Package bixo.datum

Examples of bixo.datum.ScoredUrlDatum

352
353
354
355
356
357
358
359
360
361
362
    public final void testTempRedirectHandling() throws Exception {
        FetcherPolicy policy = new FetcherPolicy();
        Server server = startServer(new RedirectResponseHandler(), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/base";
        FetchedDatum result = fetcher.get(new ScoredUrlDatum(url));
        server.stop();

        assertEquals("Redirected URL", "http://localhost:8089/redirect", result.getFetchedUrl());
        assertNull(result.getNewBaseUrl());
        assertEquals(1, result.getNumRedirects());
View Full Code Here
366
367
368
369
370
371
372
373
374
375
376
377
    public final void testPermRedirectHandling() throws Exception {
        FetcherPolicy policy = new FetcherPolicy();
        Server server = startServer(new RedirectResponseHandler(true), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/base";
        ScoredUrlDatum scoredUrl = new ScoredUrlDatum(url);
        scoredUrl.setPayloadValue("payload-field-1", 1);
        FetchedDatum result = fetcher.get(scoredUrl);
        server.stop();

        assertEquals("Redirected URL", "http://localhost:8089/redirect", result.getFetchedUrl());
        assertEquals("New base URL", "http://localhost:8089/redirect", result.getNewBaseUrl());
View Full Code Here
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
        Server server = startServer(new RedirectResponseHandler(true), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/base";
       
        try {
            fetcher.get(new ScoredUrlDatum(url));
            fail("Exception should have been thrown");
        } catch (RedirectFetchException e) {
            assertEquals("Redirected URL", "http://localhost:8089/redirect", e.getRedirectedUrl());
            assertEquals(RedirectExceptionReason.PERM_REDIRECT_DISALLOWED, e.getReason());
        } finally {
            server.stop();
        }
       
        // Now try setting the mode to follow none
        policy.setRedirectMode(RedirectMode.FOLLOW_NONE);
        server = startServer(new RedirectResponseHandler(false), 8089);
        fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
       
        try {
            fetcher.get(new ScoredUrlDatum(url));
            fail("Exception should have been thrown");
        } catch (RedirectFetchException e) {
            assertEquals("Redirected URL", "http://localhost:8089/redirect", e.getRedirectedUrl());
            assertEquals(RedirectExceptionReason.TEMP_REDIRECT_DISALLOWED, e.getReason());
        } finally {
View Full Code Here
421
422
423
424
425
426
427
428
429
430
       
        FetcherPolicy policy = new FetcherPolicy();
        Server server = startServer(new LanguageResponseHandler(englishContent, foreignContent), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/";
        FetchedDatum result = fetcher.get(new ScoredUrlDatum(url));
        server.stop();
        String contentStr = new String(result.getContentBytes(), 0, result.getContentLength());
        assertTrue( englishContent.equals(contentStr));
    }
View Full Code Here
439
440
441
442
443
444
445
446
447
448
449
        Server server = startServer(new MimeTypeResponseHandler("text/xml"), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/";
       
        try {
            fetcher.get(new ScoredUrlDatum(url));
            fail("Fetch should have failed");
        } catch (AbortedFetchException e) {
            assertEquals(AbortedFetchReason.INVALID_MIMETYPE, e.getAbortReason());
        } finally {
            server.stop();
View Full Code Here
461
462
463
464
465
466
467
468
469
470
471
        Server server = startServer(new MimeTypeResponseHandler(null), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/";
       
        try {
            fetcher.get(new ScoredUrlDatum(url));
        } catch (AbortedFetchException e) {
            fail("Fetch should not have failed if no mime-type is specified");
        } finally {
            server.stop();
        }
View Full Code Here
481
482
483
484
485
486
487
488
489
490
491
        Server server = startServer(new MimeTypeResponseHandler("text/html; charset=UTF-8"), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/";
       
        try {
            fetcher.get(new ScoredUrlDatum(url));
        } catch (AbortedFetchException e) {
            fail("Fetch should have worked");
        } finally {
            server.stop();
        }
View Full Code Here
495
496
497
498
499
500
501
502
503
504
505
    public final void testHostAddress() throws Exception {
        FetcherPolicy policy = new FetcherPolicy();
        Server server = startServer(new ResourcesResponseHandler(), 8089);
        BaseFetcher fetcher = new SimpleHttpFetcher(1, policy, ConfigUtils.BIXO_TEST_AGENT);
        String url = "http://localhost:8089/simple-page.html";
        FetchedDatum result = fetcher.get(new ScoredUrlDatum(url));
        server.stop();
       
        String hostAddress = result.getHostAddress();
        assertNotNull(hostAddress);
        assertEquals("127.0.0.1", hostAddress);
View Full Code Here
31
32
33
34
35
36
37
38
39
40
41
    public final void testNoDomain() {
        BaseFetcher fetcher = new SimpleHttpFetcher(1, ConfigUtils.BIXO_IT_AGENT);
        String url = "http://www.bogusbixodomainxxxxx.com";
       
        try {
            fetcher.get(new ScoredUrlDatum(url));
            Assert.fail("Exception not thrown");
        } catch (Exception e) {
            Assert.assertTrue(e instanceof IOFetchException);
        }
    }
View Full Code Here
124
125
126
127
128
129
130
131
132
133
134
              } else {
                url = args[index++];
              }

              System.out.println("Fetching " + url);
            FetchedDatum result = fetcher.get(new ScoredUrlDatum(url));
            System.out.println(String.format("Fetched %s: headers = %s", result.getUrl(), result.getHeaders()));
            System.out.flush();
           
            // System.out.println("Result = " + result.toString());
            ParsedDatum parsed = parser.parse(result);
View Full Code Here
TOP

Related Classes of bixo.datum.ScoredUrlDatum

  • bixo.config.DefaultFetchJobPolicyTest
  • bixo.fetcher.FetchTask
  • bixo.fetcher.http.SimpleHttpFetcherIntegrationTest
  • bixo.fetcher.SimpleHttpFetcherTest
  • bixo.operations.FilterAndScoreByUrlAndRobotsTest$MatchBlockedByRobotsKey
  • bixo.operations.MakeFetchSetsBuffer
  • bixo.operations.ProcessRobotsTask
  • bixo.operations.ResolveRedirectsTask
  • bixo.pipes.FetchPipe$MakeSkippedStatus
  • bixo.pipes.FetchPipe$SplitIntoSpecialAndRegularKeys
    • 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.