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 $

46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
    /** The node representing the document. **/
    private Node _documentNode;


    static NekoDOMParser newParser( DocumentAdapter adapter, URL url ) {
        final HTMLConfiguration configuration = new HTMLConfiguration();
        if (!HTMLParserFactory.getHTMLParserListeners().isEmpty() || HTMLParserFactory.isParserWarningsEnabled()) {
            configuration.setErrorHandler( new ErrorHandler( url ) );
            configuration.setFeature( REPORT_ERRORS, true);
        }
        configuration.setFeature( AUGMENTATIONS, true );
        final ScriptFilter javaScriptFilter = new ScriptFilter( configuration );
        configuration.setProperty( FILTERS, new XMLDocumentFilter[] { javaScriptFilter } );
        if (HTMLParserFactory.isPreserveTagCase()) {
            configuration.setProperty( TAG_NAME_CASE, "match" );
            configuration.setProperty( ATTRIBUTE_NAME_CASE, "no-change" );
        }

        try {
            final NekoDOMParser domParser = new NekoDOMParser( configuration, adapter );
            domParser.setFeature( DEFER_NODE_EXPANSION, false );
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
  @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
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
  @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
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
      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
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
   *
   * @return
   * @throws ServletException
   */
  protected HTMLConfiguration getHtmlConfig() {
    HTMLConfiguration _config = new HTMLConfiguration();
    try {
      if (this.getPublicid() != null || this.getSystemid() != null) {
        _config.setFeature(
            "http://cyberneko.org/html/features/insert-doctype",
            true);
        _config.setFeature(
            "http://cyberneko.org/html/features/override-doctype",
            true);
      }
      if (this.getPublicid() != null) {
        _config.setProperty(
            "http://cyberneko.org/html/properties/doctype/pubid",
            getPublicid());

      }
      if (this.getSystemid() != null) {
        _config.setProperty(
            "http://cyberneko.org/html/properties/doctype/sysid",
            getSystemid());

      }
      if (this.getNamespace() != null) {
        _config.setFeature("http://xml.org/sax/features/namespaces",
            true);
        _config
            .setFeature(
                "http://cyberneko.org/html/features/override-namespaces",
                true);
        _config.setFeature(
            "http://cyberneko.org/html/features/insert-namespaces",
            true);
        _config.setProperty(
            "http://cyberneko.org/html/properties/namespaces-uri",
            getNamespace());

      }
      // config
      // .setFeature(
      // "http://cyberneko.org/html/features/balance-tags/ignore-outside-content",
      // true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/cdata-sections",
              true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/script/strip-comment-delims",
              true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/style/strip-comment-delims",
              true);
      _config.setFeature(
          "http://cyberneko.org/html/features/insert-doctype", true);
      _config.setFeature(
          "http://cyberneko.org/html/features/insert-namespaces",
          true);
      // 
      // Set properties
      // http://cyberneko.org/html/features/insert-namespaces
      // _config
      // .setProperty(
      // "http://cyberneko.org/html/properties/default-encoding",
      // encoding);
      _config
          .setProperty(
              "http://cyberneko.org/html/properties/names/elems",
              "lower");
      _config
          .setProperty(
              "http://cyberneko.org/html/properties/names/attrs",
              "lower");
      _config.setProperty("http://cyberneko.org/html/properties/filters",
          _filters);
    } catch (XMLConfigurationException e) {
      // throw new ServletException("error set Neko feature ", e);
    }
    return _config;
View Full Code Here
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
   *
   * @return
   * @throws ServletException
   */
  protected HTMLConfiguration getHtmlConfig() {
    HTMLConfiguration _config = new HTMLConfiguration();
    try {
      if (this.getPublicid() != null || this.getSystemid() != null) {
        _config.setFeature(
            "http://cyberneko.org/html/features/insert-doctype",
            true);
        _config.setFeature(
            "http://cyberneko.org/html/features/override-doctype",
            true);
      }
      if (this.getPublicid() != null) {
        _config.setProperty(
            "http://cyberneko.org/html/properties/doctype/pubid",
            getPublicid());

      }
      if (this.getSystemid() != null) {
        _config.setProperty(
            "http://cyberneko.org/html/properties/doctype/sysid",
            getSystemid());

      }
      if (this.getNamespace() != null) {
        _config.setFeature("http://xml.org/sax/features/namespaces",
            true);
        _config
            .setFeature(
                "http://cyberneko.org/html/features/override-namespaces",
                true);
        _config.setFeature(
            "http://cyberneko.org/html/features/insert-namespaces",
            true);
        _config.setProperty(
            "http://cyberneko.org/html/properties/namespaces-uri",
            getNamespace());

      }
      // config
      // .setFeature(
      // "http://cyberneko.org/html/features/balance-tags/ignore-outside-content",
      // true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/cdata-sections",
              true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/script/strip-comment-delims",
              true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/style/strip-comment-delims",
              true);
      _config.setFeature(
          "http://cyberneko.org/html/features/insert-doctype", true);
      _config.setFeature(
          "http://cyberneko.org/html/features/insert-namespaces",
          true);
      // 
      // Set properties
      // http://cyberneko.org/html/features/insert-namespaces
      // _config
      // .setProperty(
      // "http://cyberneko.org/html/properties/default-encoding",
      // encoding);
      _config
          .setProperty(
              "http://cyberneko.org/html/properties/names/elems",
              "lower");
      _config
          .setProperty(
              "http://cyberneko.org/html/properties/names/attrs",
              "lower");
      _config.setProperty("http://cyberneko.org/html/properties/filters",
          _filters);
    } catch (XMLConfigurationException e) {
      // throw new ServletException("error set Neko feature ", e);
    }
    return _config;
View Full Code Here
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
   *
   * @return
   * @throws ServletException
   */
  protected HTMLConfiguration getHtmlConfig() {
    HTMLConfiguration _config = new HTMLConfiguration();
    try {
      if (this.getPublicid() != null || this.getSystemid() != null) {
        _config.setFeature(
            "http://cyberneko.org/html/features/insert-doctype",
            true);
        _config.setFeature(
            "http://cyberneko.org/html/features/override-doctype",
            true);
      }
      if (this.getPublicid() != null) {
        _config.setProperty(
            "http://cyberneko.org/html/properties/doctype/pubid",
            getPublicid());

      }
      if (this.getSystemid() != null) {
        _config.setProperty(
            "http://cyberneko.org/html/properties/doctype/sysid",
            getSystemid());

      }
      if (this.getNamespace() != null) {
        _config.setFeature("http://xml.org/sax/features/namespaces",
            true);
        _config
            .setFeature(
                "http://cyberneko.org/html/features/override-namespaces",
                true);
        _config.setFeature(
            "http://cyberneko.org/html/features/insert-namespaces",
            true);
        _config.setProperty(
            "http://cyberneko.org/html/properties/namespaces-uri",
            getNamespace());

      }
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/cdata-sections",
              true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/script/strip-comment-delims",
              true);
      _config
          .setFeature(
              "http://cyberneko.org/html/features/scanner/style/strip-comment-delims",
              true);
      _config.setFeature(
          "http://cyberneko.org/html/features/insert-doctype", true);
      _config.setFeature(
          "http://cyberneko.org/html/features/insert-namespaces",
          true);
      // 
      // Set properties
      // http://cyberneko.org/html/features/insert-namespaces
      // _config
      // .setProperty(
      // "http://cyberneko.org/html/properties/default-encoding",
      // encoding);
      _config
          .setProperty(
              "http://cyberneko.org/html/properties/names/elems",
              "lower");
      _config
          .setProperty(
              "http://cyberneko.org/html/properties/names/attrs",
              "lower");
      _config.setProperty("http://cyberneko.org/html/properties/filters",
          _filters);
       _config
       .setFeature(
       "http://cyberneko.org/html/features/balance-tags/document-fragment",
       true);
    } catch (XMLConfigurationException e) {
      // throw new ServletException("error set Neko feature ", e);
View Full Code Here
213
214
215
216
217
218
219
220
221
222
223
224
225
    // MAIN
    //

    /** Main. */
    public static void main(String[] argv) throws Exception {
        HTMLConfiguration parser = new HTMLConfiguration();
        parser.setFeature(AUGMENTATIONS, true);
        XMLDocumentFilter[] filters = { new Script(parser), new Identity(), new Writer() };
        parser.setProperty(FILTERS, filters);
        for (int i = 0; i < argv.length; i++) {
            parser.parse(new XMLInputSource(null, argv[i], null));
        }
    } // main(String[])
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
        if (fFileSets.size() == 0) {
            throw new BuildException("must specify at least one fileset");
        }

        // create parser
        XMLParserConfiguration parser = new HTMLConfiguration();

        // parse input files and produce output files
        log("Parsing test files and generating output...");
        File outdir = new File(outputdir);
        int size = fFileSets.size();
        for (int i = 0; i < size; i++) {
            FileSet fileset = (FileSet)fFileSets.elementAt(i);
            DirectoryScanner dirscanner = fileset.getDirectoryScanner(project);
            File indir = dirscanner.getBasedir();
            String[] files = dirscanner.getIncludedFiles();
            for (int j = 0; j < files.length; j++) {
                File infile = new File(indir, files[j]);
                File outfile = new File(outdir, files[j]);
                log("  "+outfile, Project.MSG_VERBOSE);
                OutputStream out = null;
                try {
                    out = new FileOutputStream(outfile);
                    XMLDocumentFilter[] filters = { new Writer(out) };
                    parser.setProperty("http://cyberneko.org/html/properties/filters", filters);
                    parser.parse(new XMLInputSource(null, infile.toString(), null));
                }
                catch (Exception e) {
                    log("  error parsing input file, "+infile);
                    throw new BuildException(e);
                }
View Full Code Here
122
123
124
125
126
127
128
129
130
131
132
    // Constructors
    //

    /** Default constructor. */
    public DOMFragmentParser() {
        fParserConfiguration = new HTMLConfiguration();
        fParserConfiguration.addRecognizedFeatures(RECOGNIZED_FEATURES);
        fParserConfiguration.addRecognizedProperties(RECOGNIZED_PROPERTIES);
        fParserConfiguration.setFeature(DOCUMENT_FRAGMENT, true);
        fParserConfiguration.setDocumentHandler(this);
    } // <init>()
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.