Package org.vietspider.html

Examples of org.vietspider.html.HTMLNode

34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
      }
    }
  }

  public HTMLDocument searchDocument(HTMLDocument document, String name) {
    HTMLNode root = document.getRoot();
    HTMLNode newRoot = searchDocument(root, name);
    if (root == newRoot) return document;
    NodePathParser pathParser = new NodePathParser();
    HTMLExtractor extractor = new HTMLExtractor();
    try {
      HTMLNode body = extractor.lookNode(root, pathParser.toPath("BODY"));
      body.clearChildren();
      body.addChild(newRoot);
    } catch (Exception e) {
      LogService.getInstance().setThrowable(e);
    }
    return document;
  }
View Full Code Here
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
    }
    return root;
  }

  private HTMLNode processNextNode(HTMLNode root, List<HTMLNode> anchors, int i) {
    HTMLNode node = anchors.get(i);
    HTMLNode nextNode = anchors.get(i + 1);
    String path = getIndexPath(node);
    String nextPath = getIndexPath(nextNode);
    String commonPath = getCommonIndexPath(path, nextPath);
    try {
      HTMLNode commonNode = getNodeByIndex(root, commonPath);
      path = path.substring(commonPath.length());
      path = path.substring(1, path.indexOf('.', 1));
      int idx = Integer.parseInt(path);

      nextPath = nextPath.substring(commonPath.length());
      nextPath = nextPath.substring(1, nextPath.indexOf('.', 1));
      int nextIdx = Integer.parseInt(nextPath);

      HTMLNode newCommonNode =
        new NodeImpl(commonNode.getValue(), commonNode.getName(), TypeToken.TAG);
      List<HTMLNode> children = commonNode.getChildren();
      for (i = idx; i < nextIdx; i++) {
        newCommonNode.addChild(children.get(i));
      }
      return newCommonNode;
    } catch (Exception e) {
      // LogService.getInstance().setMessage(e.toString());
      return root;
View Full Code Here
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
  }

  private HTMLNode processPrevNode(HTMLNode root, List<HTMLNode> anchors, int i) {
    if (i < 0 || i >= anchors.size())
      return root;
    HTMLNode node = anchors.get(i);
    HTMLNode prevNode = anchors.get(i - 1);
    String path = getIndexPath(node);
    String prevPath = getIndexPath(prevNode);
    String commonPath = getCommonIndexPath(path, prevPath);
    try {
      HTMLNode commonNode = getNodeByIndex(root, commonPath);
      path = path.substring(commonPath.length());
      int idx = path.indexOf('.', 1);
      // path = path.substring(0, idx);
      if (idx > 0) {
        path = path.substring(1, idx);
      } else {
        if (path.charAt(0) == '.')
          path = path.substring(1);
      }

      // if(idx > 1) path = path.substring(2, idx);
      // if(path.charAt(0) == '.') path = path.substring(1);
      // if(path.trim().isEmpty()) return root;
      idx = Integer.parseInt(path);

      HTMLNode newCommonNode = new NodeImpl(commonNode.getValue(), commonNode.getName(),
          TypeToken.TAG);
      List<HTMLNode> children = commonNode.getChildren();
      for (i = idx; i < children.size(); i++) {
        newCommonNode.addChild(children.get(i));
        // children.get(i).setParent(newCommonNode);
      }
      return newCommonNode;
    } catch (Exception e) {
      StringBuilder builder = new StringBuilder();
View Full Code Here
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
 
  public void handle(HTMLNode node) {
    if(node == null) return;
    NodeIterator iterator = node.iterator();
    while(iterator.hasNext()) {
      HTMLNode n = iterator.next();
     
      handleNode(n);
     
      if(!n.isTag()) continue;
      Attribute attr = getAttribute(n);
      if(attr == null) continue;
      String attrValue = attr.getValue();
      if(attrValue == null) continue;
      if (verifier == null || verifier.verify(attrValue)) list.add(attrValue);
View Full Code Here
11
12
13
14
15
16
17
18
19
20
21
22
* Oct 12, 2007 
*/
public abstract class HTMLNodePath {
 
  protected String getIndexPath(HTMLNode element){
    HTMLNode parent = element.getParent();
    HTMLNode child = element;
    StringBuilder path = new StringBuilder();
    while(parent != null){
      if(path.length() > 0) path.insert(0, '.');     
      path.insert(0, parent.getChildren().indexOf(child));
      child = parent;
View Full Code Here
28
29
30
31
32
33
34
35
36
37
38
39
40
  public HTMLNode getNodeByIndex(HTMLNode node, String indexPath) throws Exception {  
    if(indexPath == null || indexPath.trim().length() < 1) {
      throw new NullPointerException("path is empty or null");
    }
    String [] split = indexPath.split("\\.");
    HTMLNode ele = node; 
    for(String element : split) {
      ele = node.getChildren().get(Integer.parseInt(element));
      if(ele.getChildren() == null) break;
      node = ele;    
    }
    return ele;
  } 
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
88
    handler = new IdentifierAttributeHandler(list, jsVerifier, "a", "href");
    handler.handle(root);
   
    NodeIterator iterator = root.iterator();
    while(iterator.hasNext()) {
      HTMLNode n = iterator.next();
      if(!n.isNode(Name.SCRIPT)) continue;
      if(n.getChildren().size() < 1) continue;
      list.add(n.getChild(0).getTextValue());
    }
    return list;
  }
View Full Code Here
157
158
159
160
161
162
163
164
165
  public synchronized void createFullLink(HTMLNode node,
      Map<String, String> map, URL home, ValueVerifier verifier) {
    if(node == null) return;
    NodeIterator iterator = node.iterator();
    while(iterator.hasNext()) {
      HTMLNode n = iterator.next();
      if(n.isTag()) createFullSingleLink(n, map, home, verifier);
    }
  }
View Full Code Here
196
197
198
199
200
201
202
203
204
 
  public void createFullLink(HTMLNode node,
      String nodeName, String attrName, URL home, ValueVerifier verifier) {
    NodeIterator iterator = node.iterator();
    while(iterator.hasNext()) {
      HTMLNode n = iterator.next();
      if(n.isTag()) createFullSingleLink(n, nodeName, attrName, home, verifier);
    }
  }
View Full Code Here
15
16
17
18
19
20
21
22
23
24
25
26
27
28
*/
public class HTMLParentUtils {
 
  public HTMLNode getUpParent(List<HTMLNode> nodes) {
    if(nodes.size() < 1) return null;
    HTMLNode parent  = nodes.get(0);
    while(parent != null) {
      if(isChild(parent, nodes, 1)) {
        return parent;
      }
      parent = parent.getParent();
    }
    return null;
  }
View Full Code Here
TOP

Related Classes of org.vietspider.html.HTMLNode

  • AttributeExplorer
  • ExtractForum
  • HTMLNodeExplorer
  • JavascriptExtractor
  • org.vietspider.browser.LoginUtils
  • org.vietspider.browser.LoginWebsiteHandler
  • org.vietspider.browser.VirtualFormPost
  • org.vietspider.browser.VirtualFormPostBak
  • org.vietspider.document.util.ImageDescRemover
  • org.vietspider.document.util.NodeRemover
    • 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.