Package com.zhangwoo.spider.po

Examples of com.zhangwoo.spider.po.Conversation

34
35
36
37
38
39
40
41
42
43
44
45
    try {
      NodeList cons = (NodeList) xpath.evaluate(
          "//TABLE[@width='349' and @height='210']", docHtml,
          XPathConstants.NODESET);
      for (int i = 0; i < cons.getLength(); i++) {
        Conversation result = new Conversation();
        result.setContent(xpath.evaluate(
            "//TABLE[@width='349' and @height='210']",
                cons.item(i)));
        System.out.println(xpath.evaluate(
            "//TABLE[@width='349' and @height='210']",
                cons.item(i)));
View Full Code Here
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 (analyserNode != null) {
        String anSave = xpath.evaluate("@save", analyserNode);
        String anXpath = xpath.evaluate("@xpath", analyserNode);
        String anRegexp = xpath.evaluate("@regexp", analyserNode);

        Conversation convsXpath = getXpaths(analyserNode);
        Conversation convsRegExp = getRegExps(analyserNode);
        List<Conversation> convsResults = new ArrayList<Conversation>();

        // reg(可能json) 与 xpath(html标签) 走完全不同的路线
        if (!StringUtil.isEmpty(anXpath)
            && StringUtil.isEmpty(anRegexp)) { // 纯粹XPATH，给出结果一定是NodeList
          NodeList cons = (NodeList) xpath.evaluate(anXpath, docHtml,
              XPathConstants.NODESET);
          Conversation conTemp = new Conversation();
          for (int consi = 0; consi < cons.getLength(); consi++) {
            conTemp = analyserConversation(convsXpath, convsRegExp,
                cons.item(consi), null);
            conTemp.setSaveable(anSave);
            if(StringUtil.isEmpty(conTemp.getSelfLink())){
              conTemp.setSelfLink(urlReq.getUrl());
            }
            conTemp.setTid(urlReq.getTask().getTid());
            convsResults.add(conTemp);
          }
        } else if (!StringUtil.isEmpty(anRegexp)) {
          // List<String[]> cons=StringUtil.matchAll(anRegexp, html);
        }
View Full Code Here
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
  }
 

  private Conversation getRegExps(Node analyserNode)
      throws XPathExpressionException {
    Conversation convsRegExp = new Conversation();
    convsRegExp.setAuthor(xpath.evaluate("AUTHOR/@regexp", analyserNode));
    convsRegExp.setContent(xpath.evaluate("CONTENT/@regexp", analyserNode));
    convsRegExp.setMainLink(xpath
        .evaluate("MAINLINK/@regexp", analyserNode));
    convsRegExp.setPublishTime(xpath.evaluate("PUBLISHTIME/@regexp",
        analyserNode));
    convsRegExp.setSelfLink(xpath
        .evaluate("SELFLINK/@regexp", analyserNode));
    convsRegExp.setTitle(xpath.evaluate("TITLE/@regexp", analyserNode));
    convsRegExp.setUpdateTime(xpath.evaluate("UPDATETIME/@regexp",
        analyserNode));
    convsRegExp.setIsTopic(xpath.evaluate("ISTOPIC/@regexp", analyserNode));
    convsRegExp.setStopByExp(xpath.evaluate("SELFLINK/@stopByExp",
        analyserNode));
    return convsRegExp;
  }
View Full Code Here
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
    return convsRegExp;
  }

  private Conversation getXpaths(Node analyserNode)
      throws XPathExpressionException {
    Conversation convsXpath = new Conversation();
    convsXpath.setAuthor(xpath.evaluate("AUTHOR/@xpath", analyserNode).trim());
    convsXpath.setContent(xpath.evaluate("CONTENT/@xpath", analyserNode).trim());
    convsXpath.setMainLink(xpath.evaluate("MAINLINK/@xpath", analyserNode).trim());
    convsXpath.setPublishTime(xpath.evaluate("PUBLISHTIME/@xpath",
        analyserNode).trim());
    convsXpath.setSelfLink(xpath.evaluate("SELFLINK/@xpath", analyserNode).trim());
    convsXpath.setTitle(xpath.evaluate("TITLE/@xpath", analyserNode).trim());
    convsXpath.setUpdateTime(xpath.evaluate("UPDATETIME/@xpath",
        analyserNode).trim());
    convsXpath.setIsTopic(xpath.evaluate("ISTOPIC/@xpath", analyserNode).trim());
    convsXpath.setStopByXpath(xpath.evaluate("SELFLINK/@stopByXpath",
        analyserNode).trim());
    convsXpath
        .setRunable(xpath.evaluate("SELFLINK/@runable", analyserNode).trim());
    return convsXpath;
  }
View Full Code Here
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
   * @throws XPathExpressionException
   */
  private Conversation analyserConversation(Conversation convsXpath,
      Conversation convsRegExp, Node node, String str)
      throws XPathExpressionException {
    Conversation resutl = new Conversation();

    if (node == null) {

    } else if (StringUtil.isEmpty(str)) {
      if (!StringUtil.isEmpty(convsXpath.getAuthor()))
        resutl.setAuthor(xpath.evaluate(convsXpath.getAuthor(), node).trim());
      if (!StringUtil.isEmpty(convsXpath.getContent()))
        resutl.setContent(xpath.evaluate(convsXpath.getContent(), node).trim());
      if (!StringUtil.isEmpty(convsXpath.getIsTopic()))
        resutl.setIsTopic(xpath.evaluate(convsXpath.getIsTopic(), node).trim());
      if (!StringUtil.isEmpty(convsXpath.getMainLink()))
        resutl.setMainLink(xpath.evaluate(convsXpath.getMainLink(),
            node).trim());
      if (!StringUtil.isEmpty(convsXpath.getPublishTime()))
        resutl.setPublishTime(xpath.evaluate(
            convsXpath.getPublishTime(), node).trim());
      if (!StringUtil.isEmpty(convsXpath.getRunable()))
        resutl.setRunable(xpath.evaluate(convsXpath.getRunable(), node).trim());
      if (!StringUtil.isEmpty(convsXpath.getSelfLink()))
        resutl.setSelfLink(SpiderThread.formatUrl(xpath.evaluate(convsXpath.getSelfLink(),
            node).trim(),this.urlReq));
      if (!StringUtil.isEmpty(convsXpath.getTitle()))
        resutl.setTitle(xpath.evaluate(convsXpath.getTitle(), node).trim());
      if (!StringUtil.isEmpty(convsXpath.getUpdateTime()))
        resutl.setUpdateTime(xpath.evaluate(convsXpath.getUpdateTime(),
            node).trim());
      if (!StringUtil.isEmpty(convsXpath.getRunable()))
        resutl.setRunable(xpath.evaluate(convsXpath.getRunable(), node).trim());
    }

    return resutl;
  }
View Full Code Here
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
      Document docHtml) {
    List<Conversation> convsResults = new ArrayList<Conversation>();
   
    String[] tmps=StringUtil.match(html, "<a class=\"downlink\" href=\"([^\"]+)\"></a>");
    if(tmps!=null){
      Conversation c=new Conversation();
      c.setSelfLink(StringUtil.match(html, "<a class=\"downlink\" href=\"([^\"]+)\"></a>")[1]);
      tmps=StringUtil.match(html, "<a href=\"/all/\\d+.htm\">(.*?\\-.*?)</a>");
      if(tmps!=null){
        c.setAuthor(tmps[1].split("-")[0]);
        c.setTitle(tmps[1].split("-")[1]);
        convsResults.add(c);
      }
    }
   
   
 
View Full Code Here
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
      Document docHtml) {
    List<Conversation> convsResults = new ArrayList<Conversation>();
    try {
      NodeList cons = (NodeList) xpath.evaluate("//DIV[@class='text clearfix']", docHtml,XPathConstants.NODESET);
      for (int i = 0; i < cons.getLength(); i++) {
        Conversation result = new Conversation();
        result.setTitle(urlReq.getTitle());
        result.setContent(xpath.evaluate("DIV[@class='title_top']/text()", cons.item(i))+xpath.evaluate("DIV[@class='s_cont']/text()", cons.item(i)));
        result.setSelfLink(xpath.evaluate("DIV[@class='title_top']/A/@href", cons.item(i)));
        result.setPublishTime(StringUtil.match(xpath.evaluate("DIV[@class='title clearfix']/SPAN[@class='level']/text()", cons.item(i)),"\\d{4}-\\d{2}-\\d{2} \\d{2}:\\d{2}:\\d{2}")[0]);
        convsResults.add(result);
      }
    } catch (XPathExpressionException e) {
      logger.error("findConversations error",e);
    }
View Full Code Here
TOP

Related Classes of com.zhangwoo.spider.po.Conversation

  • com.zhangwoo.spider.client.analyse.XmlAnalyser
  • com.zhangwoo.spider.client.process.XmlAnalyserTemplate
  • plugins.dmall.dangdang.DangdangComment
  • plugins.music.fivend.FivendMusicDownload
    • 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.