,
166                                              Date startDate, Date endDate, Integer skipResults, Integer maxResults) {
167    
168            // Clean up the arguments so we know what we've really got
169            if(host != null && host.equals("")) host = null;
170            if(user != null && user.equals("")) user = null;
171            if(method != null && method.equals("")) method = null;
172            if(uri != null && uri.equals("")) uri = null;
173    
174            long start = startDate == null ? 0 : startDate.getTime();
175            long end = endDate == null ? 0 : endDate.getTime();
176    
177            List list = new LinkedList();
178            boolean capped = false;
179            int lineCount = 0, fileLineCount = 0;
180    
181            // Find all the files for this logName
182            File logFiles[] = getLogFiles(logName);
183    
184            if (logFiles !=null) {
185                for (int i = 0; i < logFiles.length; i++) {
186                    fileLineCount = 0;
187                    FileInputStream logInputStream = null;
188                    try {
189                        // Obtain the date for the current log file
190                        String fileName = logFiles[i].getName();
191                        Matcher fileDate = FILENAME_DATE_PATTERN.matcher(fileName);
192                        fileDate.find();
193                        SimpleDateFormat simpleFileDate = new SimpleDateFormat(LOG_FILE_NAME_FORMAT);
194                        long logFileTime = simpleFileDate.parse(fileDate.group(GROUP_FILENAME_FULL_DATE)).getTime();
195    
196                        // Check if the dates are null (ignore) or fall within the search range
197                        if (  (start==0 && end==0)
198                           || (start>0 && start<=logFileTime && end>0 && end>=logFileTime)) {
199    
200                            // It's in the range, so process the file
201                            logInputStream = new FileInputStream(logFiles[i]);
202                            BufferedReader reader = new BufferedReader(new InputStreamReader(logInputStream, "US-ASCII"));
203    
204                            Matcher target = ACCESS_LOG_PATTERN.matcher("");
205                            SimpleDateFormat format = (start == 0 && end == 0) ? null : new SimpleDateFormat(ACCESS_LOG_DATE_FORMAT);
206                            int max = maxResults == null ? MAX_SEARCH_RESULTS : Math.min(maxResults.intValue(), MAX_SEARCH_RESULTS);
207    
208                            String line;
209                            while ((line = reader.readLine()) != null) {
210                                ++lineCount;
211                                ++fileLineCount;
212                                if(capped) {
213                                    continue;
214                                }
215                                target.reset(line);
216                                if(target.find()) {
217                                    if(host != null && !host.equals(target.group(GROUP_HOST))) {
218                                        continue;
219                                    }
220                                    if(user != null && !user.equals(target.group(GROUP_USER))) {
221                                        continue;
222                                    }
223                                    if(method != null && !method.equals(target.group(GROUP_METHOD))) {
224                                        continue;
225                                    }
226                                    if(uri != null && !target.group(GROUP_URI).startsWith(uri)) {
227                                        continue;
228                                    }
229                                    if(format != null) {
230                                        try {
231                                            long entry = format.parse(target.group(GROUP_DATE)).getTime();
232                                            if(start > entry) {
233                                                continue;
234                                            }
235                                            if(end > 0 && end < entry) {
236                                                continue;
237                                            }
238                                        } catch (ParseException e) {
239                                            // can't read the date, guess this record counts.
240                                        }
241                                    }
242                                    if(skipResults != null && skipResults.intValue() > lineCount) {
243                                        continue;
244                                    }
245                                    if(list.size() > max) {
246                                        capped = true;
247                                        continue;
248                                    }
249                                    list.add(new LogMessage(fileLineCount,line.toString()));
250                                }
251                            }
252                        }
253                    } catch (Exception e) {
254                        log.error("Unexpected error processing logs", e);
255                    } finally {
256                        if (logInputStream != null) {
257                            try {
258                                logInputStream.close();
259                            } catch (IOException e) {
260                                // ignore
261                            }
262                        }
263                    }
264                }
265            }
266            return new SearchResults(lineCount, (LogMessage[]) list.toArray(new LogMessage[list.size()]), capped);
267        }
268    
269    
270        public static final GBeanInfo GBEAN_INFO;
271    
272        static {
273            GBeanInfoBuilder infoFactory = GBeanInfoBuilder.createStatic("Tomcat Log Manager", TomcatLogManagerImpl.class);
274            infoFactory.addReference("LogGBeans", ValveGBean.class);
275            infoFactory.addReference("ServerInfo", ServerInfo.class, "GBean");
276            infoFactory.addInterface(TomcatLogManager.class);
277    
278            infoFactory.setConstructor(new String[]{"ServerInfo","LogGBeans"});
279            GBEAN_INFO = infoFactory.getBeanInfo();
280        }
281    
282        public static GBeanInfo getGBeanInfo() {
283            return GBEAN_INFO;
284        }
285    
286        /*
287         * Static inner class implementation of java.io.Filename. This will help us
288         * filter for only the files that we are interested in.
289         */
290        static class PatternFilenameFilter implements FilenameFilter {
291            Pattern pattern;
292    
293            PatternFilenameFilter(String fileNamePattern) {
294                fileNamePattern = fileNamePattern.replaceAll("yyyy", "\\\\d{4}");
295                fileNamePattern = fileNamePattern.replaceAll("yy", "\\\\d{2}");
296                fileNamePattern = fileNamePattern.replaceAll("mm", "\\\\d{2}");
297                fileNamePattern = fileNamePattern.replaceAll("MM", "\\\\d{2}");
298                fileNamePattern = fileNamePattern.replaceAll("dd", "\\\\d{2}")
299                        + ".*";
300                this.pattern = Pattern.compile(fileNamePattern);
301            }
302    
303            public boolean accept(File file, String fileName) {
304                return pattern.matcher(fileName).matches();
305            }
306        }
307    
308    /*
309        public static void main(String[] args) {
310            String jetty = "127.0.0.1 - - [07/Sep/2005:19:54:41 +0000] \"GET /console/ HTTP/1.1\" 302 0 \"-\" \"Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.10) Gecko/20050715 Firefox/1.0.6 SUSE/1.0.6-4.1\" -";
311            String tomcat = "127.0.0.1 - - [07/Sep/2005:15:51:18 -0500] \"GET /console/portal/server/server_info HTTP/1.1\" 200 11708";
312    
313            SimpleDateFormat format = new SimpleDateFormat("dd/MMM/yyyy:HH:mm:ss ZZZZ");
314            try {
315                Pattern p = Pattern.compile("(\\S*) (\\S*) (\\S*) \\[(.*)\\] \\\"(\\S*) (\\S*).*?\\\" (\\S*) (\\S*).*");
316                Matcher m = p.matcher(jetty);
317                if(m.matches()) {
318                    System.out.println("Group 1: "+m.group(1)); // client
319                    System.out.println("Group 2: "+m.group(2)); // ?? server host?
320                    System.out.println("Group 3: "+m.group(3)); // username
321                    System.out.println("Group 4: "+format.parse(m.group(4))); // date
322                    System.out.println("Group 5: "+m.group(5)); // method
323                    System.out.println("Group 5: "+m.group(6)); // URI
324                    System.out.println("Group 6: "+m.group(7)); // response code
325                    System.out.println("Group 7: "+m.group(8)); // response length
326                } else {
327                    System.out.println("No match");
328                }
329                m = p.matcher(tomcat);
330                if(m.matches()) {
331                    System.out.println("Group 1: "+m.group(1));
332                    System.out.println("Group 2: "+m.group(2));
333                    System.out.println("Group 3: "+m.group(3));
334                    System.out.println("Group 4: "+format.parse(m.group(4)));
335                    System.out.println("Group 5: "+m.group(5)); // method
336                    System.out.println("Group 5: "+m.group(6)); // URI
337                    System.out.println("Group 6: "+m.group(7)); // response code
338                    System.out.println("Group 7: "+m.group(8)); // response length
339                } else {
340                    System.out.println("No match");
341                }
342            } catch (ParseException e) {
343                e.printStackTrace();
344            }
345        }
346    */
347    }