Package org.eclipse.core.internal.runtime

Examples of org.eclipse.core.internal.runtime.HashMapOfString

38
39
40
41
42
43
44
45
46
47
48
  /**
   * Crawler configuration bean class
   */
  public CrawlConfig(boolean init)
   {
    paramHash = new HashMapOfString();
    String defaultIndexDir = Constants.MUSTRU_HOME + Constants.fs + "data" + Constants.fs + "index";
    String defaultDbDir = Constants.MUSTRU_HOME + Constants.fs + "data" + Constants.fs + "bdb";
   
    //*-- initialize the crawler configuration with default values
    if (init)
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
   String filtersFile = Constants.FILTER_FILE;    //*-- file containing the list of handlers
   String errMsg = null;        //*-- Error message
   RandomAccessFile tfile = null;       //*-- a random access file containing a list of files to scan
   BufferedReader filein = null;      //*-- reader to remove dups from the task file       
   PrintWriter outp = null;        //*-- writer to create the unique list of files
   HashMapOfString h = null;        //*-- hash set to save the list of files
  
   logger.info("Start creating task file for crawl");
  
   //*-- before creating a task file, check for a restart
   if (crawlConfig.getStartPosition() != -1)
   { ctRef.fileReadTime -= new Date().getTime();
     int numFiles = 0; LineNumberReader fileno = null;
     try
     { fileno = new LineNumberReader(new FileReader(taskFile));
       while ((fileno.readLine()) != null) numFiles++;
     }
     catch (IOException ie) { logger.error("IO Error in task file " + taskFile + " " + ie.getMessage()); }
     finally { try { if (fileno != null) fileno.close(); }
               catch (IOException ie) { logger.error("Ignore error"); } }
     setNumFiles(numFiles);
     ctRef.fileReadTime += new Date().getTime();
     return;
   }
  
   ctRef.fileReadTime -= new Date().getTime();
   String iDirs = crawlConfig.getIncDirs();
   String eDirs = crawlConfig.getExcDirs();
   boolean skipHidden = crawlConfig.isSkipHidden();
   boolean followLinks = crawlConfig.isFollowLinks();
 
   if ( !(iDirs.equals("")) && !(iDirs.endsWith(";")) ) iDirs += ";";
   if ( !(eDirs.equals("")) && !(eDirs.endsWith(";")) ) eDirs += ";";
   String[] incDirs = (iDirs.equals("") ) ? new String[0]: iDirs.split(";");
   String[] excDirs = (eDirs.equals("") ) ? new String[0]: eDirs.split(";");
   Pattern[] excPatts = new Pattern[excDirs.length];
   for (int i = 0; i < excDirs.length; i++) excPatts[i] = Pattern.compile("^" + excDirs[i] + ".*$");
    
   try
    {
     tfile = new RandomAccessFile(taskFile, "rw");
     tfile.setLength(0); tfile.close();

     // *-- Run the directory scan for each directory in the list
     DirTools dirTools = new DirTools();
     for (int i = 0; i < incDirs.length; i++)
      { logger.info("Start scanning " + incDirs[i]);
        dirTools.dirScan(incDirs[i], filtersFile, taskFile, skipHidden, followLinks);
        logger.info("Finished scanning " + incDirs[i]);
      }

     // *-- Read the task file into a hashset and remove dups.
     logger.info("Removing duplicate file names from task file");
     filein = new BufferedReader(new FileReader(taskFile));
     String s; h = new HashMapOfString();
     while ((s = filein.readLine()) != null) h.put(s, "");
    }
   catch (IOException ie)
     { errMsg = "IO Error in task file " + ie.getMessage(); }
   finally
    { if (filein != null) try { filein.close(); } catch (IOException ie) { logger.error("Ignore error"); } 
      if (errMsg != null) ctRef.cleanUp(errMsg);
    }
     
   //*-- Dump the file back into tfile with a list of unique file names
   //*-- Check if any of the files belong to the list of exclusion directories
   try
   {
    tfile = new RandomAccessFile(taskFile, "rw");
    tfile.setLength(0); tfile.seek(0);
    outp = new PrintWriter(new FileWriter(taskFile));
   
    if (h != null)
     { String[] keys = h.keys();
       FLOOP: for (int i = 0; i < keys.length; i++)
        { for (int j = 0; j < excPatts.length; j++)     
           { Matcher matcher = excPatts[j].matcher(keys[i]);
             if (matcher.matches()) continue FLOOP;
          }
View Full Code Here
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
  Properties props = new Properties();
  try { props.load(new FileInputStream( Constants.DOCTYPES_FILE)); }
  catch ( IOException e) { logger.error("Could not open " + Constants.DOCTYPES_FILE + " " + e.getMessage()); }

  //*-- load the types of media from the docTypes properties file
  HashMapOfString stats = new HashMapOfString();
  String[] docTypes = new String[props.size()]; int j = 0;
  for (Enumeration keys = props.propertyNames(); keys.hasMoreElements(); )
  { String key = (String) keys.nextElement(); stats.put(key, "0" );
  docTypes[j++] = key; }

  //*-- scan the database by file name and keep track of the types of files in the index
  DbTools dbt = Constants.getDbt();  
  dbt.openDB(Constants.EXT_FILES_DB, true, false);
  Cursor cursor = null; IndexableDoc idoc = new IndexableDoc();
  try
  {
   cursor = dbt.getCurrentDB().openCursor(null, null);
   DatabaseEntry key = new DatabaseEntry();
   DatabaseEntry data = new DatabaseEntry(); int i = 0;
   LOOP: while (cursor.getNext(key, data, LockMode.DEFAULT) == OperationStatus.SUCCESS)
   {
    if (!running) break LOOP;
    idoc = (IndexableDoc) idoc.getBdbBinding().entryToObject(data);
    String docType = idoc.getFileType();
    if (docType == null) docType = "unknown";
    String num =  stats.get(docType);
    if ( num == null) { docType = "unknown"; num =  stats.get(docType); }
    int val =  Integer.parseInt(num) + 1;
   
    i++; if ( (i % 1000) == 0) System.out.println("finished " + i + " documents");
   
    stats.put(docType, String.valueOf(val) );
   } //*-- end of while

  } //*-- end of try
  catch (DatabaseException dbe)
  { logger.error("Could not open cursor to browse File index" + dbe.getMessage() ); }
  finally
  { try { if (cursor != null) cursor.close();  dbt.closeDB(); }
  catch (DatabaseException de) { logger.error("Ignore error"); }
  }

  StringBuffer out = new StringBuffer(); int numFiles = 0;
  int io;
  for (int i = 0; i < docTypes.length; i++)
  { io =   Integer.parseInt( stats.get( docTypes[i] ) );
  out.append("No. of " + docTypes[i] + " files: " + io + Constants.NEWLINE);
  numFiles += io;
  }
  out.append("Total no. of files: " + numFiles + Constants.NEWLINE); 

View Full Code Here
TOP

Related Classes of org.eclipse.core.internal.runtime.HashMapOfString

  • org.sf.mustru.crawl.CrawlConfig
  • org.sf.mustru.crawl.CrawlTask
  • org.sf.mustru.utils.IndexTools
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.