Package org.vietspider.token.attribute

Examples of org.vietspider.token.attribute.Attributes

152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
 
  private void createFullSingleLink(HTMLNode node,
      String nodeName, String attrName, URL home, ValueVerifier verifier)   {
    Attribute attr = null;
    if(node.isNode(nodeName) || (nodeName.length() == 1 || nodeName.charAt(0) == '*')) {
      Attributes attrs = node.getAttributes(); 
      int idx = attrs.indexOf(attrName);
      if(idx < 0)  return;
      attr = attrs.get(idx);
      String value = attr.getValue();
      if(verifier != null && !verifier.verify(value)) return;
      value  = urlCreator.createURL(home, value);     
      attr.setValue(value);     
      attrs.set(attr);
    }
  }
View Full Code Here
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
    NodeIterator iterator = root.iterator();
    while(iterator.hasNext()) {
      HTMLNode n = iterator.next();
      if(!n.isNode(Name.META)) continue;
      if(n.isNode(Name.BODY)) break;
      Attributes attributes = n.getAttributes();
      Attribute attribute = attributes.get("http-equiv");
      if(attribute == null || attribute.getValue() == null) continue;

      if(!"content-type".equalsIgnoreCase(attribute.getValue().trim())) continue ;

      attribute = attributes.get("content");
      if(attribute == null) continue;
      String link = attribute.getValue();
      if(link == null) continue;
      int index = link.toLowerCase().indexOf("=");
View Full Code Here
126
127
128
129
130
131
132
133
134
135
136
    return list;
  }
 
  public Attribute getAttribute(HTMLNode node, String nodeName, String attrName){
    if(node.isNode(nodeName) || (nodeName.length() == 1 && nodeName.charAt(0) == '*')){
      Attributes attrs = node.getAttributes();  
      int idx = attrs.indexOf(attrName);
      if(idx > -1) return attrs.get(idx);
    }
    return null;
  }
View Full Code Here
139
140
141
142
143
144
145
146
147
148
149
150
    Set<String> keys = map.keySet();
    Iterator<String> iter = keys.iterator();
    while(iter.hasNext()){
      String key = iter.next();
      if(node.isNode(key) || (key.length() == 1 && key.charAt(0) == '*')){
        Attributes attrs = node.getAttributes();  
        int idx = attrs.indexOf(map.get(key));
        if(idx > -1) return attrs.get(idx);
      }
    }
    return null;
  }
View Full Code Here
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
    downloadResources(address, tokens, resources);

    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl token = tokens.get(i);
      if(token.getType() != TypeToken.TAG || !token.isNode(Name.A)) continue;
      Attributes attributes = token.getAttributes();
      Attribute attribute = attributes.get("href");
      if(attribute == null) continue;
      String link  = attribute.getValue();
      if(link == null || link.trim().length() < 1) continue;
      link  = urlUtils.createURL(parent, link);
      String subName = "";
      try {
        URL subUrl = new URL(link);
        String temp = subUrl.getPath();
        if(temp != null) subName += temp;
        temp = subUrl.getQuery();
        if(temp != null) subName += "_"+temp;
        subName = toName(subName);
      }catch (Exception e) {
      }
      if(subName == null || subName.trim().length() < 1) subName = toName(link);
      try{
        crawl(new URL(address), link, subName, level+1, depth);
      }catch (Exception e) {
        System.err.println(e);
      }
      attribute.setValue(subName+".html");
      attributes.set(attribute);
    }

   
    StringBuilder contentBuilder = new StringBuilder();
    for(int i = 0; i < tokens.size(); i++) {
View Full Code Here
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
      NodeImpl token = tokens.get(i);
      if(token.getType() != TypeToken.TAG) continue;
      for(Resource resource : resources) {
        if(!token.isNode(resource.tag))  continue;
        try {
          Attributes attributes = token.getAttributes();
          for(Attribute attribute : attributes) {
            if(!attribute.getName().equalsIgnoreCase(resource.attr))  continue;
            String rscName = toName(attribute.getValue());
            File file = new File(folder, rscName);
            String link = urlUtils.createURL(url, attribute.getValue());
            System.out.println("downloading "+link+" ...");
            if(!file.exists()) loadResource(referer, link, file);
            attribute.setValue(rscName);
            attributes.set(attribute);
          }
        } catch (Exception e) {
          System.err.println(e);
        }
      }
View Full Code Here
44
45
46
47
48
49
50
51
52
53
    builder.append("style=\"width: 80px;\" type='text' ");
    builder.append(" src='resources/2D-47.gif'");
    node.setValue(builder.toString().toCharArray());
   
   
    Attributes attrs = AttributeParser.parse(node);
    for(Attribute attr : attrs){
      System.out.println(attr.getName()+" : "+attr.getValue());
    }
  }
View Full Code Here
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
public class RemoveAttribute {

  static void clean(HTMLNode node, Map<String, String[]> map){
    Iterator<String> iter = map.keySet().iterator();
    String key, value [];
    Attributes attributes = node.getAttributes();
    while(iter.hasNext()){
      key = iter.next();
      if(key.equals("*") || node.isNode(key)){
        value = map.get(key);
        for(String ele : value){
          if(attributes.contains(ele)){
            attributes.remove(ele);
          }
        }
      }
    }
    List<HTMLNode> children = node.getChildren();
View Full Code Here
39
40
41
42
43
44
45
46
47
48
49
50
   
    NodeIterator iterator = document.getRoot().iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(node.isNode(nodeName)) {
        Attributes attributes = node.getAttributes();
        Attribute attribute = attributes.get(attrName);
        if(attribute == null) continue;
        attrValues.add(attribute.getValue());
      }
    }
    return attrValues;
View Full Code Here
TOP

Related Classes of org.vietspider.token.attribute.Attributes

  • AttributeExample
  • AttributeExplorer
  • JavascriptExtractor
  • org.vietspider.browser.form.FormUtils
  • org.vietspider.browser.LoginWebsiteHandler
  • org.vietspider.browser.VirtualFormPost
  • org.vietspider.browser.VirtualFormPostBak
  • org.vietspider.html.path2.HTMLExtractor
  • org.vietspider.html.path2.LookupNode
  • org.vietspider.html.renderer.checker.ContentChecker
    • 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.