Search in sources :

Example 1 with POIFSDocumentType

use of org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType in project tika by apache.

the class RTFObjDataParser method handleEmbeddedPOIFS.

//will throw IOException if not actually POIFS
//can return null byte[]
private byte[] handleEmbeddedPOIFS(InputStream is, Metadata metadata, AtomicInteger unknownFilenameCount) throws IOException {
    byte[] ret = null;
    try (NPOIFSFileSystem fs = new NPOIFSFileSystem(is)) {
        DirectoryNode root = fs.getRoot();
        if (root == null) {
            return ret;
        }
        if (root.hasEntry("Package")) {
            Entry ooxml = root.getEntry("Package");
            TikaInputStream stream = TikaInputStream.get(new DocumentInputStream((DocumentEntry) ooxml));
            ByteArrayOutputStream out = new ByteArrayOutputStream();
            IOUtils.copy(stream, out);
            ret = out.toByteArray();
        } else {
            //try poifs
            POIFSDocumentType type = POIFSDocumentType.detectType(root);
            if (type == POIFSDocumentType.OLE10_NATIVE) {
                try {
                    // Try to un-wrap the OLE10Native record:
                    Ole10Native ole = Ole10Native.createFromEmbeddedOleObject(root);
                    ret = ole.getDataBuffer();
                } catch (Ole10NativeException ex) {
                // Not a valid OLE10Native record, skip it
                }
            } else if (type == POIFSDocumentType.COMP_OBJ) {
                DocumentEntry contentsEntry;
                try {
                    contentsEntry = (DocumentEntry) root.getEntry("CONTENTS");
                } catch (FileNotFoundException ioe) {
                    contentsEntry = (DocumentEntry) root.getEntry("Contents");
                }
                try (DocumentInputStream inp = new DocumentInputStream(contentsEntry)) {
                    ret = new byte[contentsEntry.getSize()];
                    inp.readFully(ret);
                }
            } else {
                ByteArrayOutputStream out = new ByteArrayOutputStream();
                is.reset();
                IOUtils.copy(is, out);
                ret = out.toByteArray();
                metadata.set(Metadata.RESOURCE_NAME_KEY, "file_" + unknownFilenameCount.getAndIncrement() + "." + type.getExtension());
                metadata.set(Metadata.CONTENT_TYPE, type.getType().toString());
            }
        }
    }
    return ret;
}
Also used : NPOIFSFileSystem(org.apache.poi.poifs.filesystem.NPOIFSFileSystem) Entry(org.apache.poi.poifs.filesystem.Entry) DocumentEntry(org.apache.poi.poifs.filesystem.DocumentEntry) Ole10NativeException(org.apache.poi.poifs.filesystem.Ole10NativeException) Ole10Native(org.apache.poi.poifs.filesystem.Ole10Native) DocumentEntry(org.apache.poi.poifs.filesystem.DocumentEntry) FileNotFoundException(java.io.FileNotFoundException) TikaInputStream(org.apache.tika.io.TikaInputStream) DirectoryNode(org.apache.poi.poifs.filesystem.DirectoryNode) POIFSDocumentType(org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType) ByteArrayOutputStream(java.io.ByteArrayOutputStream) DocumentInputStream(org.apache.poi.poifs.filesystem.DocumentInputStream)

Example 2 with POIFSDocumentType

use of org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType in project tika by apache.

the class AbstractOOXMLExtractor method handleEmbeddedOLE.

/**
     * Handles an embedded OLE object in the document
     */
private void handleEmbeddedOLE(PackagePart part, ContentHandler handler, String rel, Metadata parentMetadata) throws IOException, SAXException {
    // A POIFSFileSystem needs to be at least 3 blocks big to be valid
    if (part.getSize() >= 0 && part.getSize() < 512 * 3) {
        // Too small, skip
        return;
    }
    InputStream is = part.getInputStream();
    // Open the POIFS (OLE2) structure and process
    POIFSFileSystem fs = null;
    try {
        fs = new POIFSFileSystem(part.getInputStream());
    } catch (Exception e) {
        EmbeddedDocumentUtil.recordEmbeddedStreamException(e, parentMetadata);
        return;
    }
    TikaInputStream stream = null;
    try {
        Metadata metadata = new Metadata();
        metadata.set(Metadata.EMBEDDED_RELATIONSHIP_ID, rel);
        DirectoryNode root = fs.getRoot();
        POIFSDocumentType type = POIFSDocumentType.detectType(root);
        if (root.hasEntry("CONTENTS") && root.hasEntry("Ole") && root.hasEntry("CompObj")) {
            // TIKA-704: OLE 2.0 embedded non-Office document?
            //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
            stream = TikaInputStream.get(fs.createDocumentInputStream("CONTENTS"));
            if (embeddedExtractor.shouldParseEmbedded(metadata)) {
                embeddedExtractor.parseEmbedded(stream, new EmbeddedContentHandler(handler), metadata, false);
            }
        } else if (POIFSDocumentType.OLE10_NATIVE == type) {
            // TIKA-704: OLE 1.0 embedded document
            Ole10Native ole = Ole10Native.createFromEmbeddedOleObject(fs);
            if (ole.getLabel() != null) {
                metadata.set(Metadata.RESOURCE_NAME_KEY, 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();
            if (data != null) {
                stream = TikaInputStream.get(data);
            }
            if (stream != null && embeddedExtractor.shouldParseEmbedded(metadata)) {
                embeddedExtractor.parseEmbedded(stream, new EmbeddedContentHandler(handler), metadata, false);
            }
        } else {
            handleEmbeddedFile(part, handler, rel);
        }
    } catch (FileNotFoundException e) {
    // There was no CONTENTS entry, so skip this part
    } catch (Ole10NativeException e) {
    // Could not process an OLE 1.0 entry, so skip this part
    } catch (IOException e) {
        EmbeddedDocumentUtil.recordEmbeddedStreamException(e, parentMetadata);
    } finally {
        if (fs != null) {
            fs.close();
        }
        if (stream != null) {
            stream.close();
        }
    }
}
Also used : Ole10NativeException(org.apache.poi.poifs.filesystem.Ole10NativeException) Ole10Native(org.apache.poi.poifs.filesystem.Ole10Native) TikaInputStream(org.apache.tika.io.TikaInputStream) InputStream(java.io.InputStream) POIFSFileSystem(org.apache.poi.poifs.filesystem.POIFSFileSystem) NPOIFSFileSystem(org.apache.poi.poifs.filesystem.NPOIFSFileSystem) Metadata(org.apache.tika.metadata.Metadata) FileNotFoundException(java.io.FileNotFoundException) TikaInputStream(org.apache.tika.io.TikaInputStream) DirectoryNode(org.apache.poi.poifs.filesystem.DirectoryNode) POIFSDocumentType(org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType) EmbeddedContentHandler(org.apache.tika.sax.EmbeddedContentHandler) IOException(java.io.IOException) Ole10NativeException(org.apache.poi.poifs.filesystem.Ole10NativeException) TikaException(org.apache.tika.exception.TikaException) InvalidFormatException(org.apache.poi.openxml4j.exceptions.InvalidFormatException) IOException(java.io.IOException) FileNotFoundException(java.io.FileNotFoundException) XmlException(org.apache.xmlbeans.XmlException) SAXException(org.xml.sax.SAXException)

Example 3 with POIFSDocumentType

use of org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType 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();
        }
    }
}
Also used : ZipContainerDetector(org.apache.tika.parser.pkg.ZipContainerDetector) Ole10Native(org.apache.poi.poifs.filesystem.Ole10Native) Metadata(org.apache.tika.metadata.Metadata) FileNotFoundException(java.io.FileNotFoundException) TikaInputStream(org.apache.tika.io.TikaInputStream) POIFSDocumentType(org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType) IOException(java.io.IOException) DocumentInputStream(org.apache.poi.poifs.filesystem.DocumentInputStream) Ole10NativeException(org.apache.poi.poifs.filesystem.Ole10NativeException) TikaException(org.apache.tika.exception.TikaException) IOException(java.io.IOException) FileNotFoundException(java.io.FileNotFoundException) SAXException(org.xml.sax.SAXException) MimeTypeException(org.apache.tika.mime.MimeTypeException) MimeType(org.apache.tika.mime.MimeType) Entry(org.apache.poi.poifs.filesystem.Entry) DocumentEntry(org.apache.poi.poifs.filesystem.DocumentEntry) DirectoryEntry(org.apache.poi.poifs.filesystem.DirectoryEntry) Ole10NativeException(org.apache.poi.poifs.filesystem.Ole10NativeException) MimeTypeException(org.apache.tika.mime.MimeTypeException) DocumentEntry(org.apache.poi.poifs.filesystem.DocumentEntry) MediaType(org.apache.tika.mime.MediaType)

Aggregations

FileNotFoundException (java.io.FileNotFoundException)3 Ole10Native (org.apache.poi.poifs.filesystem.Ole10Native)3 Ole10NativeException (org.apache.poi.poifs.filesystem.Ole10NativeException)3 TikaInputStream (org.apache.tika.io.TikaInputStream)3 POIFSDocumentType (org.apache.tika.parser.microsoft.OfficeParser.POIFSDocumentType)3 IOException (java.io.IOException)2 DirectoryNode (org.apache.poi.poifs.filesystem.DirectoryNode)2 DocumentEntry (org.apache.poi.poifs.filesystem.DocumentEntry)2 DocumentInputStream (org.apache.poi.poifs.filesystem.DocumentInputStream)2 Entry (org.apache.poi.poifs.filesystem.Entry)2 NPOIFSFileSystem (org.apache.poi.poifs.filesystem.NPOIFSFileSystem)2 TikaException (org.apache.tika.exception.TikaException)2 Metadata (org.apache.tika.metadata.Metadata)2 SAXException (org.xml.sax.SAXException)2 ByteArrayOutputStream (java.io.ByteArrayOutputStream)1 InputStream (java.io.InputStream)1 InvalidFormatException (org.apache.poi.openxml4j.exceptions.InvalidFormatException)1 DirectoryEntry (org.apache.poi.poifs.filesystem.DirectoryEntry)1 POIFSFileSystem (org.apache.poi.poifs.filesystem.POIFSFileSystem)1 MediaType (org.apache.tika.mime.MediaType)1