Package org.apache.poi.poifs.filesystem

Examples of org.apache.poi.poifs.filesystem.DocumentInputStream

53
54
55
56
57
58
59
60
61
62
63
64
65
    public void processPOIFSReaderEvent(POIFSReaderEvent event) {
      try {
        if (!event.getName().equalsIgnoreCase("PowerPoint Document"))
          return;
        DocumentInputStream input = event.getStream();
        byte[] buffer = new byte[input.available()];
        input.read(buffer, 0, input.available());
        for (int i = 0; i < buffer.length - 20; i++) {
          long type = LittleEndian.getUShort(buffer, i + 2);
          long size = LittleEndian.getUInt(buffer, i + 4);
          if (type == 4008) {
            os.write(buffer, i + 4 + 1, (int) size + 3);
View Full Code Here
74
75
76
77
78
79
80
81
82
83
84
            throws IOException, TikaException {
        try {
            DocumentEntry entry =
                (DocumentEntry) root.getEntry(entryName);
            PropertySet properties =
                new PropertySet(new DocumentInputStream(entry));
            if (properties.isSummaryInformation()) {
                parse(new SummaryInformation(properties));
            }
            if (properties.isDocumentSummaryInformation()) {
                parse(new DocumentSummaryInformation(properties));
View Full Code Here
291
292
293
294
295
296
297
298
299
300
301
                hssfRequest.addListener(formatListener, ExtendedFormatRecord.sid);
                hssfRequest.addListener(formatListener, DrawingGroupRecord.sid);
            }

            // Create event factory and process Workbook (fire events)
            DocumentInputStream documentInputStream = root.createDocumentInputStream(WORKBOOK_ENTRY);
            HSSFEventFactory eventFactory = new HSSFEventFactory();
            try {
                eventFactory.processEvents(hssfRequest, documentInputStream);
            } catch (org.apache.poi.EncryptedDocumentException e) {
                throw new EncryptedDocumentException(e);
View Full Code Here
745
746
747
748
749
750
751
752
753
754
755
756
757
758
                    // Need to recurse
                    DirectoryEntry newDir = destDir.createDirectory(entry.getName());
                    copy((DirectoryEntry) entry, newDir);
                } else {
                    // Copy entry
                    InputStream contents = new DocumentInputStream((DocumentEntry) entry);
                    try {
                        destDir.createDocument(entry.getName(), contents);
                    } finally {
                        contents.close();
                    }
                }
            }
        }
View Full Code Here
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
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
        if (dir.hasEntry("Package")) {
            // It's OOXML (has a ZipFile):
            Entry ooxml = dir.getEntry("Package");

            TikaInputStream stream = TikaInputStream.get(
                    new DocumentInputStream((DocumentEntry) ooxml));
            try {
                ZipContainerDetector detector = new ZipContainerDetector();
                MediaType type = detector.detect(stream, new Metadata());
                handleEmbeddedResource(stream, null, dir.getName(), type.toString(), xhtml, true);
                return;
            } finally {
                stream.close();
            }
        }

        // It's regular OLE2:

        // What kind of document is it?
        Metadata metadata = new Metadata();
        metadata.set(Metadata.EMBEDDED_RELATIONSHIP_ID, dir.getName());
        POIFSDocumentType type = POIFSDocumentType.detectType(dir);
        TikaInputStream embedded = null;

        try {
            if (type == POIFSDocumentType.OLE10_NATIVE) {
                try {
                    // Try to un-wrap the OLE10Native record:
                    Ole10Native ole = Ole10Native.createFromEmbeddedOleObject((DirectoryNode)dir);
                    metadata.set(Metadata.RESOURCE_NAME_KEY, dir.getName() + '/' + ole.getLabel());
                   
                    byte[] data = ole.getDataBuffer();
                    embedded = TikaInputStream.get(data);
                } catch (Ole10NativeException ex) {
                    // Not a valid OLE10Native record, skip it
                }
            } else if (type == POIFSDocumentType.COMP_OBJ) {
                try {
                   // Grab the contents and process
                   DocumentEntry contentsEntry;
                   try {
                     contentsEntry = (DocumentEntry)dir.getEntry("CONTENTS");
                   } catch (FileNotFoundException ioe) {
                     contentsEntry = (DocumentEntry)dir.getEntry("Contents");
                   }
                   DocumentInputStream inp = new DocumentInputStream(contentsEntry);
                   byte[] contents = new byte[contentsEntry.getSize()];
                   inp.readFully(contents);
                   embedded = TikaInputStream.get(contents);
                  
                   // Try to work out what it is
                   MediaType mediaType = getDetector().detect(embedded, new Metadata());
                   String extension = type.getExtension();
View Full Code Here
59
60
61
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
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
    POIFSFileSystem fsys = new POIFSFileSystem(in);

    // load our POIFS document streams.
    DocumentEntry headerProps =
        (DocumentEntry)fsys.getRoot().getEntry("WordDocument");
    DocumentInputStream din = fsys.createDocumentInputStream("WordDocument");
    byte[] header = new byte[headerProps.getSize()];


    din.read(header);
    din.close();

    int info = LittleEndian.getShort(header, 0xa);
    if ((info & 0x4) != 0)
    {
      throw new FastSavedException("Fast-saved files are unsupported at this time");
    }
    if ((info & 0x100) != 0)
    {
      throw new PasswordProtectedException("This document is password protected");
    }

    // determine the version of Word this document came from.
    int nFib = LittleEndian.getShort(header, 0x2);
    switch (nFib)
    {
      case 101:
      case 102:
      case 103:
      case 104:
        // this is a Word 6.0 doc send it to the extractor for that version.
        Word6Extractor oldExtractor = new Word6Extractor();
        return oldExtractor.extractText(header);
    }

    //Get the information we need from the header
    boolean useTable1 = (info & 0x200) != 0;

    //get the location of the piece table
    int complexOffset = LittleEndian.getInt(header, 0x1a2);

    // determine which table stream we must use.
    String tableName = null;
    if (useTable1)
    {
      tableName = "1Table";
    }
    else
    {
      tableName = "0Table";
    }

    DocumentEntry table = (DocumentEntry)fsys.getRoot().getEntry(tableName);
    byte[] tableStream = new byte[table.getSize()];

    din = fsys.createDocumentInputStream(tableName);

    din.read(tableStream);
    din.close();

    int chpOffset = LittleEndian.getInt(header, 0xfa);
    int chpSize = LittleEndian.getInt(header, 0xfe);
    int fcMin = LittleEndian.getInt(header, 0x18);
    CHPBinTable cbt = new CHPBinTable(header, tableStream, chpOffset, chpSize, fcMin);
View Full Code Here
72
73
74
75
76
77
78
79
80
81
82
83
84
      }
      return;
    }

    try {
      final DocumentInputStream dis = event.getStream();
      final byte pptdata[] = new byte[dis.available()];
      dis.read(pptdata, 0, dis.available());
      int offset = 0;
      long offsetPD = 0;

      /*
       * Traverse Bytearray to get CurrentUserEditAtom Call to extract the Text
View Full Code Here
95
96
97
98
99
100
101
102
103
104
105
                    "author='" + documentAuthor + "'" + Layout.LINE_SEP +
                    "keywords='" + documentKeywords + "'");
        }

        // get text
        DocumentInputStream din = null;
        ExcelListener el = new ExcelListener();
        try {
            // proceed to write to file
            // create a new file input stream with the input file specified
            // at the command line
View Full Code Here
258
259
260
261
262
263
264
265
266
267
268
                        POIFSReaderListener listener =
                            ( POIFSReaderListener ) listeners.next();

                        listener.processPOIFSReaderEvent(
                            new POIFSReaderEvent(
                                new DocumentInputStream(document), path,
                                name));
                    }
                }
                else
                {
View Full Code Here
117
118
119
120
121
122
123
124
125
126
127
  /**
   * For a given named property entry, either return it or null if
   *  if it wasn't found
   */
  protected PropertySet getPropertySet(String setName) {
    DocumentInputStream dis;
    try {
      // Find the entry, and get an input stream for it
      dis = directory.createDocumentInputStream(setName);
    } catch(IOException ie) {
      // Oh well, doesn't exist
View Full Code Here
TOP

Related Classes of org.apache.poi.poifs.filesystem.DocumentInputStream

  • com.gentics.cr.lucene.indexer.transformer.ppt.MSPowerPointParser
  • com.stimulus.archiva.extraction.PowerpointExtractor
  • com.stimulus.archiva.persistence.textextraction.PowerpointExtractor
  • net.sf.mpxj.mpp.EncryptedDocumentInputStream
  • net.sf.mpxj.mpp.MPP12Reader
  • net.sf.mpxj.mpp.MPP14Reader
  • net.sf.mpxj.mpp.MPP8Reader
  • net.sf.mpxj.mpp.MPP9Reader
  • net.sf.mpxj.mpp.MPPReader
  • net.sf.mpxj.mpp.SummaryInformation
    • 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.