Package org.vietspider.token.attribute

Examples of org.vietspider.token.attribute.Attribute

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
    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();
View Full Code Here
142
143
144
145
146
147
148
149
150
151
152
153
154
     
      if(current != null && current instanceof Collection) {
        collection = (Collection) current;
      } else if(type.isInterface())  {
        Attributes attributes = AttributeParser.parse(node);
        Attribute attr = attributes.get("type");
        if(attr != null) {
          String typeName = attr.getValue();
          Class<?> clazz = Thread.currentThread().getContextClassLoader().loadClass(typeName);
          collection = (Collection) clazz.newInstance();
        } else {
          collection = new ArrayList<Object>();
        }
View Full Code Here
204
205
206
207
208
209
210
211
212
213
214
215
216
   
    if(current != null && current instanceof Map<?,?>) {
      map = (Map<Object, Object>) current;
    } else if(type.isInterface())  {
      Attributes attributes = AttributeParser.parse(node);
      Attribute attr = attributes.get("type");
      if(attr != null) {
        String typeName = attr.getValue();
        Class<?> clazz = Thread.currentThread().getContextClassLoader().loadClass(typeName);
        map =  (Map<Object, Object>) clazz.newInstance();
      } else {
        map = new Hashtable<Object, Object>();
      }
View Full Code Here
347
348
349
350
351
352
353
354
      if(ele.getName().equals(name)){
        ele.setValue(value);
        return ;
      }
    }
    Attribute attr = new Attribute(name, value);
    attrs.add(attr);
  }
View Full Code Here
103
104
105
106
107
108
109
110
111
112
113
114
115
      if(child.isNode(LINK) && child.getChildren().size() > 0) {
        item.addLink(new MetaLink(child.getChild(0).getTextValue()));
      } else if(child.isNode(ENCLOSURE)) {
        Attributes attributes = AttributeParser.parse(child);
        MetaLink metaLink  = new MetaLink();
        Attribute attribute = attributes.get(URL);
        if(attribute != null) metaLink.setHref(attribute.getValue());
        attribute = attributes.get(TYPE);
        if(attribute != null) metaLink.setType(attribute.getValue());
        item.addLink(metaLink);
      }
    }
  }
View Full Code Here
122
123
124
125
126
127
128
129
130
131
132
    for(XMLNode child : children) {
      if(!child.isNode(LINK)) continue;
      MetaLink metaLink = toBean.toBean(MetaLink.class, child);
      if(metaLink.getHref() == null || metaLink.getHref().trim().isEmpty()) {
        Attributes attributes = AttributeParser.parse(child);
        Attribute attribute = attributes.get(REF);
        if(attribute != null) metaLink.setHref(attribute.getValue());
      }
      item.addLink(metaLink);
    }
  }
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
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
//        System.out.println("== > thay "+ node + " : "+ node.isOpen()  );
        if(node.getType() == TypeToken.TAG) {
          if(form != null) forms.add(form);
          form = new Form();
          Attributes attributes = node.getAttributes();
          Attribute attribute = attributes.get(NAME_ATTR);
          if(attribute != null) form.setName(attribute.getValue());
          attribute = attributes.get("method");
          if(attribute != null) form.setMethod(attribute.getValue());
          attribute = attributes.get(ACTION);
          if(attribute != null) form.setAction(attribute.getValue());
        } else {
          if(form != null) forms.add(form);
          form = null;
        }
      } else if(node.isNode(Name.INPUT)
          || node.isNode(Name.TEXTAREA)
          || node.isNode(Name.SELECT)) {       
        if(form == null) continue;
       
        Attributes attributes = node.getAttributes();
        Attribute attribute = attributes.get(NAME_ATTR);
        if(attribute == null) attribute = attributes.get(ID_ATTR);;
        if(attribute == null) continue;
        String name  = attribute.getValue();
        if(name == null || name.trim().isEmpty()) continue;
       
        attribute = attributes.get(TYPE_ATTR);
        String type = null;
        if(attribute != null) type = attribute.getValue();
        if(type == null) type = "text";
        type = type.trim();
       
        if(type.equalsIgnoreCase("submit")) {
          if(name.toLowerCase().indexOf("cancel") > -1
              || name.toLowerCase().indexOf("reset") > -1
              || name.toLowerCase().indexOf("clear") > -1) continue;
        }
       
        if("radio".equalsIgnoreCase(type)
            || "checkbox".equalsIgnoreCase(type)) {
          attribute = attributes.get("checked");
          if(attribute == null) continue;
        }
       
        attribute = attributes.get(VALUE);
        String value = null;
        if(attribute != null) value = attribute.getValue();
        if(value == null) value = "";
        value  = value.trim();
       
        boolean newInput = true;
        for(int k = 0; k < form.getParams().size(); k++) {
View Full Code Here
43
44
45
46
47
48
49
50
51
52
53
54
55
  
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.DIV)) continue;
      Attributes attributes = node.getAttributes();
      Attribute attribute = attributes.get("id");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
     
      if("posts".equalsIgnoreCase(value)) {
        HTMLNode titleNode = searchTitleNode(root, node);
        if(titleNode != null) {
View Full Code Here
148
149
150
151
152
153
154
155
156
157
158
159
160
    NodeIterator iterator = root.iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.A)) continue;
      Attributes attributes = node.getAttributes();
      Attribute attribute = attributes.get("class");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
      value = value.toLowerCase();
      if(value.indexOf("bigusername") > -1) return node;
    }
    return null;
View Full Code Here
164
165
166
167
168
169
170
171
172
173
174
175
176
    NodeIterator iterator = root.iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.DIV)) continue;
      Attributes attributes = node.getAttributes();
      Attribute attribute = attributes.get("id");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
      value = value.toLowerCase();
      if(value.indexOf(clazz) > -1) {//"post_message"
        return upParent(node, Name.TD);
      }
View Full Code Here
TOP

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

  • AttributeExplorer
  • org.vietspider.browser.form.FormUtils
  • org.vietspider.browser.VirtualFormPost
  • org.vietspider.html.path2.NodeMatcher
  • org.vietspider.html.renderer.checker.LinkBlockChecker
  • org.vietspider.html.renderer.checker.LinkNodeChecker
  • org.vietspider.html.renderer.extractor.WebPageDataSearcher
  • org.vietspider.html.renderer.LinkChecker
  • org.vietspider.html.renderer.LinkRemover
  • org.vietspider.html.renderer.NodeComparator
    • 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.