Examples of org.w3c.tidy.Tidy

org.w3c.tidy.Tidy
3.org/People/Raggett/tidy"> HTML Tidy Release 4 Aug 2000

Copyright (c) 1998-2000 World Wide Web Consortium (Massachusetts Institute of Technology, Institut National de Recherche en Informatique et en Automatique, Keio University). All Rights Reserved.

Contributing Author(s):
Dave Raggett
Andy Quick (translation to Java)

The contributing author(s) would like to thank all those who helped with testing, bug fixes, and patience. This wouldn't have been possible without all of you.

COPYRIGHT NOTICE:
This software and documentation is provided "as is," and the copyright holders and contributing author(s) make no representations or warranties, express or implied, including but not limited to, warranties of merchantability or fitness for any particular purpose or that the use of the software or documentation will not infringe any third party patents, copyrights, trademarks or other rights.

The copyright holders and contributing author(s) will not be liable for any direct, indirect, special or consequential damages arising out of any use of the software or documentation, even if advised of the possibility of such damage.

Permission is hereby granted to use, copy, modify, and distribute this source code, or portions hereof, documentation and executables, for any purpose, without fee, subject to the following restrictions:
1. The origin of this source code must not be misrepresented.
2. Altered versions must be plainly marked as such and must not be misrepresented as being the original source.
3. This Copyright notice may not be removed or altered from any source or altered source distribution.
The copyright holders and contributing author(s) specifically permit, without fee, and encourage the use of this source code as a component for supporting the Hypertext Markup Language in commercial products. If you use this source code in a product, acknowledgment is not required but would be appreciated.
@author Dave Raggett @author Andy Quick (translation to Java) @version 1.0, 1999/05/22 @version 1.0.1, 1999/05/29 @version 1.1, 1999/06/18 Java Bean @version 1.2, 1999/07/10 Tidy Release 7 Jul 1999 @version 1.3, 1999/07/30 Tidy Release 26 Jul 1999 @version 1.4, 1999/09/04 DOM support @version 1.5, 1999/10/23 Tidy Release 27 Sep 1999 @version 1.6, 1999/11/01 Tidy Release 22 Oct 1999 @version 1.7, 1999/12/06 Tidy Release 30 Nov 1999 @version 1.8, 2000/01/22 Tidy Release 13 Jan 2000 @version 1.9, 2000/06/03 Tidy Release 30 Apr 2000 @version 1.10, 2000/07/22 Tidy Release 8 Jul 2000 @version 1.11, 2000/08/16 Tidy Release 4 Aug 2000

        }
        return false;
    }
    
    public InputStream tidy(InputStream is) {
        Tidy tidy = new Tidy();
        tidy.setXHTML(false); 
        tidy.setPrintBodyOnly(true);
        tidy.setDocType("loose");
        tidy.setXHTML(true);
        tidy.setForceOutput(true);
        java.io.ByteArrayOutputStream out = new java.io.ByteArrayOutputStream(1024);
        tidy.parse(is,out);
        return new ByteArrayInputStream(out.toByteArray());
    }

View Full Code Here


        // there is also a javax.swing.text.Document class.
        org.w3c.dom.Document document = null;


        StringWriter sw = new StringWriter();
        Tidy tidy = XPathUtil.makeTidyParser(true, true, true, sw);
        document = tidy.parseDOM(baIS, null);
        document.normalize();
        if (tidy.getParseErrors() > 0) {
            showErrorMessageDialog(sw.toString(),
                    "Tidy: " + tidy.getParseErrors() + " errors, " + tidy.getParseWarnings() + " warnings",
                    JOptionPane.WARNING_MESSAGE);
        }


        JPanel domTreePanel = new DOMTreePanel(document);
        resultsScrollPane.setViewportView(domTreePanel);

View Full Code Here

        encoding = encoding.substring(0,i).trim();
      if (encoding.indexOf("\"")!=-1)
        encoding = encoding.substring(1,encoding.length()+1);
    }


    Tidy tidy = new Tidy ();
    tidy.setXHTML (true);
    tidy.setDocType ("omit");
    tidy.setQuiet(true);
    tidy.setShowWarnings(false);
    tidy.setNumEntities(true);
    tidy.setWord2000(true);


    // If charset is specified in header, set JTidy's
    // character encoding  to either UTF-8, ISO-8859-1
    // or ISO-2022 accordingly (NOTE that these are
    // the only character encoding sets that are supported in
    // JTidy).  If character encoding is not specified,
    // UTF-8 is the default.
    if (encoding != null)
    {
      if (encoding.toLowerCase().equals("iso-8859-1"))
        tidy.setCharEncoding(org.w3c.tidy.Configuration.LATIN1);
      else if (encoding.toLowerCase().equals("iso-2022-jp"))
        tidy.setCharEncoding(org.w3c.tidy.Configuration.ISO2022);
      else
        tidy.setCharEncoding(org.w3c.tidy.Configuration.UTF8);
    }
    else
    {
      tidy.setCharEncoding(org.w3c.tidy.Configuration.UTF8);
    }


    tidy.setErrout(devNull);


    ByteArrayOutputStream stream = new ByteArrayOutputStream (1024);
    BufferedOutputStream out = new BufferedOutputStream (stream);


    tidy.parse (urlConnect.getInputStream(), out);
    String tidiedXml = stream.toString();
    stream.close();
    out.close();


    if ( tidy.getParseErrors() > 0 )
      throw new GeneralRenderingException("Unable to convert input document to XHTML");


    return tidiedXml;
  }

View Full Code Here

    protected Document getHtmlDocument(Resource resource) {
        String contentType = resource.getContentType();


        if (contentType != null && contentType.startsWith("text/html")) {
            try {
                Tidy tidy = new Tidy();
                tidy.setQuiet(true);
                tidy.setXmlOut(true);
                tidy.setErrout(new PrintWriter(new ByteArrayOutputStream()));


                org.w3c.dom.Document dom = tidy.parseDOM(resource.getInputStream(), null);


                return new DOMReader().read(dom);
            } catch (Exception e) {
            }
        }

View Full Code Here

        }


        String commentCleaned = decodeJavadocTags( description );


        // Using jTidy to clean comment
        Tidy tidy = new Tidy();
        tidy.setDocType( "loose" );
        tidy.setXHTML( true );
        tidy.setXmlOut( true );
        tidy.setInputEncoding( "UTF-8" );
        tidy.setOutputEncoding( "UTF-8" );
        tidy.setMakeClean( true );
        tidy.setNumEntities( true );
        tidy.setQuoteNbsp( false );
        tidy.setQuiet( true );
        tidy.setShowWarnings( false );
        try
        {
            ByteArrayOutputStream out = new ByteArrayOutputStream( commentCleaned.length() + 256 );
            tidy.parse( new ByteArrayInputStream( commentCleaned.getBytes( "UTF-8" ) ), out );
            commentCleaned = out.toString( "UTF-8" );
        }
        catch ( UnsupportedEncodingException e )
        {
            // cannot happen as every JVM must support UTF-8, see also class javadoc for java.nio.charset.Charset

View Full Code Here


        // there is also a javax.swing.text.Document class.
        org.w3c.dom.Document document = null;


        StringWriter sw = new StringWriter();
        Tidy tidy = XPathUtil.makeTidyParser(true, true, true, sw);
        document = tidy.parseDOM(baIS, null);
        document.normalize();
        if (tidy.getParseErrors() > 0) {
            showErrorMessageDialog(sw.toString(),
                    "Tidy: " + tidy.getParseErrors() + " errors, " + tidy.getParseWarnings() + " warnings",
                    JOptionPane.WARNING_MESSAGE);
        }


        JPanel domTreePanel = new DOMTreePanel(document);
        view = domTreePanel;

View Full Code Here

     * @throws TidyException if a ParseError is detected and report_errors is true
     */
    private static Document tidyDoc(InputStream stream, boolean quiet, boolean showWarnings, boolean report_errors,
            boolean isXML) throws TidyException {
        StringWriter sw = new StringWriter();
        Tidy tidy = makeTidyParser(quiet, showWarnings, isXML, sw);
        Document doc = tidy.parseDOM(stream, null);
        doc.normalize();
        if (tidy.getParseErrors() > 0) {
            if (report_errors) {
                log.error("TidyException: " + sw.toString());
                throw new TidyException(tidy.getParseErrors(),tidy.getParseWarnings());
            }
            log.warn("Tidy errors: " + sw.toString());
        }
        return doc;
    }

View Full Code Here

     * @param isXml - treat the content as XML?
     * @param stringWriter - if non-null, use this for Tidy errorOutput
     * @return the Tidy parser
     */
    public static Tidy makeTidyParser(boolean quiet, boolean showWarnings, boolean isXml, StringWriter stringWriter) {
        Tidy tidy = new Tidy();
        tidy.setCharEncoding(org.w3c.tidy.Configuration.UTF8);
        tidy.setQuiet(quiet);
        tidy.setShowWarnings(showWarnings);
        tidy.setMakeClean(true);
        tidy.setXmlTags(isXml);
        if (stringWriter != null) {
            tidy.setErrout(new PrintWriter(stringWriter));
        }
        return tidy;
    }

View Full Code Here

        }


        result.setFailure(false);


        // create parser
        Tidy tidy = null;
        try {
            log.debug("HTMLAssertions.getResult(): Setup tidy ...");
            log.debug("doctype: " + getDoctype());
            log.debug("errors only: " + isErrorsOnly());
            log.debug("error threshold: " + getErrorThreshold());
            log.debug("warning threshold: " + getWarningThreshold());
            log.debug("html mode: " + isHTML());
            log.debug("xhtml mode: " + isXHTML());
            log.debug("xml mode: " + isXML());
            tidy = new Tidy();
            tidy.setCharEncoding(org.w3c.tidy.Configuration.UTF8);
            tidy.setQuiet(false);
            tidy.setShowWarnings(true);
            tidy.setOnlyErrors(isErrorsOnly());
            tidy.setDocType(getDoctype());
            if (isXHTML()) {
                tidy.setXHTML(true);
            } else if (isXML()) {
                tidy.setXmlTags(true);
            }
            log.debug("err file: " + getFilename());
            tidy.setErrfile(getFilename());


            if (log.isDebugEnabled()) {
                log.debug("getParser : tidy parser created - " + tidy);
            }
            log.debug("HTMLAssertions.getResult(): Tidy instance created!");


        } catch (Exception e) {//TODO replace with proper Exception
            log.error("Unable to instantiate tidy parser", e);
            result.setFailure(true);
            result.setFailureMessage("Unable to instantiate tidy parser");
            // return with an error
            return result;
        }


        /*
         * Run tidy.
         */
        try {
            log.debug("HTMLAssertions.getResult(): start parsing with tidy ...");


            StringWriter errbuf = new StringWriter();
            tidy.setErrout(new PrintWriter(errbuf));
            // Node node = tidy.parseDOM(new
            // ByteArrayInputStream(response.getResponseData()), null);
            ByteArrayOutputStream os = new ByteArrayOutputStream();
            log.debug("Start : parse");
            Node node = tidy.parse(new ByteArrayInputStream(inResponse.getResponseData()), os);
            if (log.isDebugEnabled()) {
                log.debug("node : " + node);
            }
            log.debug("End   : parse");
            log.debug("HTMLAssertions.getResult(): parsing with tidy done!");
            log.debug("Output: " + os.toString());


            // write output to file
            writeOutput(errbuf.toString());


            // evaluate result
            if ((tidy.getParseErrors() > getErrorThreshold())
                    || (!isErrorsOnly() && (tidy.getParseWarnings() > getWarningThreshold()))) {
                log.debug("HTMLAssertions.getResult(): errors/warnings detected:");
                log.debug(errbuf.toString());
                result.setFailure(true);
                result.setFailureMessage(MessageFormat.format("Tidy Parser errors:   " + tidy.getParseErrors()
                        + " (allowed " + getErrorThreshold() + ") " + "Tidy Parser warnings: "
                        + tidy.getParseWarnings() + " (allowed " + getWarningThreshold() + ")", new Object[0]));
                // return with an error


            } else if ((tidy.getParseErrors() > 0) || (tidy.getParseWarnings() > 0)) {
                // return with no error
                log.debug("HTMLAssertions.getResult(): there were errors/warnings but threshold to high");
                result.setFailure(false);
            } else {
                // return with no error

View Full Code Here

     *
     * @return a <code>tidy</code> HTML parser
     */
    public static Tidy getParser() {
        log.debug("Start : getParser1");
        Tidy tidy = new Tidy();
        tidy.setCharEncoding(org.w3c.tidy.Configuration.UTF8);
        tidy.setQuiet(true);
        tidy.setShowWarnings(false);


        if (log.isDebugEnabled()) {
            log.debug("getParser1 : tidy parser created - " + tidy);
        }

View Full Code Here

0 1 2 3 4 5 6 7 8 9

TOP

Related Classes of org.w3c.tidy.Tidy

ag.kcmedia.JavaDocViewer

ariba.ui.aribaweb.util.AWUtil

br.com.caelum.tubaina.util.XHTMLValidator

ch.entwine.weblounge.cache.impl.filter.PrettyPrinter

ch.entwine.weblounge.cache.impl.filter.TidyReport$TidyThread

ch.entwine.weblounge.common.impl.util.TestUtils

com.alibaba.antx.config.resource.util.TextBasedPageParser

com.betfair.testing.utils.cougar.helpers.HttpHelpers

com.betfair.testing.utils.cougar.misc.DocumentHelpers

com.cardence.lawshelf.handler.UscParserHandler

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.