Package it.unimi.dsi.lang

Examples of it.unimi.dsi.lang.MutableString

192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
      if ( ! hasNext() ) return null;
      return next();
    }
   
    public String toString() {
      MutableString result = new MutableString();
      result.append( '{' );
      boolean first = true;
      for( int[] i : interval ) {
        if ( ! first ) result.append( ',' );
        first = false;
        result.append( toString( i ) );
      }
      return result.append( '}' ).toString();
    }
View Full Code Here
229
230
231
232
233
234
235
236
237
238
239
240
241
  public void reset() {
    curr = -1;
  }
 
  public String toString() {
    MutableString result = new MutableString();
    result.append( '[' );
    for( int i = 0; i < document.length; i++ ) {
      if ( i != 0 ) result.append( ", " );
      result.append( '<' ).append( document[ i ] ).append( ':' ).append( new ArrayIntervalIterator( interval[ i ], intervalTerm[ i ] ) ).append( '>' );
    }
    return result.append( ']' ).toString();
  }
View Full Code Here
88
89
90
91
92
93
94
95
96
97
98
      public void close() {}
      public Object content( int field ) throws IOException {
        ensureDocumentIndex( index );
        ObjectArrayList<Anchor> result = new ObjectArrayList<Anchor>();
        for( Map.Entry<Integer, ? extends CharSequence> entry: virtual[ index ].entrySet() )
          result.add( new Anchor( new MutableString( entry.getKey().toString() ), new MutableString( entry.getValue() ) ) );
        return result;
      }
      public CharSequence title() { return null; }
      public CharSequence uri() { return null; }
      public WordReader wordReader( int field ) { return new FastBufferedReader(); }
View Full Code Here
93
94
95
96
97
98
99
100
101
102
103
104
105
106
   *  the second field as body.
   * 
   *  @param document the document.
   *  @return the HTML version of the document.
   */  private String getHTMLDocument( String[] document ) {
    MutableString res = new MutableString();
    res.append( "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Strict//EN\" \"http://www.w3.org/TR/REC-html40/strict.dtd\">\n" );
    res.append( "<HTML>\n<HEAD>\n<TITLE>" + document[ 0 ] + "</TITLE>\n" );
    // Do NOT append the first part of the body
    res.append( "<BODY>\n" + document[ 1 ].substring( document[ 0 ].length() ) );
    res.append( "\n</BODY>\n" );
    res.append( "</HTML>" );
    return res.toString();
  }
View Full Code Here
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
   * 
   *  @param document the document.
   *  @return the HTML version of the document.
   */
  private String getMboxDocument( String[] document ) {
    MutableString res = new MutableString();
    res.append( "From MAILER-DAEMON Fri Apr 15 16:22:32 2005\n" );
    res.append( "Date: 15 Apr 2005 16:22:32 +0200\n" );
    res.append( "From: Mail System Internal Data <MAILER-DAEMON@sliver.usr.dsi.unimi.it>\n" );
    res.append( "Subject: " + document[ 0 ] + "\n" );
    res.append( "Message-ID: <1113574952@sliver.usr.dsi.unimi.it>\n" );
    res.append( "X-IMAP: 1102967122 0000138458\n" );
    res.append( "Return-Path: <matteo.xxx@unimi.it>\n" );
    res.append( "Received: from localhost (localhost.localdomain [127.0.0.1])\n" );
        res.append( "\tby sliver.usr.dsi.unimi.it (8.12.11/8.12.11) with ESMTP id iAUNtadn007305\n");
        res.append( "\tfor <vigna@localhost>; Wed, 1 Dec 2004 00:55:36 +0100\n" );
        res.append( "Received: from law5.usr.dsi.unimi.it [159.149.146.241]\n" );
        res.append( "\tby localhost with IMAP (fetchmail-6.2.5)\n" );
        res.append( "\tfor vigna@localhost (single-drop); Wed, 01 Dec 2004 00:55:36 +0100 (CET)\n" );
        res.append( "To: vigna@dsi.unimi.it\n" );
        res.append( "Message-id: <Pine.WNT.4.33.0412010051240.-209505@p233-mmx>\n" );
        res.append( "Content-type: TEXT/PLAIN; charset=iso-8859-15\n" );
        res.append( "X-Warning: UNAuthenticated Sender\n" );
        res.append( "Content-Transfer-Encoding: 8bit\n" );
        res.append( "Content-Length: " + document[ 1 ].length() + "\n" );
    res.append( "\n" );
    res.append( document[ 1 ] + "\n" );
    return res.toString();
  }
View Full Code Here
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
      }
    }

    // META REFRESH/LOCATION
    if (element == Element.META) {
      final MutableString equiv = attrMap.get(Attribute.HTTP_EQUIV);
      final MutableString content = attrMap.get(Attribute.CONTENT);
      if (equiv != null && content != null) {
        equiv.toLowerCase();

        // http-equiv="refresh" content="0;URL=http://foo.bar/..."
        if (equiv.equals("refresh") && (metaRefresh == null)) {

          final int pos = URLEQUAL_PATTERN.search(content);
          if (pos != -1)
            metaRefresh = content.substring(
                pos + URLEQUAL_PATTERN.length()).toString();
        }

        // http-equiv="location" content="http://foo.bar/..."
        if (equiv.equals("location") && (metaLocation == null))
View Full Code Here
143
144
145
146
147
148
149
150
151
152
153
154
   * @param wordReader the word reader.
   * @param tok the tokenizer.
   * @throws IOException
   */
  private void checkSameWords( WordReader wordReader, StringTokenizer tok ) throws IOException {
    MutableString word = new MutableString();
    MutableString nonWord = new MutableString();
    boolean aWordInDocum, aWordInDocument;
    boolean firstTime = true;
    for (;;) {
      aWordInDocum = wordReader.next( word, nonWord );
      if ( firstTime ) {
View Full Code Here
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
  }
 
  public void testInputStreamSequence() throws IOException, ConfigurationException {
    System.err.println( "Checking input stream (text field only)" );
    // Extract only field number 1, and write it out with separator '\u0000'
    MutableString res = new MutableString();
    String[][] justSecondField = new String[ ndoc ][ 1 ];
    for ( int i = 0; i < ndoc; i++ ) {
      res.append( document[ i ][ 1 ] + "\u0000" );
      justSecondField[ i ][ 0 ] = document[ i ][ 1 ];
    }
    String resString = res.toString();
    // Write the sequence on a file (in UTF-8)
    Writer resWriter = new OutputStreamWriter( new FileOutputStream( new File( tempDir, "stream" ) ), "UTF-8" );
    resWriter.write( resString );
    resWriter.close();
    // Read it as a input stream document sequence
View Full Code Here
355
356
357
358
359
360
361
362
363
364
365
                return new GobyAlignment(this, iterator.next());
            }

            @Override
            public MutableString getId(int targetIndex){
                return new MutableString("chrMock");
            }

        };

        ObjectArrayList<GobyAlignment> visitedEntries = new ObjectArrayList<GobyAlignment>();
View Full Code Here
87
88
89
90
91
92
93
94
95
96
97
            System.err.println(errorMessage);
            throw new UnsupportedOperationException(errorMessage);
        }
        final IndexedIdentifier identifiers = reader.getTargetIdentifiers();
        // add MT as a synonym for M:
        identifiers.put(new MutableString("M"), identifiers.getInt(new MutableString("MT")));
        targetIdentifiers = new DoubleIndexedIdentifier(identifiers);
        isIndexed = reader.isIndexed();
        // reader.close();
        // reader = null;
View Full Code Here
TOP

Related Classes of it.unimi.dsi.lang.MutableString

  • edu.uci.ics.crawler4j.crawler.LinkExtractor
  • it.unimi.dsi.mg4j.document.AbstractDocumentCollection
  • it.unimi.dsi.mg4j.document.AbstractDocumentFactory
  • it.unimi.dsi.mg4j.document.DocumentCollectionTest
  • it.unimi.dsi.mg4j.document.JavamailDocumentCollection
  • it.unimi.dsi.mg4j.document.JdbcDocumentCollection
  • it.unimi.dsi.mg4j.document.MapVirtualDocumentCollection
  • it.unimi.dsi.mg4j.document.SimpleCompressedDocumentCollection
  • it.unimi.dsi.mg4j.document.SimpleCompressedDocumentCollectionBuilder
  • it.unimi.dsi.mg4j.document.ZipDocumentCollection$ZipFactory
    • 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.