Package org.vietspider.chars.refs

Examples of org.vietspider.chars.refs.RefsDecoder

105
106
107
108
109
110
111
112
113
114
115
    return document == null || ref == null ?
        document : new HTMLAnchorUtil().searchDocument(document, ref);
  }
 
  public NodePath findNodeByText(HTMLNode node, String start, String end) throws Exception {
    RefsDecoder decoder = new RefsDecoder();
    NodePathParser pathParser = new NodePathParser();
    TextHandler textHandler = new TextHandler();
    if(start == null || start.trim().length() == 0) return pathParser.toPath(node);
    start = textHandler.trim(start);
    HTMLNode startNode = textHandler.findByText(node, start, decoder);
View Full Code Here
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
   
    return values;
  }
 
  protected HTMLNode searchBody(HTMLDocument document) throws Exception {
    RefsDecoder decoder = new RefsDecoder();
    NodeIterator iterator = document.getRoot().iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.CONTENT)) continue;
      char [] chars = node.getValue();
      chars = decoder.decode(chars);

      chars = CharsUtil.cutAndTrim(chars, 0, chars.length);
      chars =  java.text.Normalizer.normalize(new String(chars), Normalizer.Form.NFC).toCharArray();
      node.setValue(chars);             
    } 
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
    List<HTMLNode> ignores = searchBadNodes2(body, checkers);
    return new ContentRenderer(body, ignores, linkNodeChecker);
  }

  public static HTMLNode searchBody(HTMLDocument document) throws Exception {
    RefsDecoder decoder = new RefsDecoder();
    NodeIterator iterator = document.getRoot().iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.CONTENT)) continue;
      char [] chars = node.getValue();
      chars = decoder.decode(chars);

      chars = CharsUtil.cutAndTrim(chars, 0, chars.length);
      chars =  java.text.Normalizer.normalize(new String(chars), Normalizer.Form.NFC).toCharArray();
      node.setValue(chars);             
    } 
View Full Code Here
41
42
43
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
    nodeName = nodeName.toLowerCase();
   
    text = text.substring(nodeName.length());
    text = trimText(text);

    RefsDecoder refsDecoder = new RefsDecoder();
    while(text.length() > 0) {
      Object [] components = getName(text);
      if(components == null) return list;
      String name = (String)components[0];
      text = (String)components[1];
     
      if(components.length > 2) {
        list.add(new Attribute(name, null));
        continue;
      }
     
      String value = "";
      char mark = '\"';
      if(text.length() > 0) {
        components = getValue(text);
        value = (String)components[0];
        text = (String)components[1];
        mark = (Character)components[2];
      }
     
      if(mark != '\"' && mark != '\'') mark = '\"';
      if(value.indexOf(REF_START) > -1) {
        value = new String(refsDecoder.decode(value.toCharArray()));
      }
      list.addElement(new Attribute(name, value, mark));
    }
    parseStyle(list);
    return list;
View Full Code Here
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
 
  final static public Attribute[] parse(String text) {
    List<Attribute> list = new ArrayList<Attribute>();
    text = trimText(text);

    RefsDecoder refsDecoder = new RefsDecoder();
    while(text.length() > 0) {
      Object [] components = getName(text);
      if(components == null) return list.toArray(new Attribute[list.size()]);
      String name = (String)components[0];
      text = (String)components[1];
     
      if(components.length > 2) {
        list.add(new Attribute(name, null));
        continue;
      }
     
      String value = "";
      char mark = '\"';
      if(text.length() > 0) {
        components = getValue(text);
        value = (String)components[0];
//        System.out.println(" ========== > "+ value);
        text = (String)components[1];
        mark = (Character)components[2];
      }
     
      if(mark != '\"' && mark != '\'') mark = '\"';
      if(value.indexOf(REF_START) > -1) {
        value = new String(refsDecoder.decode(value.toCharArray()));
      }
      list.add(new Attribute(name, value));
    }
    return list.toArray(new Attribute[list.size()]);
  } 
View Full Code Here
33
34
35
36
37
38
39
  static {
    Arrays.sort(URICS);
  }
 
  public URLUtilsBak() {
    decoder = new RefsDecoder();
  }
View Full Code Here
44
45
46
47
48
49
50
51
52
53
  } 
 
  public HTMLDocument createDocument(byte [] bytes) throws Exception {
    if(charset != null) {
      char [] chars = CharsDecoder.decode(charset, bytes, 0, bytes.length);
      if(decode) chars = new RefsDecoder().decode(chars);
      return createDocument(chars);
    }
    return detectDocument(bytes);
  }
View Full Code Here
53
54
55
56
57
58
59
60
  }
 
  private HTMLDocument detectDocument(byte [] bytes) throws Exception {
    this.charset = detectCharset(bytes);
    char [] chars = CharsDecoder.decode(charset, bytes, 0, bytes.length);
    if(decode) chars = new RefsDecoder().decode(chars);
    return createDocument(chars);
  }
View Full Code Here
99
100
101
102
103
104
105
106
107
108
109
110
111
  }
 
  public List<NodeImpl> createTokens(byte [] bytes) throws Exception {
    if(charset != null) {
      char [] chars = CharsDecoder.decode(charset, bytes, 0, bytes.length);
      if(decode) chars = new RefsDecoder().decode(chars);
      return createTokens(chars);
    }
    this.charset = detectCharset(bytes);
    char [] chars = CharsDecoder.decode(charset, bytes, 0, bytes.length);
    if(decode) chars = new RefsDecoder().decode(chars);
    return createTokens(chars);
  }
View Full Code Here
8
9
10
11
12
13
14
15
16
17
18
19
20
*          Email:nhudinhthuan@yahoo.com
* Nov 19, 2006
*/
public class DecodeExample {
  public static void main(String arg[]){
    RefsDecoder ref = new RefsDecoder();
    String text = "&nbsp;&nbsp;&nbsp;" ;
    String value = new String(ref.decode(text.toCharArray()));
    System.out.println("|"+value+"|");
    RefsEncoder encoder = new RefsEncoder(false);
    value = value + "&";
    text = new String(encoder.encode(value.toCharArray()));
    System.out.println(text);
View Full Code Here
TOP

Related Classes of org.vietspider.chars.refs.RefsDecoder

  • DecodeExample
  • org.vietspider.html.renderer.ContentRegionSearcher
  • org.vietspider.html.renderer.ContentRendererFactory
  • org.vietspider.html.util.HTMLParserDetector
  • org.vietspider.html.util.URLUtilsBak
  • org.vietspider.token.attribute.AttributeParser
  • org.vietspider.ui.htmlexplorer.FastWebClient
    • 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.