Package org.archive.wayback.core

Examples of org.archive.wayback.core.FastCaptureSearchResult

42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
    if (exclusionFactory == null) {
      return false;
    }
   
    ExclusionFilter filter = exclusionFactory.get();
    FastCaptureSearchResult result = new FastCaptureSearchResult();
    result.setOriginalUrl(url);
   
    int status = ExclusionFilter.FILTER_EXCLUDE;
   
    String urlKey = url;
   
    try {
      // Must set canonicalizer
      urlKey = canonicalizer.urlStringToKey(url);     
      result.setUrlKey(urlKey);
     
      status = filter.filterObject(result);
    } catch (URIException e) {
      //Exclude invalid
    }
View Full Code Here
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
     
      Thread.currentThread().setName("Thread " +
          Thread.currentThread().getId() + " " + getBeanName() +
          " handling: " + urlString);
     
      CaptureSearchResult result = new FastCaptureSearchResult();
     
      r = this.getLiveWebResource(result, urlString);
     
      if (r != null) {       
        CaptureSearchResults results = new CaptureSearchResults();
        results.addSearchResult(result);
     
        wbRequest.setReplayTimestamp(result.getCaptureTimestamp());
         
        inner.getReplay().getRenderer(wbRequest, result, r).renderResource(httpRequest, httpResponse, wbRequest, result, r,
            inner.getUriConverter(), results); 
      } else {
        throw new LiveDocumentNotAvailableException(urlString);
View Full Code Here
298
299
300
301
302
303
304
305
306
307
308
    ArcResource r = null;
   
    String urlString = wbRequest.getRequestUrl();
   
    try {
      r = getLiveWebResource(new FastCaptureSearchResult(), urlString);
      return (r != null);
     
    } catch (Exception e) {
      return false;
    } finally {
View Full Code Here
41
42
43
44
45
46
47
48
    return parseCDXLineFlex(line, result);
  }
 
  // Use FastCaptureSearchResult to
  public static CaptureSearchResult parseCDXLineFlexFast(String line) { 
    CaptureSearchResult result = new FastCaptureSearchResult();
    return parseCDXLineFlex(line, result);
  }
View Full Code Here
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
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
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
    }
    }

  @Override
    public int writeLine(CDXLine line) {
    FastCaptureSearchResult result = new FastCaptureSearchResult();
   
    String timestamp = line.getTimestamp();
    String originalUrl = line.getOriginalUrl();
   
    if ((prevResult != null) && (preferContains != null) &&
       prevResult.getCaptureTimestamp().equals(timestamp) &&
       prevResult.getOriginalUrl().equals(originalUrl) &&
       prevLine.getLength().equals(line.getLength()) &&
       prevLine.getOffset().equals(line.getOffset())) {
     
      String currFile = line.getFilename();
      String prevFile = prevLine.getFilename();
     
      if (currFile.contains(preferContains) && !prevFile.contains(preferContains)) {
        prevResult.setFile(currFile);
      }
     
      return 0;
    }
       
    result.setUrlKey(line.getUrlKey());
    result.setCaptureTimestamp(timestamp);
    result.setOriginalUrl(originalUrl);
   
    // Special case: filter out captures that have userinfo
    boolean hasUserInfo = (UrlOperations.urlToUserInfo(result.getOriginalUrl()) != null);
   
    if (hasUserInfo) {
      return 0;
    }
   
    result.setRedirectUrl(line.getRedirect());
    result.setHttpCode(line.getStatusCode());
   
    if (selfRedirFilter != null && !result.getRedirectUrl().equals(CDXLine.EMPTY_VALUE)) {
      if (selfRedirFilter.filterObject(result) != ObjectFilter.FILTER_INCLUDE) {
        return 0;
      }
    }
   
    if (exclusionFilter != null) {
      if (exclusionFilter.filterObject(result) != ObjectFilter.FILTER_INCLUDE) {
        return 0;
      }
    }
   
    result.setMimeType(line.getMimeType());
    result.setDigest(line.getDigest());
    result.setOffset(NumberUtils.toLong(line.getOffset(), -1));
    result.setCompressedLength(NumberUtils.toLong(line.getLength(), -1));
    result.setFile(line.getFilename());
    result.setRobotFlags(line.getRobotFlags());
   
    boolean isRevisit = false;
   
    if (resolveRevisits) {
      isRevisit = result.getFile().equals(CDXLine.EMPTY_VALUE) ||
            result.getMimeType().equals(REVISIT_VALUE);
     
      String digest = result.getDigest();
     
      if (isRevisit) {
        if (!isReverse) {
          CaptureSearchResult payload = digestToOriginal.get(digest);
          if (payload != null) {
            result.flagDuplicateDigest(payload);
          } else {
            result.flagDuplicateDigest();
          }
        } else {
          LinkedList<CaptureSearchResult> revisits = digestToRevisits.get(digest);
          if (revisits == null) {
            revisits = new LinkedList<CaptureSearchResult>();
            digestToRevisits.put(digest, revisits);
          }
          revisits.add(result);
        }
      } else {
        if (!isReverse) {
          digestToOriginal.put(digest, result);
        } else {
          LinkedList<CaptureSearchResult> revisits = digestToRevisits.remove(digest);
          if (revisits != null) {
            for (CaptureSearchResult revisit : revisits) {
              revisit.flagDuplicateDigest(result);
            }
          }
        }
      }
    }
   
//    String payloadFile = line.getField(RevisitResolver.origfilename);
//   
//    if (!payloadFile.equals(CDXLine.EMPTY_VALUE)) {
//      FastCaptureSearchResult payload = new FastCaptureSearchResult();
//      payload.setFile(payloadFile);
//      payload.setOffset(NumberUtils.toLong(line.getField(RevisitResolver.origoffset), -1));
//      payload.setCompressedLength(NumberUtils.toLong(line.getField(RevisitResolver.origlength), -1));
//      result.flagDuplicateDigest(payload);
//    }
   
    if ((targetTimestamp != null) && (closest == null)) {
      closest = determineClosest(result);
    }
   
    results.addSearchResult(result, !isReverse);
    prevResult = result;
    prevLine = line;
   
    // Short circuit the load if seeking single capture
    if (seekSingleCapture && resolveRevisits) {
      if (closest != null) {
        // If not a revisit, we're done
        if (!isRevisit) {
          done = true;
        // Else make sure the revisit is resolved
        } else if (result.getDuplicatePayload() != null) {
          done = true;
        }
      }
    }
   
View Full Code Here
39
40
41
42
43
44
45
      this.robotsFilter = robotsFilter;
     
      this.prefixFilter1 = prefixFilter1;
      this.prefixFilter2 = prefixFilter2;
     
      this.resultTester = new FastCaptureSearchResult();
    }
View Full Code Here
TOP

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

  • org.archive.wayback.accesscontrol.robotstxt.AccessControlChecker
  • org.archive.wayback.resourceindex.cdx.format.CDXFlexFormat
  • org.archive.wayback.resourceindex.cdxserver.AccessCheckFilter
  • org.archive.wayback.resourceindex.cdxserver.CDXToCaptureSearchResultsWriter
  • org.archive.wayback.webapp.LiveWebAccessPoint
    • 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.