Package org.apache.poi.hwpf

Examples of org.apache.poi.hwpf.HWPFDocument

246
247
248
249
250
251
252
253
254
255
256
    assertTrue(b.toString().contains("TestEndnote"));
  }

  public void testComments() {
    HWPFDocument doc = HWPFTestDataSamples.openSampleFile(filename6);
    extractor = new WordExtractor(doc);

    String[] text = extractor.getCommentsText();
    StringBuffer b = new StringBuffer();
    for (int i = 0; i < text.length; i++) {
View Full Code Here
351
352
353
354
355
356
357
358
359
360
361
          assertEquals(p_text1_block, extractor.getText());
       }

       // Open via a HWPFDocument
       for(DirectoryNode dir : files) {
          HWPFDocument doc = new HWPFDocument(dir);
          WordExtractor extractor = new WordExtractor(doc);
          assertEquals(p_text1_block, extractor.getText());
       }
      
       npoifsFileSystem.close();
View Full Code Here
98
99
100
101
102
103
104
105
106
107
108
    {
        final POIFSFileSystem poifsFileSystem = HWPFDocumentCore
                .verifyAndBuildPOIFS( inputStream );
        try
        {
            return new HWPFDocument( poifsFileSystem );
        }
        catch ( OldWordFileFormatException exc )
        {
            return new HWPFOldDocument( poifsFileSystem );
        }
View Full Code Here
349
350
351
352
353
354
355
356
357
358
359
360
        {
            System.out.println( "Word 95 not supported so far" );
            return;
        }

        HWPFDocument document = (HWPFDocument) _doc;
        Bookmarks bookmarks = document.getBookmarks();
        for ( int b = 0; b < bookmarks.getBookmarksCount(); b++ )
        {
            Bookmark bookmark = bookmarks.getBookmark( b );
            System.out.println( "[" + bookmark.getStart() + "; "
                    + bookmark.getEnd() + "): " + bookmark.getName() );
View Full Code Here
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
    /**
     * Check that we do the positions correctly when working with pure-ascii
     */
    public void testAsciiParts() throws Exception {
        HWPFDocument doc = HWPFTestDataSamples
                .openSampleFile("ThreeColHeadFoot.doc");
        TextPieceTable tbl = doc.getTextTable();

        // All ascii, so stored in one big lump
        assertEquals(1, tbl.getTextPieces().size());
        TextPiece tp = tbl.getTextPieces().get(0);

        assertEquals(0, tp.getStart());
        assertEquals(339, tp.getEnd());
        assertEquals(339, tp.characterLength());
        assertEquals(339, tp.bytesLength());
        assertTrue(tp.getStringBuilder().toString()
                .startsWith("This is a sample word document"));

        // Save and re-load
        HWPFDocument docB = saveAndReload(doc);
        tbl = docB.getTextTable();

        assertEquals(1, tbl.getTextPieces().size());
        tp = tbl.getTextPieces().get(0);

        assertEquals(0, tp.getStart());
View Full Code Here
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
        {
            System.out.println( "Word 95 not supported so far" );
            return;
        }

        HWPFDocument document = (HWPFDocument) _doc;

        for ( FieldsDocumentPart part : FieldsDocumentPart.values() )
        {
            System.out.println( "=== Document part: " + part + " ===" );
            for ( Field field : document.getFields().getFields( part ) )
            {
                System.out.println( field );
            }
        }
    }
View Full Code Here
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
        {
            System.out.println( "Word 95 not supported so far" );
            return;
        }

        HWPFDocument document = (HWPFDocument) _doc;

        if ( document.getOfficeDrawingsHeaders() != null )
        {
            System.out.println( "=== Document part: HEADER ===" );
            for ( OfficeDrawing officeDrawing : document
                    .getOfficeDrawingsHeaders().getOfficeDrawings() )
            {
                System.out.println( officeDrawing );
            }
        }

        if ( document.getOfficeDrawingsHeaders() != null )
        {
            System.out.println( "=== Document part: MAIN ===" );
            for ( OfficeDrawing officeDrawing : document
                    .getOfficeDrawingsMain().getOfficeDrawings() )
            {
                System.out.println( officeDrawing );
            }
        }
View Full Code Here
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
    /**
     * Check that we do the positions correctly when working with a mix ascii,
     * unicode file
     */
    public void testUnicodeParts() throws Exception {
        HWPFDocument doc = HWPFTestDataSamples
                .openSampleFile("HeaderFooterUnicode.doc");
        TextPieceTable tbl = doc.getTextTable();

        // In three bits, split every 512 bytes
        assertEquals(3, tbl.getTextPieces().size());
        TextPiece tpA = tbl.getTextPieces().get(0);
        TextPiece tpB = tbl.getTextPieces().get(1);
        TextPiece tpC = tbl.getTextPieces().get(2);

        assertTrue(tpA.isUnicode());
        assertTrue(tpB.isUnicode());
        assertTrue(tpC.isUnicode());

        assertEquals(256, tpA.characterLength());
        assertEquals(256, tpB.characterLength());
        assertEquals(19, tpC.characterLength());

        assertEquals(512, tpA.bytesLength());
        assertEquals(512, tpB.bytesLength());
        assertEquals(38, tpC.bytesLength());

        assertEquals(0, tpA.getStart());
        assertEquals(256, tpA.getEnd());
        assertEquals(256, tpB.getStart());
        assertEquals(512, tpB.getEnd());
        assertEquals(512, tpC.getStart());
        assertEquals(531, tpC.getEnd());

        // Save and re-load
        HWPFDocument docB = saveAndReload(doc);
        tbl = docB.getTextTable();

        assertEquals(3, tbl.getTextPieces().size());
        tpA = tbl.getTextPieces().get(0);
        tpB = tbl.getTextPieces().get(1);
        tpC = tbl.getTextPieces().get(2);
View Full Code Here
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
    {
        if ( _doc instanceof HWPFDocument )
        {
            System.out.println( "binary PAP pages " );

            HWPFDocument doc = (HWPFDocument) _doc;

            java.lang.reflect.Field fMainStream = HWPFDocumentCore.class
                    .getDeclaredField( "_mainStream" );
            fMainStream.setAccessible( true );
            byte[] mainStream = (byte[]) fMainStream.get( _doc );

            PlexOfCps binTable = new PlexOfCps( doc.getTableStream(), doc
                    .getFileInformationBlock().getFcPlcfbtePapx(), doc
                    .getFileInformationBlock().getLcbPlcfbtePapx(), 4 );

            List<PAPX> papxs = new ArrayList<PAPX>();

            int length = binTable.length();
            for ( int x = 0; x < length; x++ )
            {
                GenericPropertyNode node = binTable.getProperty( x );

                int pageNum = LittleEndian.getInt( node.getBytes() );
                int pageOffset = POIFSConstants.SMALLER_BIG_BLOCK_SIZE
                        * pageNum;

                PAPFormattedDiskPage pfkp = new PAPFormattedDiskPage(
                        mainStream, doc.getDataStream(), pageOffset,
                        doc.getTextTable() );

                System.out.println( "* PFKP: " + pfkp );

                for ( PAPX papx : pfkp.getPAPXs() )
                {
View Full Code Here
159
160
161
162
163
164
165
    protected HWPFDocument saveAndReload(HWPFDocument doc) throws Exception {
        ByteArrayOutputStream baos = new ByteArrayOutputStream();
        doc.write(baos);

        return new HWPFDocument(new ByteArrayInputStream(baos.toByteArray()));
    }
View Full Code Here
TOP

Related Classes of org.apache.poi.hwpf.HWPFDocument

  • org.apache.poi.hslf.examples.DataExtraction
  • org.apache.poi.hslf.extractor.TestExtractor
  • org.apache.poi.hslf.model.TestOleEmbedding
  • org.apache.poi.hssf.usermodel.examples.EmeddedObjects
  • org.apache.poi.hwpf.converter.AbstractWordConverter
  • org.apache.poi.hwpf.converter.AbstractWordUtils
  • org.apache.poi.hwpf.converter.AbstractWordUtilsTest
  • org.apache.poi.hwpf.converter.TestWordToFoConverter
  • org.apache.poi.hwpf.converter.TestWordToHtmlConverter
  • org.apache.poi.hwpf.converter.TestWordToTextConverter
    • 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.