Package org.archive.io.arc

Examples of org.archive.io.arc.ARCWriter

202
203
204
205
206
207
208
209
210
211
212
      throw e;
    }
   
    // store URL
    ARCLocation location = null;
    ARCWriter writer = null;
    try {
      writer = cache.getWriter();
      location = storeFile(tmpFile, writer, urlString, method);
    } catch(IOException e) {
      e.printStackTrace();
View Full Code Here
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
    if(!arcDir.isDirectory()) {
      arcDir.mkdir();
    }
    File [] files = {arcDir};
    boolean compress = true;
    ARCWriter writer = new ARCWriter(new AtomicInteger(),
        Arrays.asList(files), "test", compress,
        DEFAULT_MAX_ARC_FILE_SIZE);
    Properties p = new Properties();
    p.setProperty(ARCCacheDirectory.LIVE_WEB_ARC_DIR, args[0]);
    p.setProperty(ARCCacheDirectory.LIVE_WEB_ARC_PREFIX, "test");
    p.setProperty(CACHE_PATH, arcDir.getAbsolutePath());

    URLCacher uc = new URLCacher();
    ARCCacheDirectory cache = new ARCCacheDirectory();
//    try {
////      cache.init(p);
////      uc.init(p);
//    } catch (ConfigurationException e) {
//      e.printStackTrace();
//      System.exit(1);
//    }
    for(int k = 1; k < args.length; k++) {
      try {
        url = new URL(args[k]);
      } catch (MalformedURLException e1) {
        e1.printStackTrace();
        continue;
      }
      try {
        uc.cache(cache, url.toString());
      } catch (URIException e) {
        e.printStackTrace();
      } catch (LiveDocumentNotAvailableException e) {
        e.printStackTrace();
      } catch (IOException e) {
        e.printStackTrace();
      }
    }
    try {
      writer.close();
    } catch (IOException e) {
      e.printStackTrace();
    }
  }
View Full Code Here
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
   */
  public void directoryToArc(File srcDir, File tgtDir, String prefix)
  throws IOException {
   
    File target[] = {tgtDir};
    ARCWriter writer = new ARCWriter(new AtomicInteger(),
        Arrays.asList(target),prefix,true,
        ARCConstants.DEFAULT_MAX_ARC_FILE_SIZE);
    File sources[] = srcDir.listFiles();
    logger.info("Found " + sources.length + " files in " + srcDir);
    for(int i = 0; i<sources.length; i++) {
      addFile(sources[i]);
    }
    logger.info("Associated " + sources.length + " files in " + srcDir);

    // sort keys and write them all:
    Object arr[] = components.keySet().toArray();
    Arrays.sort(arr);
    for(int i = 0; i < arr.length; i++) {
      String key = (String) arr[i];
      RecordComponents rc = components.get(key);
      rc.writeRecord(writer,srcDir);
      logger.info("Wrote record keyed " + rc.key);     
    }
    writer.close();
    logger.info("Closed arc file named " +
        writer.getFile().getAbsolutePath());
  }
View Full Code Here
82
83
84
85
86
87
88
89
90
91
92
93
       FileUtils.assertReadable(dir);
       WARCReader reader = WARCReaderFactory.get(warc);
       List<String> metadata =  new ArrayList<String>();
       metadata.add("Made from " + reader.getReaderIdentifier() + " by " +
           this.getClass().getName() + "/" + getRevision());
       ARCWriter writer =
           new ARCWriter(
                   new AtomicInteger(),
                   new WriterPoolSettingsData(
                           prefix,
                           suffix,
                           -12,
View Full Code Here
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
            setTotalBytesWritten(getTotalBytesWritten() +
              (writer.getPosition() - position));
            position = writer.getPosition();
        }
       
        ARCWriter w = (ARCWriter)writer;
        try {
            if (in instanceof ReplayInputStream) {
                w.write(curi.toString(), curi.getContentType(),
                    ip, curi.getFetchBeginTime(),
                    recordLength, (ReplayInputStream)in);
            } else {
                w.write(curi.toString(), curi.getContentType(),
                    ip, curi.getFetchBeginTime(),
                    recordLength, in);
            }
        } catch (IOException e) {
            // Invalidate this file (It gets a '.invalid' suffix).
View Full Code Here
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
  public void directoryToArc(File srcDir, File tgtDir, String prefix)
  throws IOException {
   
    File target[] = {tgtDir};

    ARCWriter writer = new ARCWriter(new AtomicInteger(),
        getSettings(true,prefix,Arrays.asList(target)));
    File sources[] = srcDir.listFiles();
    LOGGER.info("Found " + sources.length + " files in " + srcDir);
    for(int i = 0; i<sources.length; i++) {
      addFile(sources[i]);
    }
    LOGGER.info("Associated " + sources.length + " files in " + srcDir);

    // sort keys and write them all:
    Object arr[] = components.keySet().toArray();
    Arrays.sort(arr);
    for(int i = 0; i < arr.length; i++) {
      String key = (String) arr[i];
      RecordComponents rc = components.get(key);
      rc.writeRecord(writer,srcDir);
      LOGGER.info("Wrote record keyed " + rc.key);     
    }
    writer.close();
    LOGGER.info("Closed arc file named " +
        writer.getFile().getAbsolutePath());
  }
View Full Code Here
190
191
192
193
194
195
196
197
198
199
200
        getMethod.releaseConnection();
      }
    }

    // now write the content, or a fake record:
    ARCWriter writer = null;
    ReplayInputStream replayIS = null;
    try {
      writer = cache.getWriter();
      if(gotUrl) {
View Full Code Here
TOP

Related Classes of org.archive.io.arc.ARCWriter

  • org.archive.io.Warc2Arc
  • org.archive.modules.writer.ARCWriterProcessor
  • org.archive.wayback.liveweb.URLCacher
  • org.archive.wayback.liveweb.URLtoARCCacher
  • org.archive.wayback.util.ARCCreator
    • 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.