Package org.cyberneko.html

Examples of org.cyberneko.html.HTMLConfiguration

This configuration recognizes the following properties:

For complete usage information, refer to the documentation. @see HTMLScanner @see HTMLTagBalancer @see HTMLErrorReporter @author Andy Clark @version $Id: HTMLConfiguration.java,v 1.9 2005/02/14 03:56:54 andyc Exp $

136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
    htmlScanner.scanDocument(true);
    return handler;
  }

  protected HTMLConfiguration newConfiguration() {
    HTMLConfiguration config = new HTMLConfiguration();
    // Maintain original case for elements and attributes
    config.setProperty("http://cyberneko.org/html/properties/names/elems", "match");
    config.setProperty("http://cyberneko.org/html/properties/names/attrs", "no-change");
    // Parse as fragment.
    config.setFeature("http://cyberneko.org/html/features/balance-tags/document-fragment", true);
    // Get notified of entity and character references
    config.setFeature("http://apache.org/xml/features/scanner/notify-char-refs", true);
    config.setFeature("http://cyberneko.org/html/features/scanner/notify-builtin-refs", true);
    return config;
  }
View Full Code Here
53
54
55
56
57
58
59
60
61
    return true;
  }

  @Override
  protected HTMLConfiguration newConfiguration() {
    HTMLConfiguration config = super.newConfiguration();
    config.setFeature("http://xml.org/sax/features/namespaces", true);
    return config;
  }
View Full Code Here
95
96
97
98
99
100
101
102
103
104
105
  @Override
  protected Document parseDomImpl(String source) throws GadgetException {
    DocumentHandler handler;

    HTMLConfiguration config = newConfiguration();
    try {
      handler = parseHtmlImpl(source, config, new NormalizingTagBalancer());
    } catch (IOException ioe) {
      return null;
    }
View Full Code Here
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
  @Override
  protected DocumentFragment parseFragmentImpl(String source) throws GadgetException {
    DocumentHandler handler;

    HTMLConfiguration config = newConfiguration();
    // http://cyberneko.org/html/features/balance-tags/document-fragment
    // deprecated http://cyberneko.org/html/features/document-fragment
    config.setFeature("http://cyberneko.org/html/features/balance-tags/document-fragment", true);
    config.setProperty("http://cyberneko.org/html/properties/balance-tags/fragment-context-stack",
        new QName[]{new QName(null, "HTML", "HTML", null), new QName(null, "BODY", "BODY", null)});

    try {
      handler = parseHtmlImpl(source, config, new NekoPatchTagBalancer());
    } catch (IOException ioe) {
View Full Code Here
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
      bodyFirst = headScript;
    }
  }

  protected HTMLConfiguration newConfiguration() {
    HTMLConfiguration config = new HTMLConfiguration();
    // Maintain original case for elements and attributes
    config.setProperty("http://cyberneko.org/html/properties/names/elems", "match");
    config.setProperty("http://cyberneko.org/html/properties/names/attrs", "no-change");
    // Get notified of entity and character references
    config.setFeature("http://apache.org/xml/features/scanner/notify-char-refs", true);
    config.setFeature("http://cyberneko.org/html/features/scanner/notify-builtin-refs", true);
    config.setFeature("http://xml.org/sax/features/namespaces", true);
    return config;
  }
View Full Code Here
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
        public HtmlSaxParser(Properties properties) {
            super(getConfig(properties));
        }

        private static HTMLConfiguration getConfig(Properties properties) {
            HTMLConfiguration config = new HTMLConfiguration();
            config.setProperty("http://cyberneko.org/html/properties/names/elems", "lower");
            if (properties != null) {
                for (Iterator i = properties.keySet().iterator();i.hasNext();) {
                    String name = (String) i.next();
                    if (name.indexOf("/features/") > -1) {
                        config.setFeature(name, Boolean.getBoolean(properties.getProperty(name)));
                    } else if (name.indexOf("/properties/") > -1) {
                        config.setProperty(name, properties.getProperty(name));
                    }
                }
            }
            return config;
        }
View Full Code Here
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
        public HtmlSaxParser(Properties properties) {
            super(getConfig(properties));
        }
   
        private static HTMLConfiguration getConfig(Properties properties) {
            HTMLConfiguration config = new HTMLConfiguration();
            config.setProperty("http://cyberneko.org/html/properties/names/elems", "lower");
            if (properties != null) {
                for (Iterator i = properties.keySet().iterator(); i.hasNext();) {
                    String name = (String) i.next();
                    if (name.indexOf("/features/") > -1) {
                        config.setFeature(name, Boolean.getBoolean(properties.getProperty(name)));
                    } else if (name.indexOf("/properties/") > -1) {
                        config.setProperty(name, properties.getProperty(name));
                    }
                }
            }

            return config;
View Full Code Here
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
    public NekoHtmlSaxParser(Properties properties) {
        super(getConfig(properties));
    }

    private static HTMLConfiguration getConfig(Properties properties) {
        HTMLConfiguration config = new HTMLConfiguration();
        config.setProperty("http://cyberneko.org/html/properties/names/elems", "lower");
        if (properties != null) {
            for (Iterator i = properties.keySet().iterator(); i.hasNext();) {
                String name = (String) i.next();
                if (name.indexOf("/features/") > -1) {
                    config.setFeature(name, Boolean.getBoolean(properties.getProperty(name)));
                } else if (name.indexOf("/properties/") > -1) {
                    config.setProperty(name, properties.getProperty(name));
                }
            }
        }
        return config;
    }
View Full Code Here
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
        ElementRemover remover = new ElementRemover();
        remover.removeElement(REMOVE_SCRIPT);
        StringWriter contentWriter = new StringWriter();
        Writer writer = new Writer(contentWriter, CHAR_ENCODING);
        XMLDocumentFilter[] filters = { remover, writer, };
        XMLParserConfiguration parser = new HTMLConfiguration();
        parser.setProperty("http://cyberneko.org/html/properties/filters",
                filters);
        XMLInputSource source = new XMLInputSource(null, null, null,
                contentReader, CHAR_ENCODING);
        try {
            parser.parse(source);
        } catch (XNIException e) {
            throw new NotIndexableException("Can not parse html -- ", e);

        } catch (IOException e) {
            throw new NotIndexableException("Can not parse html -- ", e);
View Full Code Here
33
34
35
36
37
38
39
40
41
42
43
44
45
      new org.cyberneko.html.filters.Writer(new ByteArrayOutputStream(), "UTF-8"),
      new org.cyberneko.html.filters.Writer(new ByteArrayOutputStream(), "UTF-8")
    };
   
        // create HTML parser
    final XMLParserConfiguration parser = new HTMLConfiguration();
        parser.setProperty("http://cyberneko.org/html/properties/filters", filters);

    XMLInputSource source = new XMLInputSource(null, "currentUrl", null, inputStream, "UTF-8");
       
      parser.parse(source);
        inputStream.close();
  }
View Full Code Here
TOP

Related Classes of org.cyberneko.html.HTMLConfiguration

  • com.enigmastation.extractors.impl.NekoHTMLStripper
  • com.gargoylesoftware.htmlunit.html.HTMLParser$HtmlUnitDOMBuilder
  • com.googlecode.memwords.facade.cards.parser.FavIconSaxParser
  • com.meterware.httpunit.parsing.NekoDOMParser
  • maqetta.zazl.HTMLParser
  • nokogiri.internals.HtmlDomParserContext
  • org.ajax4jsf.webapp.nekko.NekkoParser
  • org.apache.cocoon.components.NekoHtmlSaxParser
  • org.apache.cocoon.generation.NekoHTMLGenerator$HtmlSaxParser
  • org.apache.cocoon.portal.util.HtmlSaxParser
    • 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.