Examples with DocumentEntry org.apache.poi.poifs.filesystem.DocumentEntry used on opensource projects

Example 11 with DocumentEntry

use of org.apache.poi.poifs.filesystem.DocumentEntry in project poi by apache.

the class HSLFSlideShowImpl method readPictures.

/**
     * Find and read in pictures contained in this presentation.
     * This is lazily called as and when we want to touch pictures.
     */
private void readPictures() throws IOException {
    _pictures = new ArrayList<HSLFPictureData>();
    // if the presentation doesn't contain pictures - will use a null set instead
    if (!getDirectory().hasEntry("Pictures")) {
        return;
    }
    DocumentEntry entry = (DocumentEntry) getDirectory().getEntry("Pictures");
    DocumentInputStream is = getDirectory().createDocumentInputStream(entry);
    byte[] pictstream = IOUtils.toByteArray(is, entry.getSize());
    is.close();
    HSLFSlideShowEncrypted decryptData = new HSLFSlideShowEncrypted(getDocumentEncryptionAtom());
    try {
        int pos = 0;
        // An empty picture record (length 0) will take up 8 bytes
        while (pos <= (pictstream.length - 8)) {
            int offset = pos;
            decryptData.decryptPicture(pictstream, offset);
            // Image signature
            int signature = LittleEndian.getUShort(pictstream, pos);
            pos += LittleEndianConsts.SHORT_SIZE;
            // Image type + 0xF018
            int type = LittleEndian.getUShort(pictstream, pos);
            pos += LittleEndianConsts.SHORT_SIZE;
            // Image size (excluding the 8 byte header)
            int imgsize = LittleEndian.getInt(pictstream, pos);
            pos += LittleEndianConsts.INT_SIZE;
            //  should terminate if the type isn't 0xf007 or 0xf018->0xf117
            if (!((type == 0xf007) || (type >= 0xf018 && type <= 0xf117))) {
                break;
            }
            //  time, so we won't get stuck)
            if (imgsize < 0) {
                throw new CorruptPowerPointFileException("The file contains a picture, at position " + _pictures.size() + ", which has a negatively sized data length, so we can't trust any of the picture data");
            }
            // If they type (including the bonus 0xF018) is 0, skip it
            PictureType pt = PictureType.forNativeID(type - 0xF018);
            if (pt == null) {
                logger.log(POILogger.ERROR, "Problem reading picture: Invalid image type 0, on picture with length " + imgsize + ".\nYou document will probably become corrupted if you save it!");
                logger.log(POILogger.ERROR, "" + pos);
            } else {
                //that are not used in any slide -- BUG-60305
                if (pos + imgsize > pictstream.length) {
                    logger.log(POILogger.WARN, "\"Pictures\" stream may have ended early. In some circumstances, this is not a problem; " + "in others, this could indicate a corrupt file");
                    break;
                }
                // Build the PictureData object from the data
                try {
                    HSLFPictureData pict = HSLFPictureData.create(pt);
                    pict.setSignature(signature);
                    // Copy the data, ready to pass to PictureData
                    byte[] imgdata = new byte[imgsize];
                    System.arraycopy(pictstream, pos, imgdata, 0, imgdata.length);
                    pict.setRawData(imgdata);
                    pict.setOffset(offset);
                    pict.setIndex(_pictures.size());
                    _pictures.add(pict);
                } catch (IllegalArgumentException e) {
                    logger.log(POILogger.ERROR, "Problem reading picture: " + e + "\nYou document will probably become corrupted if you save it!");
                }
            }
            pos += imgsize;
        }
    } finally {
        decryptData.close();
    }
}

Example 12 with DocumentEntry

use of org.apache.poi.poifs.filesystem.DocumentEntry in project poi by apache.

the class HWPFDocFixture method setUp.

public void setUp() throws IOException {
    POIFSFileSystem filesystem = new POIFSFileSystem(POIDataSamples.getDocumentInstance().openResourceAsStream(_testFile));
    DocumentEntry documentProps = (DocumentEntry) filesystem.getRoot().getEntry("WordDocument");
    _mainStream = new byte[documentProps.getSize()];
    filesystem.createDocumentInputStream("WordDocument").read(_mainStream);
    // use the fib to determine the name of the table stream.
    _fib = new FileInformationBlock(_mainStream);
    String name = "0Table";
    if (_fib.getFibBase().isFWhichTblStm()) {
        name = "1Table";
    }
    // read in the table stream.
    DocumentEntry tableProps = (DocumentEntry) filesystem.getRoot().getEntry(name);
    _tableStream = new byte[tableProps.getSize()];
    filesystem.createDocumentInputStream(name).read(_tableStream);
    _fib.fillVariableFields(_mainStream, _tableStream);
}

Example 13 with DocumentEntry

use of org.apache.poi.poifs.filesystem.DocumentEntry in project tika by apache.

the class AbstractPOIFSExtractor method handleEmbeddedOfficeDoc.

/**
     * Handle an office document that's embedded at the POIFS level
     */
protected void handleEmbeddedOfficeDoc(DirectoryEntry dir, String resourceName, XHTMLContentHandler xhtml) throws IOException, SAXException, TikaException {
    if (dir.hasEntry("Package")) {
        // It's OOXML (has a ZipFile):
        Entry ooxml = dir.getEntry("Package");
        try (TikaInputStream stream = TikaInputStream.get(new DocumentInputStream((DocumentEntry) ooxml))) {
            ZipContainerDetector detector = new ZipContainerDetector();
            MediaType type = null;
            try {
                //if there's a stream error while detecting...
                type = detector.detect(stream, new Metadata());
            } catch (Exception e) {
                EmbeddedDocumentUtil.recordEmbeddedStreamException(e, parentMetadata);
                return;
            }
            handleEmbeddedResource(stream, null, dir.getName(), dir.getStorageClsid(), type.toString(), xhtml, true);
            return;
        }
    }
    // It's regular OLE2:
    // What kind of document is it?
    Metadata metadata = new Metadata();
    metadata.set(Metadata.EMBEDDED_RELATIONSHIP_ID, dir.getName());
    if (dir.getStorageClsid() != null) {
        metadata.set(Metadata.EMBEDDED_STORAGE_CLASS_ID, dir.getStorageClsid().toString());
    }
    POIFSDocumentType type = POIFSDocumentType.detectType(dir);
    TikaInputStream embedded = null;
    String rName = (resourceName == null) ? dir.getName() : resourceName;
    try {
        if (type == POIFSDocumentType.OLE10_NATIVE) {
            try {
                // Try to un-wrap the OLE10Native record:
                Ole10Native ole = Ole10Native.createFromEmbeddedOleObject((DirectoryNode) dir);
                if (ole.getLabel() != null) {
                    metadata.set(Metadata.RESOURCE_NAME_KEY, rName + '/' + ole.getLabel());
                }
                if (ole.getCommand() != null) {
                    metadata.add(TikaCoreProperties.ORIGINAL_RESOURCE_NAME, ole.getCommand());
                }
                if (ole.getFileName() != null) {
                    metadata.add(TikaCoreProperties.ORIGINAL_RESOURCE_NAME, ole.getFileName());
                }
                byte[] data = ole.getDataBuffer();
                embedded = TikaInputStream.get(data);
            } catch (Ole10NativeException ex) {
            // Not a valid OLE10Native record, skip it
            } catch (Exception e) {
                EmbeddedDocumentUtil.recordEmbeddedStreamException(e, parentMetadata);
                return;
            }
        } else if (type == POIFSDocumentType.COMP_OBJ) {
            try {
                //TODO: figure out if the equivalent of OLE 1.0's
                //getCommand() and getFileName() exist for OLE 2.0 to populate
                //TikaCoreProperties.ORIGINAL_RESOURCE_NAME
                // 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();
                try {
                    MimeType mimeType = getMimeTypes().forName(mediaType.toString());
                    extension = mimeType.getExtension();
                } catch (MimeTypeException mte) {
                // No details on this type are known
                }
                // Record what we can do about it
                metadata.set(Metadata.CONTENT_TYPE, mediaType.getType().toString());
                metadata.set(Metadata.RESOURCE_NAME_KEY, rName + extension);
            } catch (Exception e) {
                EmbeddedDocumentUtil.recordEmbeddedStreamException(e, parentMetadata);
                return;
            }
        } else {
            metadata.set(Metadata.CONTENT_TYPE, type.getType().toString());
            metadata.set(Metadata.RESOURCE_NAME_KEY, rName + '.' + type.getExtension());
        }
        // Should we parse it?
        if (embeddedDocumentUtil.shouldParseEmbedded(metadata)) {
            if (embedded == null) {
                // Make a TikaInputStream that just
                // passes the root directory of the
                // embedded document, and is otherwise
                // empty (byte[0]):
                embedded = TikaInputStream.get(new byte[0]);
                embedded.setOpenContainer(dir);
            }
            embeddedDocumentUtil.parseEmbedded(embedded, xhtml, metadata, true);
        }
    } catch (IOException e) {
        EmbeddedDocumentUtil.recordEmbeddedStreamException(e, metadata);
    } finally {
        if (embedded != null) {
            embedded.close();
        }
    }
}

Example 14 with DocumentEntry

use of org.apache.poi.poifs.filesystem.DocumentEntry in project tika by apache.

the class SummaryExtractor method parseSummaryEntryIfExists.

private void parseSummaryEntryIfExists(DirectoryNode root, String entryName) 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));
        }
    } catch (FileNotFoundException e) {
    // entry does not exist, just skip it
    } catch (NoPropertySetStreamException e) {
    // no property stream, just skip it
    } catch (UnexpectedPropertySetTypeException e) {
        throw new TikaException("Unexpected HPSF document", e);
    } catch (MarkUnsupportedException e) {
        throw new TikaException("Invalid DocumentInputStream", e);
    } catch (Exception e) {
        LOG.warn("Ignoring unexpected exception while parsing summary entry {}", entryName, e);
    }
}