Package org.archive.io

Examples of org.archive.io.ArchiveRecord

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
        Iterator<ArchiveRecord> iterator = archiveReader.iterator();
        int max = 300;
        int count = 0;
        int validRecords = 0;
        while (count++ < max && iterator.hasNext()) {
            ArchiveRecord archiveRecord = iterator.next();
            ArchiveRecordHeader header = archiveRecord.getHeader();
            String url = header.getUrl();

            String protocol = "";
            try {
                protocol = new URL(url).getProtocol();
            } catch (MalformedURLException e) {
                // Ignore and skip
            }

            if (protocol.equals("http")) {
                validRecords += 1;
                int contentOffset = header.getContentBegin();
                long totalLength = header.getLength();
                int contentLength = (int) totalLength - contentOffset;

                archiveRecord.skip(contentOffset);
                byte[] content = new byte[contentLength];
                archiveRecord.read(content);

                String mimetype = header.getMimetype();
                // The Arc headers != HTTP headers, but it's at least some data we can jam
                // into the FetchedDatum as a test. Note that the Arc headers will have value
                // types other than a long, so we have do to the conversion.
View Full Code Here
235
236
237
238
239
240
241
242
243
244
245
246
247
            }
        }
        reader.close();
       
        reader = ARCReaderFactory.get(arcFile, offset);
        ArchiveRecord ar = reader.get();
        assertEquals(ar.getHeader().getUrl(), url);
        ar.close();
        reader.close();
       
        // Get reader again.  See how iterator works with offset
        reader = ARCReaderFactory.get(arcFile, offset);
        int count = 0;
View Full Code Here
662
663
664
665
666
667
668
669
670
671
        File arcFile = writeRecords("writeRecord", true,
            DEFAULT_MAX_ARC_FILE_SIZE, recordCount);
        ARCReader reader = ARCReaderFactory.get(arcFile);
        Iterator<ArchiveRecord> it = reader.iterator();
        while (it.hasNext()) {
            ArchiveRecord next = it.next();
            next.close();
        }
        reader.close();
    }
View Full Code Here
40
41
42
43
44
45
46
47
48
49
50
51
        ArchiveReader reader = null;
        try {
            reader = ArchiveReaderFactory.
                get(new URL("file:////" + arc.getAbsolutePath()));
            for (Iterator<ArchiveRecord> i = reader.iterator(); i.hasNext();) {
                ArchiveRecord r = (ArchiveRecord)i.next();
                assertTrue("mime unread",StringUtils.isNotBlank(r.getHeader().getMimetype()));
            }
        } finally {
            if (reader != null) {
                reader.close();
            }
View Full Code Here
60
61
62
63
64
65
66
67
68
69
70
71
        File arc = ARCWriterTest.createARCFile(getTmpDir(), true);
        ArchiveReader reader = null;
        try {
            reader = ArchiveReaderFactory.get(arc.getAbsoluteFile());
            for (Iterator<ArchiveRecord> i = reader.iterator(); i.hasNext();) {
                ArchiveRecord r = (ArchiveRecord)i.next();
                assertTrue("mime unread",StringUtils.isNotBlank(r.getHeader().getMimetype()));
            }
        } finally {
            if (reader != null) {
                reader.close();
            }
View Full Code Here
80
81
82
83
84
85
86
87
88
89
90
91
        File arc = ARCWriterTest.createARCFile(getTmpDir(), true);
        ArchiveReader reader = null;
        try {
            reader = ArchiveReaderFactory.get(arc.getAbsoluteFile().getAbsolutePath());
            for (Iterator<ArchiveRecord> i = reader.iterator(); i.hasNext();) {
                ArchiveRecord r = (ArchiveRecord)i.next();
                assertTrue("mime unread",StringUtils.isNotBlank(r.getHeader().getMimetype()));
            }
        } finally {
            if (reader != null) {
                reader.close();
            }
View Full Code Here
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
            String[] warcs = warcDirs.get(0).list();
            assertEquals(1, warcs.length);
            WARCReader warcReader = WARCReaderFactory.get(new File(warcDirs.get(0), warcs[0]));
            Iterator<ArchiveRecord> recordIterator = warcReader.iterator();
           
            ArchiveRecord record = recordIterator.next();
            assertEquals(WARCRecordType.warcinfo.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
           
            assertTrue(recordIterator.hasNext());
            record = recordIterator.next();
            assertEquals(WARCRecordType.response.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
            assertEquals("141", record.getHeader().getHeaderValue(CONTENT_LENGTH));
            assertEquals(expectedDigest, record.getHeader().getHeaderValue(HEADER_KEY_PAYLOAD_DIGEST));
            assertEquals(curi1.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_URI));
            assertEquals(payloadRecordIdWithBrackets, record.getHeader().getHeaderValue(HEADER_KEY_ID));
           
            assertTrue(recordIterator.hasNext());
            record = recordIterator.next();
            assertEquals(WARCRecordType.request.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
            assertEquals(curi1.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_URI));
            assertEquals(payloadRecordIdWithBrackets, record.getHeader().getHeaderValue(HEADER_KEY_CONCURRENT_TO));
           
            assertTrue(recordIterator.hasNext());
            record = recordIterator.next();
            assertEquals(WARCRecordType.metadata.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
            assertEquals(curi1.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_URI));
            assertEquals(payloadRecordIdWithBrackets, record.getHeader().getHeaderValue(HEADER_KEY_CONCURRENT_TO));
           
            // the all-important revisit record
            assertTrue(recordIterator.hasNext());
            record = recordIterator.next();
            assertEquals(WARCRecordType.revisit.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
            assertEquals(curi2.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_URI));
            assertEquals(payloadRecordIdWithBrackets, record.getHeader().getHeaderValue(HEADER_KEY_REFERS_TO));
            assertEquals(NAMED_FIELD_TRUNCATED_VALUE_LENGTH, record.getHeader().getHeaderValue(HEADER_KEY_TRUNCATED));
            assertEquals(HTTP_RESPONSE_MIMETYPE, record.getHeader().getHeaderValue(CONTENT_TYPE));
            assertEquals(expectedDigest, record.getHeader().getHeaderValue(HEADER_KEY_PAYLOAD_DIGEST));
            assertEquals(PROFILE_REVISIT_IDENTICAL_DIGEST,
                    record.getHeader().getHeaderValue(HEADER_KEY_PROFILE));
            assertEquals(curi1.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_REFERS_TO_TARGET_URI));
            assertEquals(historyStore().store.get(expectedDigest).get(A_ORIGINAL_DATE),
                    record.getHeader().getHeaderValue(HEADER_KEY_REFERS_TO_DATE));
            assertNull(record.getHeader().getHeaderValue(HEADER_KEY_REFERS_TO_FILENAME));
            assertNull(record.getHeader().getHeaderValue(HEADER_KEY_REFERS_TO_FILE_OFFSET));

            assertTrue(recordIterator.hasNext());
            record = recordIterator.next();
            assertEquals(WARCRecordType.request.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
            assertEquals(curi2.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_URI));
           
            assertTrue(recordIterator.hasNext());
            record = recordIterator.next();
            assertEquals(WARCRecordType.metadata.toString(), record.getHeader().getHeaderValue(HEADER_KEY_TYPE));
            assertEquals(curi2.getUURI().toString(), record.getHeader().getHeaderValue(HEADER_KEY_URI));

            assertFalse(recordIterator.hasNext());
           
        } finally {
            warcWriter.stop();
View Full Code Here
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
        // and the previous.
       
        for (int i = headers.size() - 1; i >= 0; i--) {
            reader = WARCReaderFactory.get(f);
            ArchiveRecordHeader h = (ArchiveRecordHeader)headers.get(i);
            ArchiveRecord r = reader.get(h.getOffset());
            String mimeType = r.getHeader().getMimetype();
            assertTrue("Record is bogus",
                mimeType != null && mimeType.length() > 0);
            reader.close();
        }
       
        assertTrue("Metadatas not equal", headers.size() == recordCount);
        for (Iterator<ArchiveRecordHeader> i = headers.iterator(); i.hasNext();) {
            ArchiveRecordHeader r = (ArchiveRecordHeader)i.next();
            assertTrue("Record is empty", r.getLength() > 0);
        }
    }
View Full Code Here
332
333
334
335
336
337
338
339
340
341
342
343
344
            }
        }
        reader.close();
       
        reader = WARCReaderFactory.get(f, offset);
        ArchiveRecord ar = reader.get();
        assertEquals(ar.getHeader().getUrl(), url);
        ar.close();
        reader.close();
       
        // Get reader again.  See how iterator works with offset
        reader = WARCReaderFactory.get(f, offset);
        int count = 0;
View Full Code Here
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
   
    protected int iterateRecords(WARCReader r)
    throws IOException {
        int count = 0;
        for (Iterator<ArchiveRecord> i = r.iterator(); i.hasNext();) {
            ArchiveRecord ar = i.next();
            ar.close();
            if (count != 0) {
                assertTrue("Unexpected URL " + ar.getHeader().getUrl(),
                    ar.getHeader().getUrl().equals(SOME_URL));
            }
            count++;
        }
        return count;
    }
View Full Code Here
TOP

Related Classes of org.archive.io.ArchiveRecord

  • bixo.pipes.ParsePipeTest
  • org.archive.io.arc.ARCWriterTest
  • org.archive.io.ArchiveReaderFactoryTest
  • org.archive.io.warc.WARCWriterTest
  • org.archive.modules.recrawl.ContentDigestHistoryTest
    • 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.