Package org.vietspider.html.path2

Examples of org.vietspider.html.path2.NodePathParser

115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
        tree.removeAll();
        if(document != null) handler.createTreeItem(tree, document);
        new AutoSelectDataNode2(document, url, handler, tree);
       
        if(paths.length < 1) {
          HTMLExtractor extractor  = new HTMLExtractor();
          NodePathParser pathParser = new NodePathParser();
          if(hyperlinkUtil == null) hyperlinkUtil = new HyperLinkUtil();
          HTMLNode header = null;
          HTMLNode body = null;
          try {
            NodePath nodePath  = pathParser.toPath("HEAD");
            header = extractor.lookNode(document.getRoot(), nodePath);
            nodePath  = pathParser.toPath("BODY");
            body = extractor.lookNode(document.getRoot(), nodePath);
          } catch (Exception e) {
            ClientLog.getInstance().setException(getShell(), e);
          }
         
          if(header == null || body == null) return;
View Full Code Here
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
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
    String txt = handler.getConfig(item);
    txtPath.setText(txt);
   
    treeAddButton.computeShowArea(item);
   
    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();
   
    try {
      NodePath nodePath  = pathParser.toPath(txt);
      HTMLNode node = extractor.lookNode(document.getRoot(), nodePath);
      if(node == null) return;
      if(node.isNode(Name.CONTENT)
          || node.isNode(Name.COMMENT)
          || node.isNode(Name.UNKNOWN)) {
        browser.setText(node.getTextValue());
        return;
      }
     
      NodePath headerPath  = pathParser.toPath("HEAD");
      HTMLNode header = extractor.lookNode(document.getRoot(), headerPath);
   
      if(toolbar.isShowAll()) {
        if(hyperlinkUtil == null) hyperlinkUtil = new HyperLinkUtil();
        try {
          String address = toolbar.getText();
          URL home = new URL(address);
         
          hyperlinkUtil.createFullNormalLink(node, home);
          hyperlinkUtil.createFullImageLink(node, home);
          HashMap<String, String> map = new HashMap<String,String>();
          map.put("link","href");
          map.put("script","src");
          hyperlinkUtil.createFullLink(header, map, home, null);
        } catch(MalformedURLException me) {
        } catch (Exception e) {
          ClientLog.getInstance().setException(getShell(), e);
        }
      }
     
      StringBuilder builder = new StringBuilder();
      builder.append("<html>");
      builder.append("<head>");
      if(toolbar.isShowAll() && header != null && header.getChildren() != null) {
        for(HTMLNode ele : header.getChildren()){
          builder.append(ele.getTextValue()).append('\n');
        }
      }
//      String baseHref = HTMLExplorer.class.getResource("").toString();
//      builder.append("<base href=\""+baseHref+"\">");
      if(node.isNode(Name.BODY)) {
        HTMLNode body = null;
        try {
          nodePath  = pathParser.toPath("BODY");
          body = extractor.lookNode(document.getRoot(), nodePath);
        } catch (Exception e) {
          ClientLog.getInstance().setException(getShell(), e);
        }
       
        if(header == null || body == null) return;
View Full Code Here
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
  void viewItem(){
    TreeItem[] items = tree.getSelection();
    if( items == null || items.length  < 1) return;
    int x = getShell().getLocation().x+120, y = getShell().getLocation().y +130;
    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();
    for(TreeItem item : items) {
      String pathIndex = handler.getConfig(item);
      try {
        NodeInfoViewer viewer = new NodeInfoViewer(getShell(), x, y);
        NodePath nodePath = pathParser.toPath(pathIndex);
        HTMLNode node = extractor.lookNode(document.getRoot(), nodePath);
        viewer.setNode(node);  
        x += 10;
        y += 10;
        nodeViewers.add(viewer);
      } catch(Exception exp) {
View Full Code Here
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
    }
    return upParent(node.getParent(), names);
  }
 
  private HTMLNode searchTitleNode(HTMLNode root, HTMLNode node) {
    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();
    String title = "title";
    try {
      NodePath nodePath  = pathParser.toPath("HEAD.TITLE");
      HTMLNode titleNode = extractor.lookNode(root, nodePath);
      if(titleNode.hasChildren()) {
        title  = titleNode.getChild(0).getTextValue();
      }
    } catch (Exception e) {
      return null;
View Full Code Here
36
37
38
39
40
41
42
43
44
45
46
47
48
49
  }
 
  private String [] build(List<NodeImpl> tokens) throws Exception {
    HTMLDocument doc = new HTMLParser2().createDocument(tokens);
    NodePathParser pathParser = new NodePathParser();
    HTMLExtractor extractor  = new HTMLExtractor();
   
    NodePath nodePath  = pathParser.toPath("BODY");
    HTMLNode body = extractor.lookNode(doc.getRoot(), nodePath);
   
    TextRenderer renderer = new TextRenderer(body, TextRenderer.HANDLER);
    String value = renderer.getTextValue().toString();
    return value.trim().split("\n");
  }
View Full Code Here
232
233
234
235
236
237
238
239
240
241
242
      chars = CharsUtil.cutAndTrim(chars, 0, chars.length);
      chars =  java.text.Normalizer.normalize(new String(chars), Normalizer.Form.NFC).toCharArray();
      node.setValue(chars);             
    } 

    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();

    NodePath nodePath  = pathParser.toPath("BODY");
    return extractor.lookNode(document.getRoot(), nodePath);
  }
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
      chars = CharsUtil.cutAndTrim(chars, 0, chars.length);
      chars =  java.text.Normalizer.normalize(new String(chars), Normalizer.Form.NFC).toCharArray();
      node.setValue(chars);             
    } 

    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();

    NodePath nodePath  = pathParser.toPath("BODY");
    return extractor.lookNode(document.getRoot(), nodePath);
  }
View Full Code Here
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
    if(popup != null) popup.dispose();
  }*/
 
  private HTMLNode createNode(String path) {
    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();
    if(path.endsWith("[.")) {
      path  = path.substring(0, path.length() - 2);
    } else  if(path.endsWith("[")) {
      path  = path.substring(0, path.length() - 1);
    }
    try {
      NodePath nodePath = pathParser.toPath(path);
      return extractor.lookNode(explorer.getDocument().getRoot(), nodePath);
    } catch (Exception e) {
    }
    return null;
  }
View Full Code Here
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
      }
    }
  }
 
  private HTMLNode createNode(String path) {
    HTMLExtractor extractor  = new HTMLExtractor();
    NodePathParser pathParser = new NodePathParser();
    if(path.endsWith("[.")) {
      path  = path.substring(0, path.length() - 2);
    } else  if(path.endsWith("[")) {
      path  = path.substring(0, path.length() - 1);
    }
    try {
      NodePath nodePath = pathParser.toPath(path);
      return extractor.lookNode(explorer.getDocument().getRoot(), nodePath);
    } catch (Exception e) {
    }
    return null;
  }
View Full Code Here
52
53
54
55
56
57
58
59
60
61
62
    }catch (Exception e) {
      ClientLog.getInstance().setException(tree.getShell(), e);
    }
   
    if(bodyPath == null || document == null) return;
    HTMLNode body = new HTMLExtractor().lookNode(document.getRoot(), bodyPath);
    List<HTMLNode> list = new ArrayList<HTMLNode>();
    List<HTMLNode> commons = new ArrayList<HTMLNode>();
    if(HTMLExplorer.SECTION == type) {
      searchSectionCSS(commons, list, body);
View Full Code Here
TOP

Related Classes of org.vietspider.html.path2.NodePathParser

  • ExtractForum
  • HASTCIndexEx
  • HTMLNodeExplorer
  • NewsComExtractor
  • org.vietspider.browser.LoginUtils
  • org.vietspider.html.HTMLDocument
  • org.vietspider.html.HTMLNode
  • org.vietspider.html.Name
  • org.vietspider.html.renderer.ContentRegionSearcher
  • org.vietspider.html.renderer.ContentRendererFactory
    • 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.