Package org.htmlcleaner

Examples of org.htmlcleaner.CleanerProperties

73
74
75
76
77
78
79
80
81
82
83
        xml = CommonUtil.toXml((Node)node, keepHeader);
        return CommonUtil.toHTML(xml);
      }else if (node instanceof TagNode){
        StringWriter sw = new StringWriter();
        //TODO 从配置文件里加载这个CleanerProperties
        CleanerProperties prop = new HtmlCleaner().getProperties();
        SimpleXmlSerializer ser = new SimpleXmlSerializer(prop);
        ser.write((TagNode)node, sw, "UTF-8");
          String html = sw.getBuffer().toString();
          if (keepHeader)
            xml = html;
View Full Code Here
26
27
28
29
30
31
32
33
34
35
36
37
    StringBuffer sbResult = new StringBuffer();
    try
    {
      HtmlCleaner cleaner = new HtmlCleaner();
     
      CleanerProperties props = cleaner.getProperties();
      props.setUseEmptyElementTags(false);
     
      TagNode node = cleaner.clean(sb.toString());
      Document myJDom = new JDomSerializer(props, true).createJDom(node);
      XMLOutputter outputter = new XMLOutputter(Format.getPrettyFormat());
      sbResult.append(outputter.outputString(myJDom));
View Full Code Here
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
    this.html = html;

    // create an instance of HtmlCleaner
    cleaner = new HtmlCleaner();
    // take default cleaner properties
    CleanerProperties props = cleaner.getProperties();

    // customize cleaner's behaviour with property setters
    // props.setXXX(...);
    props.setOmitXmlDeclaration(true);
    props.setOmitHtmlEnvelope(true);

    // Clean HTML taken from simple string, file, URL, input stream,
    // input source or reader. Result is root node of created
    // tree-like structure. Single cleaner instance may be safely used
    // multiple times.
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
    this.url = url;

    // create an instance of HtmlCleaner
    cleaner = new HtmlCleaner();
    // take default cleaner properties
    CleanerProperties props = cleaner.getProperties();

    // customize cleaner's behaviour with property setters
    // props.setXXX(...);
    props.setOmitXmlDeclaration(true);
    props.setOmitHtmlEnvelope(true);

    // Clean HTML taken from simple string, file, URL, input stream,
    // input source or reader. Result is root node of created
    // tree-like structure. Single cleaner instance may be safely used
    // multiple times.
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
  {
    // serialize a node to a file, output stream, DOM, JDom...
    System.out.println( "\n\n\nINICIO");

    HtmlCleaner cleaner = new HtmlCleaner();
    CleanerProperties props = cleaner.getProperties();
    PrettyXmlSerializer serializer = new PrettyXmlSerializer( props);
 
    for( int i=0 ; i<nodes.length ; i++) {
      System.out.println( "######################## Va uno " + i);
View Full Code Here
20
21
22
23
24
25
26
27
28
29
30
     
      // create an instance of HtmlCleaner
    HtmlCleaner cleaner = new HtmlCleaner();

    // take default cleaner properties
    CleanerProperties props = cleaner.getProperties();

    // customize cleaner's behaviour with property setters
//    props.setXXX(...);

    // Clean HTML taken from simple string, file, URL, input stream,
View Full Code Here
23
24
25
26
27
28
29
30
31
32
33
     
      // create an instance of HtmlCleaner
    HtmlCleaner cleaner = new HtmlCleaner();

    // take default cleaner properties
    CleanerProperties props = cleaner.getProperties();

    // customize cleaner's behaviour with property setters
//    props.setXXX(...);

    // Clean HTML taken from simple string, file, URL, input stream,
View Full Code Here
23
24
25
26
27
28
29
30
31
32
33
34
    {
        // 1. get ally
      String page = HttpServer.getHttpServer().httpGetPage( "http://s3.travian.net/allianz.php?aid=" + ally.getId());
       
        HtmlCleaner cleaner = new HtmlCleaner();
        CleanerProperties props = cleaner.getProperties();
        props.setRecognizeUnicodeChars( true);
//        TagNode node = cleaner.clean( new File("tests/ally.html"));
        TagNode node = cleaner.clean( page);

        updatePlayers( ally, node);
      
View Full Code Here
169
170
171
172
173
174
175
176
177
178
179
180
        // 1. get overview/innerview pantalla actual
      String page = HttpServer.getHttpServer().httpGetPage( "http://s3.travian.net/spieler.php?uid=" + player.getId());
      Thread.sleep( (long) (1000 + 3000*Math.random()));

        HtmlCleaner cleaner = new HtmlCleaner();
        CleanerProperties props = cleaner.getProperties();
        props.setRecognizeUnicodeChars( true);
//        TagNode node = cleaner.clean( new File("tests/perfil.html"));
        TagNode node = cleaner.clean( page);

        updateMapVillages2( player, node);

View Full Code Here
19
20
21
22
23
24
25
26
27
28
29
30
  {
    // 1. get overview/innerview pantalla actual
//    String page = HttpServer.getHttpServer().httpGetPage( "http://s3.travian.net/dorf1.php");
     
    HtmlCleaner cleaner = new HtmlCleaner();
    CleanerProperties props = cleaner.getProperties();
    props.setRecognizeUnicodeChars( true);
        TagNode node = cleaner.clean( new File("tests/overview8.html"));
//        TagNode node = cleaner.clean( page);

        updateVillages2( villages, node);
View Full Code Here
TOP

Related Classes of org.htmlcleaner.CleanerProperties

  • at.newmedialab.ldpath.model.functions.CleanHtmlFunction
  • ch.entwine.weblounge.preview.xhtmlrenderer.XhtmlRendererPagePreviewGenerator
  • com.atlantbh.nutch.filter.xpath.XPathHtmlParserFilter
  • com.cubusmail.mail.text.test.HtmlParserTest
  • com.ikanow.infinit.e.harvest.enrichment.custom.UnstructuredAnalysisHarvester
  • com.jeck.microblogging.utils.HtmlUtils
  • com.netfever.site.dynovisz.tools.utils.XmlUtils
  • fiftyfive.wicket.test.WicketTestUtils
  • net.sf.jwan.servlet.util.JwanHtmlFormatter
  • net.traviangui.hostInterface.IOAlly
    • 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.