Package org.apache.nutch.net.protocols

Examples of org.apache.nutch.net.protocols.Response

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
    String urlString = url.toString();
    try {
      URL u = new URL(urlString);
     
      long startTime = System.currentTimeMillis();
      Response response = getResponse(u, datum, false); // make a request
     
      if(this.responseTime) {
        int elapsedTime = (int) (System.currentTimeMillis() - startTime);
        datum.getMetaData().put(RESPONSE_TIME, new IntWritable(elapsedTime));
      }
     
      int code = response.getCode();
      byte[] content = response.getContent();
      Content c = new Content(u.toString(), u.toString(),
                              (content == null ? EMPTY_CONTENT : content),
                              response.getHeader("Content-Type"),
                              response.getHeaders(), this.conf);
     
      if (code == 200) { // got a good response
        return new ProtocolOutput(c); // return it
       
      } else if (code == 410) { // page is gone
        return new ProtocolOutput(c, new ProtocolStatus(ProtocolStatus.GONE, "Http: " + code + " url=" + url));
       
      } else if (code >= 300 && code < 400) { // handle redirect
        String location = response.getHeader("Location");
        // some broken servers, such as MS IIS, use lowercase header name...
        if (location == null) location = response.getHeader("location");
        if (location == null) location = "";
        u = new URL(u, location);
        int protocolStatusCode;
        switch (code) {
          case 300:   // multiple choices, preferred value in Location
View Full Code Here
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
   
    String urlString = url.toString();
    try {
      URL u = new URL(urlString);
      String host = null;
      Response response = getResponse(u, datum, false); // make a request
     
      int code = response.getCode();
      byte[] content = response.getContent();
      Content c = new Content(u.toString(), u.toString(),
                              (content == null ? EMPTY_CONTENT : content),
                              response.getHeader("Content-Type"),
                              response.getHeaders(), this.conf);
     
      if (code == 200) { // got a good response
        return new ProtocolOutput(c); // return it
       
      } else if (code == 410) { // page is gone
        return new ProtocolOutput(c, new ProtocolStatus(ProtocolStatus.GONE, "Http: " + code + " url=" + url));
       
      } else if (code >= 300 && code < 400) { // handle redirect
        String location = response.getHeader("Location");
        // some broken servers, such as MS IIS, use lowercase header name...
        if (location == null) location = response.getHeader("location");
        if (location == null) location = "";
        u = new URL(u, location);
        int protocolStatusCode;
        switch (code) {
          case 300:   // multiple choices, preferred value in Location
View Full Code Here
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
  public ProtocolOutput getProtocolOutput(String url, WebPage page) {

    try {
      URL u = new URL(url);
      String host = null;
      Response response = getResponse(u, page, false); // make a request
      int code = response.getCode();
      byte[] content = response.getContent();
      Content c = new Content(u.toString(), u.toString(),
          (content == null ? EMPTY_CONTENT : content),
          response.getHeader("Content-Type"),
          response.getHeaders(), mimeTypes);

      if (code == 200) { // got a good response
        return new ProtocolOutput(c); // return it

      } else if (code == 410) { // page is gone
        return new ProtocolOutput(c,
            ProtocolStatusUtils.makeStatus(ProtocolStatusCodes.GONE, "Http: " + code + " url=" + url));
      } else if (code >= 300 && code < 400) { // handle redirect
        String location = response.getHeader("Location");
        // some broken servers, such as MS IIS, use lowercase header name...
        if (location == null) location = response.getHeader("location");
        if (location == null) location = "";
        u = new URL(u, location);
        int protocolStatusCode;
        switch (code) {
        case 300:   // multiple choices, preferred value in Location
View Full Code Here
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
    if (robotRules == null) {                     // cache miss
      URL redir = null;
      if (LOG.isTraceEnabled()) { LOG.trace("cache miss " + url); }
      try {
        Response response = http.getResponse(new URL(url, "/robots.txt"),
                                             new WebPage(), true);
        // try one level of redirection ?
        if (response.getCode() == 301 || response.getCode() == 302) {
          String redirection = response.getHeader("Location");
          if (redirection == null) {
            // some versions of MS IIS are known to mangle this header
            redirection = response.getHeader("location");
          }
          if (redirection != null) {
            if (!redirection.startsWith("http")) {
              // RFC says it should be absolute, but apparently it isn't
              redir = new URL(url, redirection);
            } else {
              redir = new URL(redirection);
            }

            response = http.getResponse(redir, new WebPage(), true);
          }
        }

        if (response.getCode() == 200)               // found rules: parse them
          robotRules = parseRules(response.getContent());
        else if ( (response.getCode() == 403) && (!allowForbidden) )
          robotRules = FORBID_ALL_RULES;            // use forbid all
        else if (response.getCode() >= 500) {
          cacheRule = false;
          robotRules = EMPTY_RULES;
        }else
          robotRules = EMPTY_RULES;                 // use default rules
      } catch (Throwable t) {
View Full Code Here
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
  public ProtocolOutput getProtocolOutput(Text url, CrawlDatum datum) {
   
    String urlString = url.toString();
    try {
      URL u = new URL(urlString);
      Response response = getResponse(u, datum, false); // make a request
     
      int code = response.getCode();
      byte[] content = response.getContent();
      Content c = new Content(u.toString(), u.toString(),
                              (content == null ? EMPTY_CONTENT : content),
                              response.getHeader("Content-Type"),
                              response.getHeaders(), this.conf);
     
      if (code == 200) { // got a good response
        return new ProtocolOutput(c); // return it
       
      } else if (code == 410) { // page is gone
        return new ProtocolOutput(c, new ProtocolStatus(ProtocolStatus.GONE, "Http: " + code + " url=" + url));
       
      } else if (code >= 300 && code < 400) { // handle redirect
        String location = response.getHeader("Location");
        // some broken servers, such as MS IIS, use lowercase header name...
        if (location == null) location = response.getHeader("location");
        if (location == null) location = "";
        u = new URL(u, location);
        int protocolStatusCode;
        switch (code) {
          case 300:   // multiple choices, preferred value in Location
View Full Code Here
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
   
    if (robotRules == null) {                     // cache miss
      URL redir = null;
      if (LOG.isTraceEnabled()) { LOG.trace("cache miss " + url); }
      try {
        Response response = ((HttpBase)http).getResponse(new URL(url, "/robots.txt"),
                                             new CrawlDatum(), true);
        // try one level of redirection ?
        if (response.getCode() == 301 || response.getCode() == 302) {
          String redirection = response.getHeader("Location");
          if (redirection == null) {
            // some versions of MS IIS are known to mangle this header
            redirection = response.getHeader("location");
          }
          if (redirection != null) {
            if (!redirection.startsWith("http")) {
              // RFC says it should be absolute, but apparently it isn't
              redir = new URL(url, redirection);
            } else {
              redir = new URL(redirection);
            }
           
            response = ((HttpBase)http).getResponse(redir, new CrawlDatum(), true);
          }
        }

        if (response.getCode() == 200)               // found rules: parse them
          robotRules =  parseRules(url.toString(), response.getContent(),
                                   response.getHeader("Content-Type"),
                                   agentNames);

        else if ( (response.getCode() == 403) && (!allowForbidden) )
          robotRules = FORBID_ALL_RULES;            // use forbid all
        else if (response.getCode() >= 500) {
          cacheRule = false;
          robotRules = EMPTY_RULES;
        }else                                       
          robotRules = EMPTY_RULES;                 // use default rules
      } catch (Throwable t) {
View Full Code Here
170
171
172
173
174
175
176
177
178
179
180
   * @throws Exception    When an error occurs or test case fails.
   */
  private void fetchPage(String page, int expectedCode)
      throws Exception {
    URL url = new URL("http", "127.0.0.1", port, page);
    Response response = null;
    response = http.getResponse(url, new CrawlDatum(), true);

    int code = response.getCode();
    assertEquals("HTTP Status Code for " + url, expectedCode, code);
  }
View Full Code Here
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
      }
     
      long crawlDelay = robots.getCrawlDelay(this, u);
      long delay = crawlDelay > 0 ? crawlDelay : serverDelay;
      String host = blockAddr(u, delay);
      Response response;
      try {
        response = getResponse(u, datum, false); // make a request
      } finally {
        unblockAddr(host, delay);
      }
     
      int code = response.getCode();
      byte[] content = response.getContent();
      Content c = new Content(u.toString(), u.toString(),
                              (content == null ? EMPTY_CONTENT : content),
                              response.getHeader("Content-Type"),
                              response.getHeaders(), this.conf);
     
      if (code == 200) { // got a good response
        return new ProtocolOutput(c); // return it
       
      } else if (code == 410) { // page is gone
        return new ProtocolOutput(c, new ProtocolStatus(ProtocolStatus.GONE, "Http: " + code + " url=" + url));
       
      } else if (code >= 300 && code < 400) { // handle redirect
        String location = response.getHeader("Location");
        // some broken servers, such as MS IIS, use lowercase header name...
        if (location == null) location = response.getHeader("location");
        if (location == null) location = "";
        u = new URL(u, location);
        int protocolStatusCode;
        switch (code) {
          case 300:   // multiple choices, preferred value in Location
View Full Code Here
173
174
175
176
177
178
179
180
181
182
183
   * @throws Exception    When an error occurs or test case fails.
   */
  private void fetchPage(String page, int expectedCode)
      throws Exception {
    URL url = new URL("http", "127.0.0.1", port, page);
    Response response = null;
    response = http.getResponse(url, new CrawlDatum(), true);

    int code = response.getCode();
    assertEquals("HTTP Status Code for " + url, expectedCode, code);
  }
View Full Code Here
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
   
    String urlString = url.toString();
    try {
      URL u = new URL(urlString);
      String host = null;
      Response response = getResponse(u, datum, false); // make a request
     
      int code = response.getCode();
      byte[] content = response.getContent();
      Content c = new Content(u.toString(), u.toString(),
                              (content == null ? EMPTY_CONTENT : content),
                              response.getHeader("Content-Type"),
                              response.getHeaders(), this.conf);
     
      if (code == 200) { // got a good response
        return new ProtocolOutput(c); // return it
       
      } else if (code == 410) { // page is gone
        return new ProtocolOutput(c, new ProtocolStatus(ProtocolStatus.GONE, "Http: " + code + " url=" + url));
       
      } else if (code >= 300 && code < 400) { // handle redirect
        String location = response.getHeader("Location");
        // some broken servers, such as MS IIS, use lowercase header name...
        if (location == null) location = response.getHeader("location");
        if (location == null) location = "";
        u = new URL(u, location);
        int protocolStatusCode;
        switch (code) {
          case 300:   // multiple choices, preferred value in Location
View Full Code Here
TOP

Related Classes of org.apache.nutch.net.protocols.Response

  • org.apache.nutch.protocol.http.api.HttpBase
  • org.apache.nutch.protocol.http.api.HttpRobotRulesParser
  • org.apache.nutch.protocol.http.api.RobotRulesParser
  • org.apache.nutch.protocol.http.TestProtocolHttp
  • org.apache.nutch.protocol.httpclient.TestProtocolHttpClient
    • 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.