/** Creates a Workbook from the given NPOIFSFileSystem, which may be password protected */
private static Workbook create(NPOIFSFileSystem fs, String password)
throws IOException, InvalidFormatException {
DirectoryNode root = fs.getRoot();
// Encrypted OOXML files go inside OLE2 containers, is this one?
if (root.hasEntry(Decryptor.DEFAULT_POIFS_ENTRY)) {
EncryptionInfo info = new EncryptionInfo(fs);
Decryptor d = Decryptor.getInstance(info);
boolean passwordCorrect = false;
InputStream stream = null;
try {
if (password != null && d.verifyPassword(password)) {
passwordCorrect = true;
}
if (!passwordCorrect && d.verifyPassword(Decryptor.DEFAULT_PASSWORD)) {
passwordCorrect = true;
}
if (passwordCorrect) {
stream = d.getDataStream(root);
}
} catch (GeneralSecurityException e) {
throw new IOException(e);
}
if (!passwordCorrect) {
if (password != null) throw new EncryptedDocumentException("Password incorrect");
else
throw new EncryptedDocumentException(
"The supplied spreadsheet is protected, but no password was supplied");
}
OPCPackage pkg = OPCPackage.open(stream);
return create(pkg);
}
// If we get here, it isn't an encrypted XLSX file
// So, treat it as a regular HSSF XLS one
if (password != null) {
Biff8EncryptionKey.setCurrentUserPassword(password);
}
Workbook wb = new HSSFWorkbook(root, true);
Biff8EncryptionKey.setCurrentUserPassword(null);
return wb;
}
/**
* Processes a file into essentially record events.
*
* @param req an Instance of HSSFRequest which has your registered listeners
* @param dir a DirectoryNode containing your workbook
*/
public void processWorkbookEvents(HSSFRequest req, DirectoryNode dir) throws IOException {
// some old documents have "WORKBOOK" or "BOOK"
final String name;
Set<String> entryNames = dir.getEntryNames();
if (entryNames.contains("Workbook")) {
name = "Workbook";
} else if (entryNames.contains("WORKBOOK")) {
name = "WORKBOOK";
} else if (entryNames.contains("BOOK")) {
name = "BOOK";
} else {
name = "Workbook";
}
InputStream in = dir.createDocumentInputStream(name);
processEvents(req, in);
}
/**
* This constructor loads a Word document from a specific point in a POIFSFileSystem, probably not
* the default. Used typically to open embeded documents.
*
* @param directory The DirectoryNode that contains the Word document.
* @throws IOException If there is an unexpected IOException from the passed in POIFSFileSystem.
*/
public HWPFDocumentCore(DirectoryNode directory) throws IOException {
// Sort out the hpsf properties
super(directory);
// read in the main stream.
DocumentEntry documentProps = (DocumentEntry) directory.getEntry("WordDocument");
_mainStream = new byte[documentProps.getSize()];
directory.createDocumentInputStream(STREAM_WORD_DOCUMENT).read(_mainStream);
// Create our FIB, and check for the doc being encrypted
_fib = new FileInformationBlock(_mainStream);
DirectoryEntry objectPoolEntry;
try {
objectPoolEntry = (DirectoryEntry) directory.getEntry(STREAM_OBJECT_POOL);
} catch (FileNotFoundException exc) {
objectPoolEntry = null;
}
_objectPool = new ObjectPoolImpl(objectPoolEntry);
}
/**
* Add a embedded object to this presentation
*
* @return 0-based index of the embedded object
*/
public int addEmbed(POIFSFileSystem poiData) {
DirectoryNode root = poiData.getRoot();
// prepare embedded data
if (new ClassID().equals(root.getStorageClsid())) {
// need to set class id
Map<String, ClassID> olemap = getOleMap();
ClassID classID = null;
for (Map.Entry<String, ClassID> entry : olemap.entrySet()) {
if (root.hasEntry(entry.getKey())) {
classID = entry.getValue();
break;
}
}
if (classID == null) {
throw new IllegalArgumentException("Unsupported embedded document");
}
root.setStorageClsid(classID);
}
ExEmbed exEmbed = new ExEmbed();
// remove unneccessary infos, so we don't need to specify the type
// of the ole object multiple times
Record children[] = exEmbed.getChildRecords();
exEmbed.removeChild(children[2]);
exEmbed.removeChild(children[3]);
exEmbed.removeChild(children[4]);
ExEmbedAtom eeEmbed = exEmbed.getExEmbedAtom();
eeEmbed.setCantLockServerB(true);
ExOleObjAtom eeAtom = exEmbed.getExOleObjAtom();
eeAtom.setDrawAspect(ExOleObjAtom.DRAW_ASPECT_VISIBLE);
eeAtom.setType(ExOleObjAtom.TYPE_EMBEDDED);
// eeAtom.setSubType(ExOleObjAtom.SUBTYPE_EXCEL);
// should be ignored?!?, see MS-PPT ExOleObjAtom, but Libre Office sets it ...
eeAtom.setOptions(1226240);
ExOleObjStg exOleObjStg = new ExOleObjStg();
try {
final String OLESTREAM_NAME = "\u0001Ole";
if (!root.hasEntry(OLESTREAM_NAME)) {
// the following data was taken from an example libre office document
// beside this "\u0001Ole" record there were several other records, e.g. CompObj,
// OlePresXXX, but it seems, that they aren't neccessary
byte oleBytes[] = {1, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
poiData.createDocument(new ByteArrayInputStream(oleBytes), OLESTREAM_NAME);
}
ByteArrayOutputStream bos = new ByteArrayOutputStream();
poiData.writeFilesystem(bos);
exOleObjStg.setData(bos.toByteArray());
} catch (IOException e) {
throw new HSLFException(e);
}
int psrId = addPersistentObject(exOleObjStg);
exOleObjStg.setPersistId(psrId);
eeAtom.setObjStgDataRef(psrId);
int objectId = addToObjListAtom(exEmbed);
eeAtom.setObjID(objectId);
return objectId;
}
/**
* Processes a file into essentially record events.
*
* @param req an Instance of HSSFRequest which has your registered listeners
* @param dir a DirectoryNode containing your workbook
* @return numeric user-specified result code.
*/
public short abortableProcessWorkbookEvents(HSSFRequest req, DirectoryNode dir)
throws IOException, HSSFUserException {
InputStream in = dir.createDocumentInputStream("Workbook");
return abortableProcessEvents(req, in);
}
// will throw IOException if not actually POIFS
// can return null byte[]
private byte[] handleEmbeddedPOIFS(
InputStream is, Metadata metadata, AtomicInteger unknownFilenameCount) throws IOException {
NPOIFSFileSystem fs = null;
byte[] ret = null;
try {
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");
}
DocumentInputStream inp = null;
try {
inp = new DocumentInputStream(contentsEntry);
ret = new byte[contentsEntry.getSize()];
inp.readFully(ret);
} finally {
if (inp != null) {
inp.close();
}
}
} 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());
}
}
} finally {
if (fs != null) {
fs.close();
}
}
return ret;
}
/**
* This constructor loads a Word document from a specific point in a POIFSFileSystem, probably not
* the default. Used typically to open embeded documents.
*
* @param directory The DirectoryNode that contains the Word document.
* @throws IOException If there is an unexpected IOException from the passed in POIFSFileSystem.
*/
public HWPFDocument(DirectoryNode directory) throws IOException {
// Load the main stream and FIB
// Also handles HPSF bits
super(directory);
// Is this document too old for us?
if (_fib.getFibBase().getNFib() < 106) {
throw new OldWordFileFormatException(
"The document is too old - Word 95 or older. Try HWPFOldDocument instead?");
}
// use the fib to determine the name of the table stream.
String name = STREAM_TABLE_0;
if (_fib.getFibBase().isFWhichTblStm()) {
name = STREAM_TABLE_1;
}
// Grab the table stream.
DocumentEntry tableProps;
try {
tableProps = (DocumentEntry) directory.getEntry(name);
} catch (FileNotFoundException fnfe) {
throw new IllegalStateException(
"Table Stream '"
+ name
+ "' wasn't found - Either the document is corrupt, or is Word95 (or earlier)");
}
// read in the table stream.
_tableStream = new byte[tableProps.getSize()];
directory.createDocumentInputStream(name).read(_tableStream);
_fib.fillVariableFields(_mainStream, _tableStream);
// read in the data stream.
try {
DocumentEntry dataProps = (DocumentEntry) directory.getEntry(STREAM_DATA);
_dataStream = new byte[dataProps.getSize()];
directory.createDocumentInputStream(STREAM_DATA).read(_dataStream);
} catch (java.io.FileNotFoundException e) {
_dataStream = new byte[0];
}
// Get the cp of the start of text in the main stream
// The latest spec doc says this is always zero!
int fcMin = 0;
// fcMin = _fib.getFcMin()
// Start to load up our standard structures.
_dop = new DocumentProperties(_tableStream, _fib.getFcDop(), _fib.getLcbDop());
_cft = new ComplexFileTable(_mainStream, _tableStream, _fib.getFcClx(), fcMin);
TextPieceTable _tpt = _cft.getTextPieceTable();
// Now load the rest of the properties, which need to be adjusted
// for where text really begin
_cbt =
new CHPBinTable(
_mainStream, _tableStream, _fib.getFcPlcfbteChpx(), _fib.getLcbPlcfbteChpx(), _tpt);
_pbt =
new PAPBinTable(
_mainStream,
_tableStream,
_dataStream,
_fib.getFcPlcfbtePapx(),
_fib.getLcbPlcfbtePapx(),
_tpt);
_text = _tpt.getText();
/*
* in this mode we preserving PAPX/CHPX structure from file, so text may
* miss from output, and text order may be corrupted
*/
boolean preserveBinTables = false;
try {
preserveBinTables = Boolean.parseBoolean(System.getProperty(PROPERTY_PRESERVE_BIN_TABLES));
} catch (Exception exc) {
// ignore;
}
if (!preserveBinTables) {
_cbt.rebuild(_cft);
_pbt.rebuild(_text, _cft);
}
/*
* Property to disable text rebuilding. In this mode changing the text
* will lead to unpredictable behavior
*/
boolean preserveTextTable = false;
try {
preserveTextTable = Boolean.parseBoolean(System.getProperty(PROPERTY_PRESERVE_TEXT_TABLE));
} catch (Exception exc) {
// ignore;
}
if (!preserveTextTable) {
_cft = new ComplexFileTable();
_tpt = _cft.getTextPieceTable();
final TextPiece textPiece = new SinglentonTextPiece(_text);
_tpt.add(textPiece);
_text = textPiece.getStringBuilder();
}
// Read FSPA and Escher information
// _fspa = new FSPATable(_tableStream, _fib.getFcPlcspaMom(),
// _fib.getLcbPlcspaMom(), getTextTable().getTextPieces());
_fspaHeaders = new FSPATable(_tableStream, _fib, FSPADocumentPart.HEADER);
_fspaMain = new FSPATable(_tableStream, _fib, FSPADocumentPart.MAIN);
if (_fib.getFcDggInfo() != 0) {
_escherRecordHolder =
new EscherRecordHolder(_tableStream, _fib.getFcDggInfo(), _fib.getLcbDggInfo());
} else {
_escherRecordHolder = new EscherRecordHolder();
}
// read in the pictures stream
_pictures = new PicturesTable(this, _dataStream, _mainStream, _fspaMain, _escherRecordHolder);
// And the art shapes stream
_officeArts = new ShapesTable(_tableStream, _fib);
// And escher pictures
_officeDrawingsHeaders = new OfficeDrawingsImpl(_fspaHeaders, _escherRecordHolder, _mainStream);
_officeDrawingsMain = new OfficeDrawingsImpl(_fspaMain, _escherRecordHolder, _mainStream);
_st =
new SectionTable(
_mainStream,
_tableStream,
_fib.getFcPlcfsed(),
_fib.getLcbPlcfsed(),
fcMin,
_tpt,
_fib.getSubdocumentTextStreamLength(SubdocumentType.MAIN));
_ss = new StyleSheet(_tableStream, _fib.getFcStshf());
_ft = new FontTable(_tableStream, _fib.getFcSttbfffn(), _fib.getLcbSttbfffn());
int listOffset = _fib.getFcPlcfLst();
int lfoOffset = _fib.getFcPlfLfo();
if (listOffset != 0 && _fib.getLcbPlcfLst() != 0) {
_lt = new ListTables(_tableStream, _fib.getFcPlcfLst(), _fib.getFcPlfLfo());
}
int sbtOffset = _fib.getFcSttbSavedBy();
int sbtLength = _fib.getLcbSttbSavedBy();
if (sbtOffset != 0 && sbtLength != 0) {
_sbt = new SavedByTable(_tableStream, sbtOffset, sbtLength);
}
int rmarkOffset = _fib.getFcSttbfRMark();
int rmarkLength = _fib.getLcbSttbfRMark();
if (rmarkOffset != 0 && rmarkLength != 0) {
_rmat = new RevisionMarkAuthorTable(_tableStream, rmarkOffset, rmarkLength);
}
_bookmarksTables = new BookmarksTables(_tableStream, _fib);
_bookmarks = new BookmarksImpl(_bookmarksTables);
_endnotesTables = new NotesTables(NoteType.ENDNOTE, _tableStream, _fib);
_endnotes = new NotesImpl(_endnotesTables);
_footnotesTables = new NotesTables(NoteType.FOOTNOTE, _tableStream, _fib);
_footnotes = new NotesImpl(_footnotesTables);
_fieldsTables = new FieldsTables(_tableStream, _fib);
_fields = new FieldsImpl(_fieldsTables);
}