Package org.w3c.tidy

Examples of org.w3c.tidy.Tidy

60
61
62
63
64
65
66
67
68
69
70
71
     *
     * @param file the <code>File</code> containing the HTML to parse
     * @exception IOException if an I/O exception occurs
     */
    public HtmlDocument(File file) throws IOException {
        Tidy tidy = new Tidy();
        tidy.setQuiet(true);
        tidy.setShowWarnings(false);

        org.w3c.dom.Document root = tidy.parseDOM(new FileInputStream(file), null);
        rawDoc = root.getDocumentElement();
    }
View Full Code Here
75
76
77
78
79
80
81
82
83
84
85
86
     *
     * @param is the <code>InputStream</code> containing the HTML
     * @exception IOException if I/O exception occurs
     */
    public HtmlDocument(InputStream is) throws IOException {
        Tidy tidy = new Tidy();
        tidy.setQuiet(true);
        tidy.setShowWarnings(false);

        org.w3c.dom.Document root = tidy.parseDOM(is, null);
        rawDoc = root.getDocumentElement();
    }
View Full Code Here
190
191
192
193
194
195
196
197
198
199
200
201
202
203
     * @return  a <code>tidy</code> HTML parser
     */
    private static Tidy getTidyParser()
    {
        log.debug("Start : getParser");
        Tidy tidy = new Tidy();
        tidy.setCharEncoding(org.w3c.tidy.Configuration.UTF8);
        tidy.setQuiet(true);
        tidy.setShowWarnings(false);
        if(log.isDebugEnabled())
        {
            log.debug("getParser : tidy parser created - " + tidy);
        }
        log.debug("End   : getParser");
View Full Code Here
190
191
192
193
194
195
196
197
198
199
200
201
202
203
     * @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
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
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
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
    }

    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) {
      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
35
36
37
38
39
40
41
42
43
44
    // noop
  }

  public static void prettyPrint(InputStream in, OutputStream writer)
      throws IOException {
    Tidy tidy = getDefaultTidy();

    tidy.parse( in, writer );

  }
View Full Code Here
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
    tidy.parse( in, writer );

  }

  static Tidy getDefaultTidy() throws IOException {
    Tidy tidy = new Tidy();

    // no output please!
    tidy.setErrout( new PrintWriter( new Writer() {
      public void close() throws IOException {
      }

      public void flush() throws IOException {
      }

      public void write(char[] cbuf, int off, int len) throws IOException {
       
      }
    } ) );

    Properties properties = new Properties();

    properties.load( XMLPrettyPrinter.class
        .getResourceAsStream( "jtidy.properties" ) );

    tidy.setConfigurationFromProps( properties );

    return tidy;
  }
View Full Code Here
149
150
151
152
153
154
155
156
      public boolean accept(File dir, String name) {
        return name.endsWith( prefix );
      }
    } );

    Tidy tidy = getDefaultTidy();
    prettyPrintFiles( tidy, files, files, silent );
  }
View Full Code Here
51
52
53
54
55
56
57
58
   protected Properties props;

   public TidyTransformerImpl()
   {
      super();
      tidy = new Tidy();
      initProps();
   }
View Full Code Here
21
22
23
24
25
26
27
28
29
   * Tidy a HTML Stream.
   * @param is
   */
  public static Reader tidy(InputStream is) {
    ByteArrayOutputStream out = new ByteArrayOutputStream();
    Tidy tidy = new Tidy();
    tidy.parse(is, out);
    return new InputStreamReader(new ByteArrayInputStream(out.toByteArray()));
  }
View Full Code Here
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
    • 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.