Package org.archive.wayback.core

Examples of org.archive.wayback.core.WaybackRequest

192
193
194
195
196
197
198
199
200
201
  }

  private void testFindRange(final AlphaPartitionedIndex apIndex,
      final String url, final String wantGroup) throws URIException,
      BadQueryException, ResourceIndexNotAvailableException {
    WaybackRequest r = new WaybackRequest();
    r.put(WaybackConstants.REQUEST_URL,apIndex.canonicalize(url));
    RangeGroup g = apIndex.getRangeGroupForRequest(r);
    assertEquals(g.getName(),wantGroup);   
  }
View Full Code Here
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
130
131
132
      }     
     
      LOGGER.fine(String.format("Parsed format(%s) URL(%s)", format,
          urlStr));

      WaybackRequest wbRequest = new WaybackRequest();
      // ludab changes nov 30 2012 to add timemap paging

      Matcher matcher = WB_REQUEST_REGEX.matcher(urlStr);
      String startDate = getEarliestTimestamp();

      if (matcher != null && matcher.matches()) {

        String dateStr = matcher.group(1);
        urlStr = matcher.group(4);
        wbRequest.put(PAGE_STARTS, dateStr);

        if (dateStr.length() == 0) {
          startDate = getEarliestTimestamp();

        } else {
          startDate = Timestamp.parseAfter(dateStr).getDateStr();

        }

      }

      if (wbRequest.getStartTimestamp() == null) {
        // ludab nov30 timemap paging
        wbRequest.setStartTimestamp(startDate);

        // wbRequest.setStartTimestamp(getEarliestTimestamp());
      }
      wbRequest.setAnchorTimestamp(getLatestTimestamp());
      if (wbRequest.getEndTimestamp() == null) {
        wbRequest.setEndTimestamp(getLatestTimestamp());
      }
      wbRequest.setCaptureQueryRequest();
      wbRequest.setMementoTimemapFormat(format);
      wbRequest.setRequestUrl(urlStr);
      int pagemax = MementoUtils.getPageMaxRecord(accessPoint);
      if (pagemax == 0) {
        wbRequest.setResultsPerPage(getMaxRecords());
      } else {
        wbRequest.setResultsPerPage(pagemax);
      }
      return wbRequest;
    }
    return null;
  }
View Full Code Here
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
     * <p>%-encoded timestamp.</p>
     * @throws Exception
     */
    public void testPathPrefixDatePrefixEncoded() throws Exception {
    {
      WaybackRequest wbr = parse("/web/2010%2A/http://www.yahoo.com/*");
      assertNotNull(wbr);
      assertTrue(wbr.isUrlQueryRequest());
      assertEquals("20100101000000", wbr.getStartTimestamp());
      assertEquals("20101231235959", wbr.getEndTimestamp());
      assertEquals("http://www.yahoo.com/", wbr.getRequestUrl());
    }
    // negative case - %2A doesn't make it path-prefix.
    {
      WaybackRequest wbr = parse("/web/2010%2A/http://www.yahoo.com/%2A");
      assertNotNull(wbr);
      assertTrue(wbr.isCaptureQueryRequest());
      assertEquals("20100101000000", wbr.getStartTimestamp());
      assertEquals("20101231235959", wbr.getEndTimestamp());
      assertEquals("http://www.yahoo.com/%2A", wbr.getRequestUrl());
    }
    }
View Full Code Here
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
     * start and end timestamp.</p>
     * <p>timerange without "*" is not recognized. it could be.</p>
     */
  public void testPathPrefixDateRange() throws Exception {
    {
      WaybackRequest wbr1 = parse("/web/20100101-20100531*/http://www.yahoo.com/*");
      assertNotNull(wbr1);
      assertTrue(wbr1.isUrlQueryRequest());
      assertEquals("20100101000000", wbr1.getStartTimestamp());
      assertEquals("20100531235959", wbr1.getEndTimestamp());
      assertEquals("http://www.yahoo.com/", wbr1.getRequestUrl());
    }

    // TODO: date range without "*"
  }
 
View Full Code Here
297
298
299
300
301
302
303
304
305
306
307
308
309
   * %-encoded version.
   * @throws Exception
   */
  public void testPathPrefixdateRangeEncoded() throws Exception {
    {
      WaybackRequest wbr = parse("/web/20100101%2D20100531%2A/http://www.yahoo.com/*");
      assertNotNull(wbr);
      assertTrue(wbr.isUrlQueryRequest());
      assertEquals("20100101000000", wbr.getStartTimestamp());
      assertEquals("20100531235959", wbr.getEndTimestamp());
      assertEquals("http://www.yahoo.com/", wbr.getRequestUrl());
    }
  }
View Full Code Here
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
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
412
413
414
415
416
417
418
419
420
421
422
423
   * for non-URL paths.</p>
   * @throws Exception
   */
  public void testPathDateless() throws Exception {
    {
      WaybackRequest wbr = parse("/web/http://www.yahoo.com/");
      checkPathDateless(wbr, "http://www.yahoo.com/");
    }
    {
      WaybackRequest wbr = parse("/web/https://www.yahoo.com/");
      checkPathDateless(wbr, "https://www.yahoo.com/");
    }
    {
      WaybackRequest wbr = parse("/web/http://www.yahoo.com:8080/");
      checkPathDateless(wbr, "http://www.yahoo.com:8080/");
    }
    // some client canonicalizes "//" in path into "/".
    // (why this isn't done for other patterns of requests?)
    {
      WaybackRequest wbr = parse("/web/http:/www.yahoo.com/");
      checkPathDateless(wbr, "http://www.yahoo.com/");
    }
    // doesn't repair "https:/"
    {
      WaybackRequest wbr = parse("/web/https:/www.yahoo.com/");
      //checkPathDateless(wbr, "https://www.yahoo.com/");
      assertNull(wbr);
    }
    // doesn't repair "ftp:/" either.
    {
      WaybackRequest wbr = parse("/web/ftp:/www.yahoo.com/afile");
      assertNull(wbr);
    }
    // scheme-relative - results in NullPointerException FIXME
    {
      try {
        WaybackRequest wbr = parse("/web///www.yahoo.com/");
        assertNull(wbr);
      } catch (NullPointerException ex) {
        // current behavior - FIXME.
      }
    }
    // regular case.
    {
      WaybackRequest wbr = parse("/web/www.yahoo.com/");
      checkPathDateless(wbr, "http://www.yahoo.com/");
    }
    // scheme-less URL with user info is rejected
    // TODO: why is this rejected?
    {
      try {
        @SuppressWarnings("unused")
        WaybackRequest wbr = parse("/web/user@www.yahoo.com/");
        fail("BadQueryException was not thrown");
      } catch (BadQueryException ex) {
        // expected;
      }
    }
    // but it's accepted with scheme.
    // TODO: should this be rejected as well?
    {
      WaybackRequest wbr = parse("/web/http://user@www.yahoo.com/");
      checkPathDateless(wbr, "http://user@www.yahoo.com/");
    }
    // just make sure path and query parts in requestUrl are preserved.
    {
      WaybackRequest wbr = parse("/web/www.yahoo.com/apis?v=2");
      checkPathDateless(wbr, "http://www.yahoo.com/apis?v=2");
    }
    // doesn't look like an URL.
    {
      WaybackRequest wbr = parse("/web/images/foo.gif");
      assertNull(wbr);
    }
    {
      WaybackRequest wbr = parse("/web/handler.php?url=http://www.yahoo.com/");
      assertNull(wbr);
    }
    // TODO: shouldn't this be parsed as dateless URL-Query?
    {
      WaybackRequest wbr = parse("/web/http://www.yahoo.com/*");
      checkPathDateless(wbr, "http://www.yahoo.com/*");
    }
    // ditto
    {
      WaybackRequest wbr = parse("/web/www.yahoo.com/*");
      checkPathDateless(wbr, "http://www.yahoo.com/*");
    }
  }
 
View Full Code Here
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
  public void testPathDatelessWithDateHeader() throws Exception {
    final String dateHeader = "Thu, 24 Apr 2014 21:15:51 UTC+00:00";
    final Date date = (new SimpleDateFormat("EEE, dd MMM yyyy hh:mm:ss z", Locale.ENGLISH)).parse(dateHeader);
    {
      acceptDatetimeHeader = dateHeader;
      WaybackRequest wbr = parse("/web/http://www.yahoo.com/");
      assertNotNull(wbr);
      assertTrue(wbr.isReplayRequest());
      assertFalse(wbr.isBestLatestReplayRequest());
      assertEquals(EXPECTED_START_TIMESTAMP, wbr.getStartTimestamp());
      assertEquals(EXPECTED_END_TIMESTAMP, wbr.getEndTimestamp());
      assertEquals(date, wbr.getReplayDate());
      assertEquals(date, wbr.getAnchorDate());
    }
    // invalid Accept-Datetime header
    {
      acceptDatetimeHeader = "invalid date";
      try {
        @SuppressWarnings("unused")
        WaybackRequest wbr = parse("/web/http://www.yahoo.com/");
        fail("did not throw exception");
      } catch (TimeGateBadQueryException ex) {
        // expected
      }
    }
    // alternate Accept-Timestamp header. as long as it is valid,
    // invalid value in Accept-Datetime header doesn't cause exception.
    {
      acceptDatetimeHeader = "invalid date";
      acceptTimestampHeader = "20140424211551";
      WaybackRequest wbr = parse("/web/http://www.yahoo.com/");
      assertNotNull(wbr);
      assertTrue(wbr.isReplayRequest());
      assertFalse(wbr.isBestLatestReplayRequest());
      assertEquals(EXPECTED_START_TIMESTAMP, wbr.getStartTimestamp());
      assertEquals(EXPECTED_END_TIMESTAMP, wbr.getEndTimestamp());
      assertEquals(date, wbr.getReplayDate());
      assertEquals(date, wbr.getAnchorDate());
    }
    // invalid value in Accept-Timestamp header is silently ignored,
    // unless Accept-Datetime also has an invalid value.
    {
      acceptDatetimeHeader = null;
      acceptTimestampHeader = "*INVALID*";
      WaybackRequest wbr = parse("/web/http://www.yahoo.com/");
      assertNotNull(wbr);
      assertTrue(wbr.isReplayRequest());
      assertTrue(wbr.isBestLatestReplayRequest());
      assertEquals(EXPECTED_START_TIMESTAMP, wbr.getStartTimestamp());
      assertEquals(EXPECTED_END_TIMESTAMP, wbr.getEndTimestamp());
    }
  }
View Full Code Here
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
   * some pathological cases.
   * @throws Exception
   */
    public void testPathological() throws Exception {
    {
      WaybackRequest wbr = parse("/web/20100101*30/http://www.yahoo.com/?p=*");
      assertNull(wbr);
    }
    {
      WaybackRequest wbr = parse("/web/*20100101*/http://www.yahoo.com/");
      assertNull(wbr);
    }
    {
      WaybackRequest wbr = parse("/web/20100101*im_/http://www.yahoo.com/a.png");
      assertNull(wbr);
    }
    // TODO: should we accept this?
    {
      WaybackRequest wbr = parse("/web//20100101*/http://www.yahoo.com/");
      assertNull(wbr);
    }
    }
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
        uriConverter = EasyMock.createMock(ResultURIConverter.class);
       
        response = EasyMock.createMock(HttpServletResponse.class);
        EasyMock.expect(response.getOutputStream()).andReturn(servletOutput);
       
        wbRequest = new WaybackRequest();
        wbRequest.setFrameWrapperContext(false);
       
        result = new CaptureSearchResult();
        result.setOriginalUrl("http://www.example.com/");
        result.setCaptureTimestamp("20100101123456");
View Full Code Here
41
42
43
44
45
46
47
48
49
50
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
   * @throws BetterRequestException
   */
  public void testParseString() throws Exception {
    BaseRequestParser wrapped = new ArchivalUrlRequestParser();
    ReplayRequestParser p = new ReplayRequestParser(wrapped);
    WaybackRequest r;
    AccessPoint ap = null;
    r = p.parse("",ap);
    assertNull("Should not parse empty string", r);
    r = p.parse("20070101000000/foo.com",ap);
    assertNotNull("Should parse legit request sans scheme", r);
    assertEquals("parsed request Url",r.getRequestUrl(),"http://foo.com");
    assertEquals("Parsed timestamp","20070101000000",r.getReplayTimestamp());

    r = p.parse("20070101000000/foo.com/",ap);
    assertEquals("parsed request Url, maintaining trailing slash",
        "http://foo.com/",r.getRequestUrl());

    r = p.parse("200701010000/foo.com",ap);
    assertEquals("parsed partial date",
        "http://foo.com",r.getRequestUrl());
    assertEquals("Parsed partial timestamp to earliest",
        "20070101000000",r.getReplayTimestamp());

    r = p.parse("20070101000000/http://foo.com",ap);
    assertEquals("parsed request Url with scheme",
        "http://foo.com",r.getRequestUrl());

    r = p.parse("20070101000000/http://foo.com/",ap);
    assertEquals("parsed request Url with scheme and trailing slash",
        "http://foo.com/",r.getRequestUrl());

    r = p.parse("20070101000000/ftp://foo.com/",ap);
    assertEquals("parsed request Url with ftp scheme",
        "ftp://foo.com/",r.getRequestUrl());
   
    r = p.parse("20070101000000/https://foo.com/",ap);
    assertEquals("parsed request Url with https scheme",
        "https://foo.com/",r.getRequestUrl());

    r = p.parse("20070101000000js_/http://foo.com/",ap);
    assertEquals("parsed request Url with js_ flag",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed js_ flag",r.isJSContext());
    assertFalse("css not set",r.isCSSContext());

    r = p.parse("20070101000000cs_/http://foo.com/",ap);
    assertEquals("parsed request Url with cs_ flag",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed cs_ flag",r.isCSSContext());
    assertFalse("js not set",r.isJSContext());

    r = p.parse("20070101000000cs_js_/http://foo.com/",ap);
    assertEquals("parsed request Url with cs_ and js_ flags",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed cs_ flag",r.isCSSContext());
    assertTrue("parsed js_ flag",r.isJSContext());

    r = p.parse("20070101000000js_cs_/http://foo.com/",ap);
    assertEquals("parsed request Url with cs_ and js_ flags, backvards",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed cs_ flag",r.isCSSContext());
    assertTrue("parsed js_ flag",r.isJSContext());

    r = p.parse("20070101000000un_/http://foo.com/",ap);
    assertEquals("parsed request Url with unknown flag",
        "http://foo.com/",r.getRequestUrl());
    assertFalse("no cs_ flag",r.isCSSContext());
    assertFalse("no js_ flag",r.isJSContext());

    r = p.parse("20070101000000un_js_cs_/http://foo.com/",ap);
    assertEquals("parsed request Url with falgs and unknown flag",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed cs_ flag",r.isCSSContext());
    assertTrue("parsed js_ flag",r.isJSContext());

    r = p.parse("20070101000000js_cs_un_/http://foo.com/",ap);
    assertEquals("parsed request Url with falgs and unknown flag at end",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed cs_ flag",r.isCSSContext());
    assertTrue("parsed js_ flag",r.isJSContext());

    r = p.parse("20070101000000un_js_cs_un_/http://foo.com/",ap);
    assertEquals("parsed request Url with falgs and unknown flags",
        "http://foo.com/",r.getRequestUrl());
    assertTrue("parsed cs_ flag",r.isCSSContext());
    assertTrue("parsed js_ flag",r.isJSContext());

  }
View Full Code Here
TOP

Related Classes of org.archive.wayback.core.WaybackRequest

  • org.archive.net.UURI
  • org.archive.wayback.archivalurl.ArchivalUrlCSSReplayRendererTest
  • org.archive.wayback.archivalurl.ArchivalURLJSStringTransformerReplayRendererTest
  • org.archive.wayback.archivalurl.ArchivalUrlRequestParserTest
  • org.archive.wayback.archivalurl.ArchivalUrlSAXRewriteReplayRendererTest
  • org.archive.wayback.archivalurl.ArchivalUrlTest
  • org.archive.wayback.archivalurl.FastArchivalUrlReplayParseEventHandler
  • org.archive.wayback.archivalurl.FastArchivalUrlReplayParseEventHandlerTest
  • org.archive.wayback.archivalurl.FastArchivalUrlReplayParseEventHandlerTest$TestJSPExecutor
  • org.archive.wayback.archivalurl.requestparser.ArchivalUrlFormRequestParser
    • 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.