/// <summary>
/// Set the hash key for the specified object.
/// </summary>
public void SetHashKey(PdfObjectID id)
{
#if !NETFX_CORE
//#if !SILVERLIGHT
byte[] objectId = new byte[5];
_md5.Initialize();
// Split the object number and generation
objectId[0] = (byte)id.ObjectNumber;
objectId[1] = (byte)(id.ObjectNumber >> 8);
objectId[2] = (byte)(id.ObjectNumber >> 16);
objectId[3] = (byte)id.GenerationNumber;
objectId[4] = (byte)(id.GenerationNumber >> 8);
_md5.TransformBlock(_encryptionKey, 0, _encryptionKey.Length, _encryptionKey, 0);
_md5.TransformBlock(objectId, 0, objectId.Length, objectId, 0);
if (_document.revision == 1)
{
_md5.TransformFinalBlock(salt, 0, salt.Length);
}
else
{
_md5.TransformFinalBlock(salt, 0, 0);
}
_key = _md5.Hash;
_md5.Initialize();
_keySize = _encryptionKey.Length + 5;
if (_keySize > 16)
{
_keySize = 16;
}
//#endif
#endif
}
/// <summary>
/// Reads the compressed object with the specified index.
/// </summary>
internal void ReadReferences(PdfCrossReferenceTable xrefTable)
{
////// Create parser for stream.
////Parser parser = new Parser(_document, new MemoryStream(Stream.Value));
for (int idx = 0; idx < _header.Length; idx++)
{
int objectNumber = _header[idx][0];
int offset = _header[idx][1];
PdfObjectID objectID = new PdfObjectID(objectNumber);
// HACK: -1 indicates compressed object.
PdfReference iref = new PdfReference(objectID, -1);
////iref.ObjectID = objectID;
////iref.Value = xrefStream;
if (!xrefTable.Contains(iref.ObjectID))
{
xrefTable.Add(iref);
}
else
{
GetType();
}
}
}
/// <summary>
/// Set the hash key for the specified object.
/// </summary>
internal void SetHashKey(PdfObjectID id)
{
#if !NETFX_CORE && !DNC10
//#if !SILVERLIGHT
byte[] objectId = new byte[5];
// Split the object number and generation
objectId[0] = (byte)id.ObjectNumber;
objectId[1] = (byte)(id.ObjectNumber >> 8);
objectId[2] = (byte)(id.ObjectNumber >> 16);
objectId[3] = (byte)id.GenerationNumber;
objectId[4] = (byte)(id.GenerationNumber >> 8);
_md5.Initialize();
_md5.TransformBlock(_encryptionKey, 0, _encryptionKey.Length, _encryptionKey, 0);
_md5.TransformBlock(objectId, 0, objectId.Length, objectId, 0);
if (_document._securitySettings.DocumentSecurityLevel == PdfDocumentSecurityLevel.Encrypted128BitAes)
{
// Additional padding needed for AES encryption
byte[] aesPadding = new byte[] { 0x73, 0x41, 0x6C, 0x54 }; // 'sAlT'
_md5.TransformFinalBlock(aesPadding, 0, aesPadding.Length);
}
else
{
_md5.TransformFinalBlock(objectId, 0, 0);
}
_key = _md5.Hash;
_md5.Initialize();
_keySize = _encryptionKey.Length + 5;
if (_keySize > 16)
{
_keySize = 16;
}
//#endif
#endif
}
// /// <summary>
// /// Reads a real value directly or (optionally) indirectly from the PDF data stream.
// /// </summary>
// double ReadReal(bool canBeIndirect)
// {
// Symbol symbol = this.lexer.ScanNextToken(canBeIndirect);
// if (symbol == Symbol.Real || symbol == Symbol.Integer)
// return this.lexer.TokenToReal;
// else if (symbol == Symbol.R)
// {
// int position = this.lexer.Position;
//// MoveToObject(this.lexer.Token);
// ReadObjectID(null);
// double f = ReadReal();
// ReadSymbol(Symbol.EndObj);
// this.lexer.Position = position;
// return f;
// }
// throw new PdfReaderException(PSSR.UnexpectedToken(this.lexer.Token));
// }
//
// double ReadReal()
// {
// return ReadReal(false);
// }
// /// <summary>
// /// Reads an object from the PDF input stream. If the object has a specialized parser, it it used.
// /// </summary>
// public static PdfObject ReadObject(PdfObject pdfObject, PdfObjectID objectID)
// {
// if (pdfObject == null)
// throw new ArgumentNullException("pdfObject");
// if (pdfObject.Document == null)
// throw new ArgumentException(PSSR.OwningDocumentRequired, "pdfObject");
//
// Type type = pdfObject.GetType();
// PdfParser parser = CreateParser(pdfObject.Document, type);
// return parser.ReadObject(pdfObject, objectID, false);
// }
/// <summary>
/// Reads an object from the PDF input stream using the default parser.
/// </summary>
public static PdfObject ReadObject(PdfDocument owner, PdfObjectID objectID)
{
if (owner == null)
throw new ArgumentNullException("owner");
Parser parser = new Parser(owner);
return parser.ReadObject(null, objectID, false);
}
/// <summary>
/// Initializes a new PdfReference instance from the specified object identifier and file position.
/// </summary>
public PdfReference(PdfObjectID objectID, int position)
{
_objectID = objectID;
_position = position;
#if UNIQUE_IREF && DEBUG
_uid = ++s_counter;
#endif
}
/// <summary>
/// Initializes a new PdfReference instance from the specified object identifier and file position.
/// </summary>
public PdfReference(PdfObjectID objectID, int position)
{
this.objectID = objectID;
this.position = position;
#if UNIQUE_IREF && DEBUG
this.uid = ++PdfReference.counter;
#endif
}
/// <summary>
/// Initializes a new PdfReference instance from the specified object identifier and file position.
/// </summary>
public PdfReference(PdfObjectID objectID, int position)
{
this.objectID = objectID;
this.position = position;
#if UNIQUE_IREF && DEBUG
uid = ++counter;
#endif
}
/// <summary>
/// Gets a cross reference entry from an object identifier.
/// Returns null if no object with the specified ID exists in the object table.
/// </summary>
public PdfReference this[PdfObjectID objectID]
{
get
{
PdfReference iref;
ObjectTable.TryGetValue(objectID, out iref);
return(iref);
}
}
/// <summary>
///
/// </summary>
PdfTrailer ReadXRefTableAndTrailer(PdfReferenceTable xrefTable)
{
Debug.Assert(xrefTable != null);
Symbol symbol = ScanNextToken();
// Is it an xref stream?
if (symbol == Symbol.Integer)
throw new PdfReaderException(PSSR.CannotHandleXRefStreams);
// TODO: It is very high on the todo list, but still undone
Debug.Assert(symbol == Symbol.XRef);
while (true)
{
symbol = ScanNextToken();
if (symbol == Symbol.Integer)
{
int start = this.lexer.TokenToInteger;
int length = ReadInteger();
for (int id = start; id < start + length; id++)
{
int position = ReadInteger();
int generation = ReadInteger();
ReadSymbol(Symbol.Keyword);
string token = lexer.Token;
// Skip start entry
if (id == 0)
continue;
// Skip unused entries.
if (token != "n")
continue;
// Even it is restricted, an object can exists in more than one subsection.
// (PDF Reference Implementation Notes 15).
PdfObjectID objectID = new PdfObjectID(id, generation);
// Ignore the latter one
if (xrefTable.Contains(objectID))
continue;
xrefTable.Add(new PdfReference(objectID, position));
}
}
else if (symbol == Symbol.Trailer)
{
ReadSymbol(Symbol.BeginDictionary);
PdfTrailer trailer = new PdfTrailer(this.document);
this.ReadDictionary(trailer, false);
return trailer;
}
else
throw new PdfReaderException(PSSR.UnexpectedToken(this.lexer.Token));
}
}
/// <summary>
/// Set the hash key for the specified object.
/// </summary>
internal void SetHashKey(PdfObjectID id)
{
byte[] objectId = new byte[5];
_md5.Initialize();
// Split the object number and generation
objectId[0] = (byte)id.ObjectNumber;
objectId[1] = (byte)(id.ObjectNumber >> 8);
objectId[2] = (byte)(id.ObjectNumber >> 16);
objectId[3] = (byte)id.GenerationNumber;
objectId[4] = (byte)(id.GenerationNumber >> 8);
_md5.TransformBlock(_encryptionKey, 0, _encryptionKey.Length, _encryptionKey, 0);
_md5.TransformFinalBlock(objectId, 0, objectId.Length);
_key = _md5.Hash;
_md5.Initialize();
_keySize = _encryptionKey.Length + 5;
if (_keySize > 16)
{
_keySize = 16;
}
}
/// <summary>
/// Pdf Reference 1.7, Chapter 7.6.2, Algorithm #1
/// </summary>
/// <param name="id"></param>
public void CreateHashKey(PdfObjectID id)
{
var objectId = new byte[5];
md5.Initialize();
// Split the object number and generation
objectId[0] = (byte)id.ObjectNumber;
objectId[1] = (byte)(id.ObjectNumber >> 8);
objectId[2] = (byte)(id.ObjectNumber >> 16);
objectId[3] = (byte)id.GenerationNumber;
objectId[4] = (byte)(id.GenerationNumber >> 8);
md5.TransformBlock(encryptionKey, 0, encryptionKey.Length, encryptionKey, 0); // ?? incomplete
md5.TransformFinalBlock(objectId, 0, objectId.Length);
key = md5.Hash;
md5.Initialize();
keySize = encryptionKey.Length + 5;
if (keySize > 16)
{
keySize = 16;
}
}
/// <summary>
/// Set the hash key for the specified object.
/// </summary>
internal void SetHashKey(PdfObjectID id)
{
#if !SILVERLIGHT
byte[] objectId = new byte[5];
this.md5.Initialize();
// Split the object number and generation
objectId[0] = (byte)id.ObjectNumber;
objectId[1] = (byte)(id.ObjectNumber >> 8);
objectId[2] = (byte)(id.ObjectNumber >> 16);
objectId[3] = (byte)id.GenerationNumber;
objectId[4] = (byte)(id.GenerationNumber >> 8);
this.md5.TransformBlock(this.encryptionKey, 0, this.encryptionKey.Length, this.encryptionKey, 0);
this.md5.TransformFinalBlock(objectId, 0, objectId.Length);
this.key = this.md5.Hash;
this.md5.Initialize();
this.keySize = this.encryptionKey.Length + 5;
if (this.keySize > 16)
{
this.keySize = 16;
}
#endif
}
/// <summary>
/// Set the hash key for the specified object.
/// </summary>
internal void SetHashKey(PdfObjectID id)
{
stringEncryptor.CreateHashKey(id);
streamEncryptor.CreateHashKey(id);
}
/// <summary>
/// Opens an existing PDF document.
/// </summary>
public static PdfDocument Open(Stream stream, string password, PdfDocumentOpenMode openmode, PdfPasswordProvider passwordProvider)
{
PdfDocument document;
try
{
Lexer lexer = new Lexer(stream);
document = new PdfDocument(lexer);
document._state |= DocumentState.Imported;
document._openMode = openmode;
document._fileSize = stream.Length;
// Get file version.
byte[] header = new byte[1024];
stream.Position = 0;
stream.Read(header, 0, 1024);
document._version = GetPdfFileVersion(header);
if (document._version == 0)
throw new InvalidOperationException(PSSR.InvalidPdf);
document._irefTable.IsUnderConstruction = true;
Parser parser = new Parser(document);
// Read all trailers or cross-reference streams, but no objects.
document._trailer = parser.ReadTrailer();
Debug.Assert(document._irefTable.IsUnderConstruction);
document._irefTable.IsUnderConstruction = false;
// Is document encrypted?
PdfReference xrefEncrypt = document._trailer.Elements[PdfTrailer.Keys.Encrypt] as PdfReference;
if (xrefEncrypt != null)
{
//xrefEncrypt.Value = parser.ReadObject(null, xrefEncrypt.ObjectID, false);
PdfObject encrypt = parser.ReadObject(null, xrefEncrypt.ObjectID, false, false);
encrypt.Reference = xrefEncrypt;
xrefEncrypt.Value = encrypt;
PdfStandardSecurityHandler securityHandler = document.SecurityHandler;
TryAgain:
PasswordValidity validity = securityHandler.ValidatePassword(password);
if (validity == PasswordValidity.Invalid)
{
if (passwordProvider != null)
{
PdfPasswordProviderArgs args = new PdfPasswordProviderArgs();
passwordProvider(args);
if (args.Abort)
return null;
password = args.Password;
goto TryAgain;
}
else
{
if (password == null)
throw new PdfReaderException(PSSR.PasswordRequired);
else
throw new PdfReaderException(PSSR.InvalidPassword);
}
}
else if (validity == PasswordValidity.UserPassword && openmode == PdfDocumentOpenMode.Modify)
{
if (passwordProvider != null)
{
PdfPasswordProviderArgs args = new PdfPasswordProviderArgs();
passwordProvider(args);
if (args.Abort)
return null;
password = args.Password;
goto TryAgain;
}
else
throw new PdfReaderException(PSSR.OwnerPasswordRequired);
}
}
else
{
if (password != null)
{
// Password specified but document is not encrypted.
// ignore
}
}
PdfReference[] irefs2 = document._irefTable.AllReferences;
int count2 = irefs2.Length;
// 3rd: Create iRefs for all compressed objects.
Dictionary<int, object> objectStreams = new Dictionary<int, object>();
for (int idx = 0; idx < count2; idx++)
{
PdfReference iref = irefs2[idx];
PdfCrossReferenceStream xrefStream = iref.Value as PdfCrossReferenceStream;
if (xrefStream != null)
{
for (int idx2 = 0; idx2 < xrefStream.Entries.Count; idx2++)
{
PdfCrossReferenceStream.CrossReferenceStreamEntry item = xrefStream.Entries[idx2];
// Is type xref to compressed object?
if (item.Type == 2)
{
//PdfReference irefNew = parser.ReadCompressedObject(new PdfObjectID((int)item.Field2), (int)item.Field3);
//document._irefTable.Add(irefNew);
int objectNumber = (int)item.Field2;
if (!objectStreams.ContainsKey(objectNumber))
{
objectStreams.Add(objectNumber, null);
PdfObjectID objectID = new PdfObjectID((int)item.Field2);
parser.ReadIRefsFromCompressedObject(objectID);
}
}
}
}
}
// 4th: Read compressed objects.
for (int idx = 0; idx < count2; idx++)
{
PdfReference iref = irefs2[idx];
PdfCrossReferenceStream xrefStream = iref.Value as PdfCrossReferenceStream;
if (xrefStream != null)
{
for (int idx2 = 0; idx2 < xrefStream.Entries.Count; idx2++)
{
PdfCrossReferenceStream.CrossReferenceStreamEntry item = xrefStream.Entries[idx2];
// Is type xref to compressed object?
if (item.Type == 2)
{
PdfReference irefNew = parser.ReadCompressedObject(new PdfObjectID((int)item.Field2),
(int)item.Field3);
Debug.Assert(document._irefTable.Contains(iref.ObjectID));
//document._irefTable.Add(irefNew);
}
}
}
}
PdfReference[] irefs = document._irefTable.AllReferences;
int count = irefs.Length;
// Read all indirect objects.
for (int idx = 0; idx < count; idx++)
{
PdfReference iref = irefs[idx];
if (iref.Value == null)
{
#if DEBUG_
if (iref.ObjectNumber == 1074)
iref.GetType();
#endif
try
{
Debug.Assert(document._irefTable.Contains(iref.ObjectID));
PdfObject pdfObject = parser.ReadObject(null, iref.ObjectID, false, false);
Debug.Assert(pdfObject.Reference == iref);
pdfObject.Reference = iref;
Debug.Assert(pdfObject.Reference.Value != null, "Something went wrong.");
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
// 4STLA rethrow exception to notify caller.
throw;
}
}
else
{
Debug.Assert(document._irefTable.Contains(iref.ObjectID));
//iref.GetType();
}
// Set maximum object number.
document._irefTable._maxObjectNumber = Math.Max(document._irefTable._maxObjectNumber,
iref.ObjectNumber);
}
// Encrypt all objects.
if (xrefEncrypt != null)
{
document.SecurityHandler.EncryptDocument();
}
// Fix references of trailer values and then objects and irefs are consistent.
document._trailer.Finish();
#if DEBUG_
// Some tests...
PdfReference[] reachables = document.xrefTable.TransitiveClosure(document.trailer);
reachables.GetType();
reachables = document.xrefTable.AllXRefs;
document.xrefTable.CheckConsistence();
#endif
if (openmode == PdfDocumentOpenMode.Modify)
{
// Create new or change existing document IDs.
if (document.Internals.SecondDocumentID == "")
document._trailer.CreateNewDocumentIDs();
else
{
byte[] agTemp = Guid.NewGuid().ToByteArray();
document.Internals.SecondDocumentID = PdfEncoders.RawEncoding.GetString(agTemp, 0, agTemp.Length);
}
// Change modification date
document.Info.ModificationDate = DateTime.Now;
// Remove all unreachable objects
int removed = document._irefTable.Compact();
if (removed != 0)
Debug.WriteLine("Number of deleted unreachable objects: " + removed);
// Force flattening of page tree
PdfPages pages = document.Pages;
Debug.Assert(pages != null);
//bool b = document.irefTable.Contains(new PdfObjectID(1108));
//b.GetType();
document._irefTable.CheckConsistence();
document._irefTable.Renumber();
document._irefTable.CheckConsistence();
}
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
throw;
}
return document;
}
/// <summary>
/// Reads the compressed object with the specified index.
/// </summary>
internal void ReadReferences(PdfCrossReferenceTable xrefTable)
{
////// Create parser for stream.
////Parser parser = new Parser(_document, new MemoryStream(Stream.Value));
for (int idx = 0; idx < _header.Length; idx++)
{
int objectNumber = _header[idx][0];
int offset = _header[idx][1];
PdfObjectID objectID = new PdfObjectID(objectNumber);
// HACK: -1 indicates compressed object.
PdfReference iref = new PdfReference(objectID, -1);
////iref.ObjectID = objectID;
////iref.Value = xrefStream;
if (!xrefTable.Contains(iref.ObjectID))
{
xrefTable.Add(iref);
}
else
{
GetType();
}
}
}
/// <summary>
/// Reads the compressed object with the specified index in the object stream
/// of the object with the specified object id.
/// </summary>
internal PdfReference ReadCompressedObject(PdfObjectID objectID, int index)
{
PdfReference iref;
#if true
Debug.Assert(_document._irefTable.ObjectTable.ContainsKey(objectID));
if (!_document._irefTable.ObjectTable.TryGetValue(objectID, out iref))
{
throw new NotImplementedException("This case is not coded or something else went wrong");
}
#else
// We should never come here because the object stream must be a type 1 entry in the xref stream
// and iref was created before.
// Has the specified object already an iref in the object table?
if (!_document._irefTable.ObjectTable.TryGetValue(objectID, out iref))
{
try
{
#if true_
iref = new PdfReference(objectID,);
iref.ObjectID = objectID;
_document._irefTable.Add(os);
#else
PdfDictionary dict = (PdfDictionary)ReadObject(null, objectID, false, false);
PdfObjectStream os = new PdfObjectStream(dict);
iref = new PdfReference(os);
iref.ObjectID = objectID;
_document._irefTable.Add(os);
#endif
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
throw;
}
}
#endif
// Read in object stream object when we come here for the very first time.
if (iref.Value == null)
{
try
{
Debug.Assert(_document._irefTable.Contains(iref.ObjectID));
PdfDictionary pdfObject = (PdfDictionary)ReadObject(null, iref.ObjectID, false, false);
PdfObjectStream objectStream = new PdfObjectStream(pdfObject);
Debug.Assert(objectStream.Reference == iref);
// objectStream.Reference = iref; Superfluous, see Assert in line before.
Debug.Assert(objectStream.Reference.Value != null, "Something went wrong.");
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
throw;
}
}
Debug.Assert(iref.Value != null);
PdfObjectStream objectStreamStream = iref.Value as PdfObjectStream;
if (objectStreamStream == null)
{
Debug.Assert(((PdfDictionary)iref.Value).Elements.GetName("/Type") == "/ObjStm");
objectStreamStream = new PdfObjectStream((PdfDictionary)iref.Value);
Debug.Assert(objectStreamStream.Reference == iref);
// objectStream.Reference = iref; Superfluous, see Assert in line before.
Debug.Assert(objectStreamStream.Reference.Value != null, "Something went wrong.");
}
Debug.Assert(objectStreamStream != null);
//PdfObjectStream objectStreamStream = (PdfObjectStream)iref.Value;
if (objectStreamStream == null)
throw new Exception("Something went wrong here.");
return objectStreamStream.ReadCompressedObject(index);
}
/// <summary>
/// Reads cross reference table(s) and trailer(s).
/// </summary>
private PdfTrailer ReadXRefTableAndTrailer(PdfCrossReferenceTable xrefTable)
{
Debug.Assert(xrefTable != null);
Symbol symbol = ScanNextToken();
if (symbol == Symbol.XRef) // Is it a cross-reference table?
{
// Reference: 3.4.3 Cross-Reference Table / Page 93
while (true)
{
symbol = ScanNextToken();
if (symbol == Symbol.Integer)
{
int start = _lexer.TokenToInteger;
int length = ReadInteger();
for (int id = start; id < start + length; id++)
{
int position = ReadInteger();
int generation = ReadInteger();
ReadSymbol(Symbol.Keyword);
string token = _lexer.Token;
// Skip start entry
if (id == 0)
continue;
// Skip unused entries.
if (token != "n")
continue;
// Even it is restricted, an object can exists in more than one subsection.
// (PDF Reference Implementation Notes 15).
PdfObjectID objectID = new PdfObjectID(id, generation);
// Ignore the latter one.
if (xrefTable.Contains(objectID))
continue;
xrefTable.Add(new PdfReference(objectID, position));
}
}
else if (symbol == Symbol.Trailer)
{
ReadSymbol(Symbol.BeginDictionary);
PdfTrailer trailer = new PdfTrailer(_document);
ReadDictionary(trailer, false);
return trailer;
}
else
ParserDiagnostics.HandleUnexpectedToken(_lexer.Token);
}
}
// ReSharper disable once RedundantIfElseBlock because of code readability.
else if (symbol == Symbol.Integer) // Is it an cross-reference stream?
{
// Reference: 3.4.7 Cross-Reference Streams / Page 93
// TODO: Handle PDF files larger than 2 GiB, see implementation note 21 in Appendix H.
// The parsed integer is the object id of the cross-refernece stream.
return ReadXRefStream(xrefTable);
}
return null;
}
/// <summary>
/// Indicates whether the specified object is already imported.
/// </summary>
public bool Contains(PdfObjectID externalID)
{
return(_externalIDs.ContainsKey(externalID.ToString()));
}
/// <summary>
/// Parses whatever comes until the specified stop symbol is reached.
/// </summary>
void ParseObject(Symbol stop)
{
#if DEBUG_
ParseObjectCounter++;
Debug.WriteLine(ParseObjectCounter.ToString());
if (ParseObjectCounter == 178)
{
GetType();
}
#endif
Symbol symbol;
while ((symbol = ScanNextToken()) != Symbol.Eof)
{
if (symbol == stop)
{
return;
}
switch (symbol)
{
case Symbol.Comment:
// ignore comments
break;
case Symbol.Null:
this.stack.Shift(PdfNull.Value);
break;
case Symbol.Boolean:
this.stack.Shift(new PdfBoolean(this.lexer.TokenToBoolean));
break;
case Symbol.Integer:
this.stack.Shift(new PdfInteger(this.lexer.TokenToInteger));
break;
case Symbol.UInteger:
this.stack.Shift(new PdfUInteger(this.lexer.TokenToUInteger));
break;
case Symbol.Real:
this.stack.Shift(new PdfReal(this.lexer.TokenToReal));
break;
case Symbol.String:
//this.stack.Shift(new PdfString(this.lexer.Token, PdfStringFlags.PDFDocEncoding));
this.stack.Shift(new PdfString(this.lexer.Token, PdfStringFlags.RawEncoding));
break;
case Symbol.UnicodeString:
this.stack.Shift(new PdfString(this.lexer.Token, PdfStringFlags.Unicode));
break;
case Symbol.HexString:
this.stack.Shift(new PdfString(this.lexer.Token, PdfStringFlags.HexLiteral));
break;
case Symbol.UnicodeHexString:
this.stack.Shift(new PdfString(this.lexer.Token, PdfStringFlags.Unicode | PdfStringFlags.HexLiteral));
break;
case Symbol.Name:
this.stack.Shift(new PdfName(this.lexer.Token));
break;
case Symbol.R:
{
Debug.Assert(this.stack.GetItem(-1) is PdfInteger && this.stack.GetItem(-2) is PdfInteger);
PdfObjectID objectID = new PdfObjectID(this.stack.GetInteger(-2), this.stack.GetInteger(-1));
PdfReference iref = this.document.irefTable[objectID];
if (iref == null)
{
// If a document has more than one PdfXRefTable it is possible that the first trailer has
// indirect references to objects whos iref entry is not yet read in.
if (this.document.irefTable.IsUnderConstruction)
{
// XRefTable not complete when trailer is read. Create temporary irefs that are
// removed later in PdfTrailer.FixXRefs.
iref = new PdfReference(objectID, 0);
this.stack.Reduce(iref, 2);
break;
}
// PDF Reference section 3.2.9:
// An indirect reference to an undefined object is not an error;
// it is simply treated as a reference to the null object.
this.stack.Reduce(PdfNull.Value, 2);
// Let's see what null objects are good for...
//Debug.Assert(false, "Null object detected!");
//this.stack.Reduce(PdfNull.Value, 2);
}
else
{
this.stack.Reduce(iref, 2);
}
break;
}
case Symbol.BeginArray:
PdfArray array = new PdfArray(this.document);
ReadArray(array, false);
this.stack.Shift(array);
break;
case Symbol.BeginDictionary:
PdfDictionary dict = new PdfDictionary(this.document);
ReadDictionary(dict, false);
this.stack.Shift(dict);
break;
case Symbol.BeginStream:
throw new NotImplementedException();
default:
string error = this.lexer.Token;
Debug.Assert(false, "Unexpected: " + error);
break;
}
}
throw new PdfReaderException("Unexpected end of file.");
}
/// <summary>
/// Sets PDF input stream position to the specified object.
/// </summary>
public int MoveToObject(PdfObjectID objectID)
{
int position = _document._irefTable[objectID].Position;
return _lexer.Position = position;
}
/// <summary>
/// Adds a cloned object to this table.
/// </summary>
/// <param name="externalID">The object identifier in the foreign object.</param>
/// <param name="iref">The cross reference to the clone of the foreign object, which belongs to
/// this document. In general the clone has a different object identifier.</param>
public void Add(PdfObjectID externalID, PdfReference iref)
{
_externalIDs[externalID.ToString()] = iref;
}
/// <summary>
/// Reads the compressed object with the specified number at the given offset.
/// The parser must be initialized with the stream an object stream object.
/// </summary>
internal PdfReference ReadCompressedObject(int objectNumber, int offset)
{
#if DEBUG__
if (objectNumber == 1034)
GetType();
#endif
// Generation is always 0 for compressed objects.
PdfObjectID objectID = new PdfObjectID(objectNumber);
_lexer.Position = offset;
PdfObject obj = ReadObject(null, objectID, false, true);
return obj.Reference;
}
/// <summary>
/// Indicates whether the specified object is already imported.
/// </summary>
public bool Contains(PdfObjectID externalID)
{
return this.externalIDs.Contains(externalID.ToString());
}
/// <summary>
/// Gets the cloned object that corresponds to the specified external identifier.
/// </summary>
public PdfReference this[PdfObjectID externalID]
{
get { return (PdfReference)this.externalIDs[externalID.ToString()]; }
}
/// <summary>
/// Adds a cloned object to this table.
/// </summary>
/// <param name="externalID">The object identifier in the forein object.</param>
/// <param name="iref">The cross reference to the clone of the forein object, which belongs to
/// this document. In general the clone has a different object identifier.</param>
public void Add(PdfObjectID externalID, PdfReference iref)
{
this.externalIDs[externalID.ToString()] = iref;
}
/// <summary> /// Gets the cloned object that corresponds to the specified external identifier. /// </summary> public PdfReference this[PdfObjectID externalID] => (PdfReference)externalIDs[externalID.ToString()];
public PositionNotFoundException(PdfObjectID id) : base($"Object with ID {id} resolved with negative position ")
{
}
/// <summary>
/// Opens an existing PDF document asynchronously.
/// </summary>
public static async Task <PdfDocument> OpenAsync(
Stream stream,
string password = null,
PdfDocumentOpenMode openmode = PdfDocumentOpenMode.Modify,
PdfPasswordProvider passwordProvider = null)
{
PdfDocument document;
#if !DEBUG
try
#endif
{
Lexer lexer = new Lexer(stream);
document = new PdfDocument(lexer);
document._state |= DocumentState.Imported;
document._openMode = openmode;
document._fileSize = stream.Length;
// Get file version.
byte[] header = new byte[1024];
stream.Position = 0;
stream.Read(header, 0, 1024);
document._version = GetPdfFileVersion(header);
if (document._version == 0)
{
throw new InvalidOperationException(PSSR.InvalidPdf);
}
document._irefTable.IsUnderConstruction = true;
Parser parser = new Parser(document);
// Read all trailers or cross-reference streams, but no objects.
document._trailer = await parser.ReadTrailerAsync();
if (document._trailer == null)
{
ParserDiagnostics.ThrowParserException("Invalid PDF file: no trailer found."); // TODO L10N using PSSR.
}
Debug.Assert(document._irefTable.IsUnderConstruction);
document._irefTable.IsUnderConstruction = false;
// Is document encrypted?
PdfReference xrefEncrypt = document._trailer.Elements[PdfTrailer.Keys.Encrypt] as PdfReference;
if (xrefEncrypt != null)
{
PdfObject encrypt = await parser.ReadObjectAsync(null, xrefEncrypt.ObjectID, false, false);
encrypt.Reference = xrefEncrypt;
xrefEncrypt.Value = encrypt;
PdfStandardSecurityHandler securityHandler = document.SecurityHandler;
TryAgain:
PasswordValidity validity = securityHandler.ValidatePassword(password);
if (validity == PasswordValidity.Invalid)
{
if (passwordProvider != null)
{
PdfPasswordProviderArgs args = new PdfPasswordProviderArgs();
passwordProvider(args);
if (args.Abort)
{
return(null);
}
password = args.Password;
goto TryAgain;
}
else
{
if (password == null)
{
throw new PdfReaderException(PSSR.PasswordRequired);
}
else
{
throw new PdfReaderException(PSSR.InvalidPassword);
}
}
}
else if (validity == PasswordValidity.UserPassword && openmode == PdfDocumentOpenMode.Modify)
{
if (passwordProvider != null)
{
PdfPasswordProviderArgs args = new PdfPasswordProviderArgs();
passwordProvider(args);
if (args.Abort)
{
return(null);
}
password = args.Password;
goto TryAgain;
}
else
{
throw new PdfReaderException(PSSR.OwnerPasswordRequired);
}
}
}
else
{
if (password != null)
{
// Password specified but document is not encrypted.
// ignore
}
}
PdfReference[] irefs2 = document._irefTable.AllReferences;
int count2 = irefs2.Length;
// 3rd: Create iRefs for all compressed objects.
Dictionary <int, object> objectStreams = new Dictionary <int, object>();
for (int idx = 0; idx < count2; idx++)
{
PdfReference iref = irefs2[idx];
if (iref.Value is PdfCrossReferenceStream xrefStream)
{
for (int idx2 = 0; idx2 < xrefStream.Entries.Count; idx2++)
{
PdfCrossReferenceStream.CrossReferenceStreamEntry item = xrefStream.Entries[idx2];
// Is type xref to compressed object?
if (item.Type == 2)
{
//PdfReference irefNew = parser.ReadCompressedObject(new PdfObjectID((int)item.Field2), (int)item.Field3);
//document._irefTable.Add(irefNew);
int objectNumber = (int)item.Field2;
if (!objectStreams.ContainsKey(objectNumber))
{
objectStreams.Add(objectNumber, null);
PdfObjectID objectID = new PdfObjectID((int)item.Field2);
parser.ReadIRefsFromCompressedObject(objectID);
}
}
}
}
}
// 4th: Read compressed objects.
for (int idx = 0; idx < count2; idx++)
{
PdfReference iref = irefs2[idx];
if (iref.Value is PdfCrossReferenceStream xrefStream)
{
for (int idx2 = 0; idx2 < xrefStream.Entries.Count; idx2++)
{
PdfCrossReferenceStream.CrossReferenceStreamEntry item = xrefStream.Entries[idx2];
// Is type xref to compressed object?
if (item.Type == 2)
{
PdfReference irefNew = parser.ReadCompressedObject(new PdfObjectID((int)item.Field2),
(int)item.Field3);
Debug.Assert(document._irefTable.Contains(iref.ObjectID));
//document._irefTable.Add(irefNew);
}
}
}
}
PdfReference[] irefs = document._irefTable.AllReferences;
int count = irefs.Length;
// Read all indirect objects.
for (int idx = 0; idx < count; idx++)
{
PdfReference iref = irefs[idx];
if (iref.Value == null)
{
#if DEBUG_
if (iref.ObjectNumber == 1074)
{
iref.GetType();
}
#endif
try
{
Debug.Assert(document._irefTable.Contains(iref.ObjectID));
PdfObject pdfObject = await parser.ReadObjectAsync(null, iref.ObjectID, false, false);
Debug.Assert(pdfObject.Reference == iref);
pdfObject.Reference = iref;
Debug.Assert(pdfObject.Reference.Value != null, "Something went wrong.");
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
// 4STLA rethrow exception to notify caller.
throw;
}
}
else
{
Debug.Assert(document._irefTable.Contains(iref.ObjectID));
//iref.GetType();
}
// Set maximum object number.
document._irefTable._maxObjectNumber = Math.Max(document._irefTable._maxObjectNumber,
iref.ObjectNumber);
}
// Decrypt all objects.
if (xrefEncrypt != null)
{
document.SecurityHandler.DecryptDocument();
}
// Fix references of trailer values and then objects and irefs are consistent.
document._trailer.Finish();
#if DEBUG_
// Some tests...
PdfReference[] reachables = document.xrefTable.TransitiveClosure(document.trailer);
reachables.GetType();
reachables = document.xrefTable.AllXRefs;
document.xrefTable.CheckConsistence();
#endif
if (openmode == PdfDocumentOpenMode.Modify)
{
// Create new or change existing document IDs.
if (document.Internals.SecondDocumentID == "")
{
document._trailer.CreateNewDocumentIDs();
}
else
{
byte[] agTemp = Guid.NewGuid().ToByteArray();
document.Internals.SecondDocumentID = PdfEncoders.RawEncoding.GetString(agTemp, 0, agTemp.Length);
}
// Change modification date
document.Info.ModificationDate = DateTime.Now;
// Remove all unreachable objects
int removed = document._irefTable.Compact();
if (removed != 0)
{
Debug.WriteLine("Number of deleted unreachable objects: " + removed);
}
// Force flattening of page tree
PdfPages pages = document.Pages;
Debug.Assert(pages != null);
//bool b = document.irefTable.Contains(new PdfObjectID(1108));
//b.GetType();
document._irefTable.CheckConsistence();
document._irefTable.Renumber();
document._irefTable.CheckConsistence();
}
}
#if !DEBUG
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
throw;
}
#endif
return(document);
}
/// <summary>
/// Gets the cloned object that corresponds to the specified external identifier.
/// </summary>
public PdfReference this[PdfObjectID externalID]
{
get { return(_externalIDs[externalID.ToString()]); }
}
// /// <summary>
// /// Reads a real value directly or (optionally) indirectly from the PDF data stream.
// /// </summary>
// double ReadReal(bool canBeIndirect)
// {
// Symbol symbol = lexer.ScanNextToken(canBeIndirect);
// if (symbol == Symbol.Real || symbol == Symbol.Integer)
// return lexer.TokenToReal;
// else if (symbol == Symbol.R)
// {
// int position = lexer.Position;
//// MoveToObject(lexer.Token);
// ReadObjectID(null);
// double f = ReadReal();
// ReadSymbol(Symbol.EndObj);
// lexer.Position = position;
// return f;
// }
// thr ow new PdfReaderException(PSSR.UnexpectedToken(lexer.Token));
// }
//
// double ReadReal()
// {
// return ReadReal(false);
// }
// /// <summary>
// /// Reads an object from the PDF input stream. If the object has a specialized parser, it it used.
// /// </summary>
// public static PdfObject ReadObject(PdfObject pdfObject, PdfObjectID objectID)
// {
// if (pdfObject == null)
// thr ow new ArgumentNullException("pdfObject");
// if (pdfObject.Document == null)
// th row new ArgumentException(PSSR.OwningDocumentRequired, "pdfObject");
//
// Type type = pdfObject.GetType();
// PdfParser parser = CreateParser(pdfObject.Document, type);
// return parser.ReadObject(pdfObject, objectID, false);
// }
/// <summary>
/// Reads an object from the PDF input stream using the default parser.
/// </summary>
public static PdfObject ReadObject(PdfDocument owner, PdfObjectID objectID)
{
if (owner == null)
throw new ArgumentNullException("owner");
Parser parser = new Parser(owner);
return parser.ReadObject(null, objectID, false, false);
}
/// <summary>
/// Reads PDF object from input stream.
/// </summary>
/// <param name="pdfObject">Either the instance of a derived type or null. If it is null
/// an appropriate object is created.</param>
/// <param name="objectID">The address of the object.</param>
/// <param name="includeReferences">If true, specifies that all indirect objects
/// are included recursively.</param>
public PdfObject ReadObject(PdfObject pdfObject, PdfObjectID objectID, bool includeReferences)
{
MoveToObject(objectID);
int objectNumber = ReadInteger();
int generationNumber = ReadInteger();
#if DEBUG
// The following assertion sometime failed (see below)
//Debug.Assert(objectID == new PdfObjectID(objectNumber, generationNumber));
if (objectID != new PdfObjectID(objectNumber, generationNumber))
{
// A special kind of bug? Or is this an undocumented PDF feature?
// PDF4NET 2.6 provides a sample called 'Unicode', which produces a file 'unicode.pdf'
// The iref table of this file contains the following entries:
// iref
// 0 148
// 0000000000 65535 f
// 0000000015 00000 n
// 0000000346 00000 n
// ....
// 0000083236 00000 n
// 0000083045 00000 n
// 0000083045 00000 n
// 0000083045 00000 n
// 0000083045 00000 n
// 0000080334 00000 n
// ....
// Object 84, 85, 86, and 87 maps to the same dictionary, but all PDF readers I tested
// ignores this mismatch! The following assertion failed about 50 times with this file.
#if true_
string message = String.Format("xref entry {0} {1} maps to object {2} {3}.",
objectID.ObjectNumber, objectID.GenerationNumber, objectNumber, generationNumber);
Debug.Assert(false, message);
#endif
}
#endif
// Always use object ID from iref table (see above)
objectNumber = objectID.ObjectNumber;
generationNumber = objectID.GenerationNumber;
#if true_
Debug.WriteLine(String.Format("obj: {0} {1}", objectNumber, generationNumber));
#endif
ReadSymbol(Symbol.Obj);
bool checkForStream = false;
Symbol symbol = ScanNextToken();
switch (symbol)
{
case Symbol.BeginArray:
PdfArray array;
if (pdfObject == null)
{
array = new PdfArray(this.document);
}
else
{
array = (PdfArray)pdfObject;
}
//PdfObject.RegisterObject(array, objectID, generation);
pdfObject = ReadArray(array, includeReferences);
pdfObject.SetObjectID(objectNumber, generationNumber);
break;
case Symbol.BeginDictionary:
PdfDictionary dict;
if (pdfObject == null)
{
dict = new PdfDictionary(this.document);
}
else
{
dict = (PdfDictionary)pdfObject;
}
//PdfObject.RegisterObject(dict, objectID, generation);
checkForStream = true;
pdfObject = ReadDictionary(dict, includeReferences);
pdfObject.SetObjectID(objectNumber, generationNumber);
break;
// Acrobat 6 Professional proudly presents: The Null object!
// Even with a one-digit object number an indirect reference �x 0 R� to this object is
// one character larger than the direct use of �null�. Probable this is the reason why
// it is true that Acrobat Web Capture 6.0 creates this object, but obviously never
// creates a reference to it!
case Symbol.Null:
pdfObject = new PdfNullObject(this.document);
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.Boolean:
pdfObject = new PdfBooleanObject(this.document, string.Compare(this.lexer.Token, Boolean.TrueString, true) == 0); //!!!mod THHO 19.11.09
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.Integer:
pdfObject = new PdfIntegerObject(this.document, this.lexer.TokenToInteger);
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.UInteger:
pdfObject = new PdfUIntegerObject(this.document, this.lexer.TokenToUInteger);
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.Real:
pdfObject = new PdfRealObject(this.document, this.lexer.TokenToReal);
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.String:
pdfObject = new PdfStringObject(this.document, this.lexer.Token);
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.Name:
pdfObject = new PdfNameObject(this.document, this.lexer.Token);
pdfObject.SetObjectID(objectNumber, generationNumber);
ReadSymbol(Symbol.EndObj);
return(pdfObject);
case Symbol.Keyword:
// Should not come here anymore
throw new NotImplementedException("Keyword");
default:
// Should not come here anymore
throw new NotImplementedException("unknown token \"" + symbol + "\"");
}
symbol = ScanNextToken();
if (symbol == Symbol.BeginStream)
{
PdfDictionary dict = (PdfDictionary)pdfObject;
Debug.Assert(checkForStream, "Unexpected stream...");
int length = GetStreamLength(dict);
byte[] bytes = this.lexer.ReadStream(length);
#if true_
if (dict.Elements.GetString("/Filter") == "/FlateDecode")
{
if (dict.Elements["/Subtype"] == null)
{
try
{
byte[] decoded = Filtering.FlateDecode.Decode(bytes);
if (decoded.Length == 0)
{
goto End;
}
string pageContent = Filtering.FlateDecode.DecodeToString(bytes);
if (pageContent.Length > 100)
{
pageContent = pageContent.Substring(pageContent.Length - 100);
}
pageContent.GetType();
bytes = decoded;
dict.Elements.Remove("/Filter");
dict.Elements.SetInteger("/Length", bytes.Length);
}
catch
{
}
}
End :;
}
#endif
PdfDictionary.PdfStream stream = new PdfDictionary.PdfStream(bytes, dict);
dict.Stream = stream;
ReadSymbol(Symbol.EndStream);
symbol = ScanNextToken();
}
if (symbol != Symbol.EndObj)
{
throw new PdfReaderException(PSSR.UnexpectedToken(this.lexer.Token));
}
return(pdfObject);
}
/// <summary>
/// Parses whatever comes until the specified stop symbol is reached.
/// </summary>
private void ParseObject(Symbol stop)
{
#if DEBUG_
ParseObjectCounter++;
Debug.WriteLine(ParseObjectCounter.ToString());
if (ParseObjectCounter == 178)
GetType();
#endif
Symbol symbol;
while ((symbol = ScanNextToken()) != Symbol.Eof)
{
if (symbol == stop)
return;
switch (symbol)
{
case Symbol.Comment:
// ignore comments
break;
case Symbol.Null:
_stack.Shift(PdfNull.Value);
break;
case Symbol.Boolean:
_stack.Shift(new PdfBoolean(_lexer.TokenToBoolean));
break;
case Symbol.Integer:
_stack.Shift(new PdfInteger(_lexer.TokenToInteger));
break;
case Symbol.UInteger:
_stack.Shift(new PdfUInteger(_lexer.TokenToUInteger));
break;
case Symbol.Real:
_stack.Shift(new PdfReal(_lexer.TokenToReal));
break;
case Symbol.String:
//stack.Shift(new PdfString(lexer.Token, PdfStringFlags.PDFDocEncoding));
_stack.Shift(new PdfString(_lexer.Token, PdfStringFlags.RawEncoding));
break;
case Symbol.UnicodeString:
_stack.Shift(new PdfString(_lexer.Token, PdfStringFlags.Unicode));
break;
case Symbol.HexString:
_stack.Shift(new PdfString(_lexer.Token, PdfStringFlags.HexLiteral));
break;
case Symbol.UnicodeHexString:
_stack.Shift(new PdfString(_lexer.Token, PdfStringFlags.Unicode | PdfStringFlags.HexLiteral));
break;
case Symbol.Name:
_stack.Shift(new PdfName(_lexer.Token));
break;
case Symbol.R:
{
Debug.Assert(_stack.GetItem(-1) is PdfInteger && _stack.GetItem(-2) is PdfInteger);
PdfObjectID objectID = new PdfObjectID(_stack.GetInteger(-2), _stack.GetInteger(-1));
PdfReference iref = _document._irefTable[objectID];
if (iref == null)
{
// If a document has more than one PdfXRefTable it is possible that the first trailer has
// indirect references to objects whos iref entry is not yet read in.
if (_document._irefTable.IsUnderConstruction)
{
// XRefTable not complete when trailer is read. Create temporary irefs that are
// removed later in PdfTrailer.FixXRefs.
iref = new PdfReference(objectID, 0);
_stack.Reduce(iref, 2);
break;
}
// PDF Reference section 3.2.9:
// An indirect reference to an undefined object is not an error;
// it is simply treated as a reference to the null object.
_stack.Reduce(PdfNull.Value, 2);
// Let's see what null objects are good for...
//Debug.Assert(false, "Null object detected!");
//stack.Reduce(PdfNull.Value, 2);
}
else
_stack.Reduce(iref, 2);
break;
}
case Symbol.BeginArray:
PdfArray array = new PdfArray(_document);
ReadArray(array, false);
_stack.Shift(array);
break;
case Symbol.BeginDictionary:
PdfDictionary dict = new PdfDictionary(_document);
ReadDictionary(dict, false);
_stack.Shift(dict);
break;
case Symbol.BeginStream:
throw new NotImplementedException();
// Not expected here:
//case Symbol.None:
//case Symbol.Keyword:
//case Symbol.EndStream:
//case Symbol.EndArray:
//case Symbol.EndDictionary:
//case Symbol.Obj:
//case Symbol.EndObj:
//case Symbol.XRef:
//case Symbol.Trailer:
//case Symbol.StartXRef:
//case Symbol.Eof:
default:
ParserDiagnostics.HandleUnexpectedToken(_lexer.Token);
SkipCharsUntil(stop);
return;
}
}
ParserDiagnostics.ThrowParserException("Unexpected end of file.");
}
/// <summary>
/// Sets PDF input stream position to the specified object.
/// </summary>
public int MoveToObject(PdfObjectID objectID)
{
int position = this.document.irefTable[objectID].Position;
return(this.lexer.Position = position);
}
/// <summary>
/// Reads the irefs from the compressed object with the specified index in the object stream
/// of the object with the specified object id.
/// </summary>
internal void ReadIRefsFromCompressedObject(PdfObjectID objectID)
{
PdfReference iref;
Debug.Assert(_document._irefTable.ObjectTable.ContainsKey(objectID));
if (!_document._irefTable.ObjectTable.TryGetValue(objectID, out iref))
{
// We should never come here because the object stream must be a type 1 entry in the xref stream
// and iref was created before.
throw new NotImplementedException("This case is not coded or something else went wrong");
}
// Read in object stream object when we come here for the very first time.
if (iref.Value == null)
{
try
{
Debug.Assert(_document._irefTable.Contains(iref.ObjectID));
PdfDictionary pdfObject = (PdfDictionary)ReadObject(null, iref.ObjectID, false, false);
PdfObjectStream objectStream = new PdfObjectStream(pdfObject);
Debug.Assert(objectStream.Reference == iref);
// objectStream.Reference = iref; Superfluous, see Assert in line before.
Debug.Assert(objectStream.Reference.Value != null, "Something went wrong.");
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
throw;
}
}
Debug.Assert(iref.Value != null);
PdfObjectStream objectStreamStream = iref.Value as PdfObjectStream;
if (objectStreamStream == null)
{
Debug.Assert(((PdfDictionary)iref.Value).Elements.GetName("/Type") == "/ObjStm");
objectStreamStream = new PdfObjectStream((PdfDictionary)iref.Value);
Debug.Assert(objectStreamStream.Reference == iref);
// objectStream.Reference = iref; Superfluous, see Assert in line before.
Debug.Assert(objectStreamStream.Reference.Value != null, "Something went wrong.");
}
Debug.Assert(objectStreamStream != null);
//PdfObjectStream objectStreamStream = (PdfObjectStream)iref.Value;
if (objectStreamStream == null)
throw new Exception("Something went wrong here.");
objectStreamStream.ReadReferences(_document._irefTable);
}
/// <summary>
/// Returns the object with the specified Identifier, or null, if no such object exists.
/// </summary>
public PdfObject GetObject(PdfObjectID objectID)
{
return(_document._irefTable[objectID].Value);
}
/// <summary>
/// Indicates whether the specified object identifier is in the table.
/// </summary>
public bool Contains(PdfObjectID objectID)
{
return ObjectTable.ContainsKey(objectID);
}
/// <summary>
/// Reads PDF object from input stream.
/// </summary>
/// <param name="pdfObject">Either the instance of a derived type or null. If it is null
/// an appropriate object is created.</param>
/// <param name="objectID">The address of the object.</param>
/// <param name="includeReferences">If true, specifies that all indirect objects
/// are included recursively.</param>
/// <param name="fromObjecStream">If true, the objects is parsed from an object stream.</param>
public PdfObject ReadObject(PdfObject pdfObject, PdfObjectID objectID, bool includeReferences, bool fromObjecStream)
{
#if DEBUG_
Debug.WriteLine("ReadObject: " + objectID);
if (objectID.ObjectNumber == 20)
GetType();
#endif
int objectNumber = objectID.ObjectNumber;
int generationNumber = objectID.GenerationNumber;
if (!fromObjecStream)
{
MoveToObject(objectID);
objectNumber = ReadInteger();
generationNumber = ReadInteger();
}
#if DEBUG
// The following assertion sometime failed (see below)
//Debug.Assert(objectID == new PdfObjectID(objectNumber, generationNumber));
if (!fromObjecStream && objectID != new PdfObjectID(objectNumber, generationNumber))
{
// A special kind of bug? Or is this an undocumented PDF feature?
// PDF4NET 2.6 provides a sample called 'Unicode', which produces a file 'unicode.pdf'
// The iref table of this file contains the following entries:
// iref
// 0 148
// 0000000000 65535 f
// 0000000015 00000 n
// 0000000346 00000 n
// ....
// 0000083236 00000 n
// 0000083045 00000 n
// 0000083045 00000 n
// 0000083045 00000 n
// 0000083045 00000 n
// 0000080334 00000 n
// ....
// Object 84, 85, 86, and 87 maps to the same dictionary, but all PDF readers I tested
// ignores this mismatch! The following assertion failed about 50 times with this file.
#if true_
string message = String.Format("xref entry {0} {1} maps to object {2} {3}.",
objectID.ObjectNumber, objectID.GenerationNumber, objectNumber, generationNumber);
Debug.Assert(false, message);
#endif
}
#endif
// Always use object ID from iref table (see above).
objectNumber = objectID.ObjectNumber;
generationNumber = objectID.GenerationNumber;
#if true_
Debug.WriteLine(String.Format("obj: {0} {1}", objectNumber, generationNumber));
#endif
if (!fromObjecStream)
ReadSymbol(Symbol.Obj);
bool checkForStream = false;
Symbol symbol = ScanNextToken();
switch (symbol)
{
case Symbol.BeginArray:
PdfArray array;
if (pdfObject == null)
array = new PdfArray(_document);
else
array = (PdfArray)pdfObject;
//PdfObject.RegisterObject(array, objectID, generation);
pdfObject = ReadArray(array, includeReferences);
pdfObject.SetObjectID(objectNumber, generationNumber);
break;
case Symbol.BeginDictionary:
PdfDictionary dict;
if (pdfObject == null)
dict = new PdfDictionary(_document);
else
dict = (PdfDictionary)pdfObject;
//PdfObject.RegisterObject(dict, objectID, generation);
checkForStream = true;
pdfObject = ReadDictionary(dict, includeReferences);
pdfObject.SetObjectID(objectNumber, generationNumber);
break;
// Acrobat 6 Professional proudly presents: The Null object!
// Even with a one-digit object number an indirect reference «x 0 R» to this object is
// one character larger than the direct use of «null». Probable this is the reason why
// it is true that Acrobat Web Capture 6.0 creates this object, but obviously never
// creates a reference to it!
case Symbol.Null:
pdfObject = new PdfNullObject(_document);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.Boolean:
pdfObject = new PdfBooleanObject(_document, String.Compare(_lexer.Token, Boolean.TrueString, StringComparison.OrdinalIgnoreCase) == 0);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.Integer:
pdfObject = new PdfIntegerObject(_document, _lexer.TokenToInteger);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.UInteger:
pdfObject = new PdfUIntegerObject(_document, _lexer.TokenToUInteger);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.Real:
pdfObject = new PdfRealObject(_document, _lexer.TokenToReal);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.String:
pdfObject = new PdfStringObject(_document, _lexer.Token);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.Name:
pdfObject = new PdfNameObject(_document, _lexer.Token);
pdfObject.SetObjectID(objectNumber, generationNumber);
if (!fromObjecStream)
ReadSymbol(Symbol.EndObj);
return pdfObject;
case Symbol.Keyword:
// Should not come here anymore.
ParserDiagnostics.HandleUnexpectedToken(_lexer.Token);
break;
default:
// Should not come here anymore.
ParserDiagnostics.HandleUnexpectedToken(_lexer.Token);
break;
}
symbol = ScanNextToken();
if (symbol == Symbol.BeginStream)
{
PdfDictionary dict = (PdfDictionary)pdfObject;
Debug.Assert(checkForStream, "Unexpected stream...");
#if true_
ReadStream(dict);
#else
int length = GetStreamLength(dict);
byte[] bytes = _lexer.ReadStream(length);
#if true_
if (dict.Elements.GetString("/Filter") == "/FlateDecode")
{
if (dict.Elements["/Subtype"] == null)
{
try
{
byte[] decoded = Filtering.FlateDecode.Decode(bytes);
if (decoded.Length == 0)
goto End;
string pageContent = Filtering.FlateDecode.DecodeToString(bytes);
if (pageContent.Length > 100)
pageContent = pageContent.Substring(pageContent.Length - 100);
pageContent.GetType();
bytes = decoded;
dict.Elements.Remove("/Filter");
dict.Elements.SetInteger("/Length", bytes.Length);
}
catch
{
}
}
End: ;
}
#endif
PdfDictionary.PdfStream stream = new PdfDictionary.PdfStream(bytes, dict);
dict.Stream = stream;
ReadSymbol(Symbol.EndStream);
symbol = ScanNextToken();
#endif
}
if (!fromObjecStream && symbol != Symbol.EndObj)
ParserDiagnostics.ThrowParserException(PSSR.UnexpectedToken(_lexer.Token));
return pdfObject;
}
/// <summary>
/// Reads cross reference stream(s).
/// </summary>
private PdfTrailer ReadXRefStream(PdfCrossReferenceTable xrefTable)
{
// Read cross reference stream.
//Debug.Assert(_lexer.Symbol == Symbol.Integer);
int number = _lexer.TokenToInteger;
int generation = ReadInteger();
Debug.Assert(generation == 0);
ReadSymbol(Symbol.Obj);
ReadSymbol(Symbol.BeginDictionary);
PdfObjectID objectID = new PdfObjectID(number, generation);
PdfCrossReferenceStream xrefStream = new PdfCrossReferenceStream(_document);
ReadDictionary(xrefStream, false);
ReadSymbol(Symbol.BeginStream);
ReadStream(xrefStream);
//xrefTable.Add(new PdfReference(objectID, position));
PdfReference iref = new PdfReference(xrefStream);
iref.ObjectID = objectID;
iref.Value = xrefStream;
xrefTable.Add(iref);
Debug.Assert(xrefStream.Stream != null);
//string sValue = new RawEncoding().GetString(xrefStream.Stream.UnfilteredValue,);
//sValue.GetType();
byte[] bytesRaw = xrefStream.Stream.UnfilteredValue;
byte[] bytes = bytesRaw;
// HACK: Should be done in UnfilteredValue.
if (xrefStream.Stream.HasDecodeParams)
{
int predictor = xrefStream.Stream.DecodePredictor;
int columns = xrefStream.Stream.DecodeColumns;
bytes = DecodeCrossReferenceStream(bytesRaw, columns, predictor);
}
#if DEBUG_
for (int idx = 0; idx < bytes.Length; idx++)
{
if (idx % 4 == 0)
Console.WriteLine();
Console.Write("{0:000} ", (int)bytes[idx]);
}
Console.WriteLine();
#endif
// bytes.GetType();
// Add to table.
// xrefTable.Add(new PdfReference(objectID, -1));
int size = xrefStream.Elements.GetInteger(PdfCrossReferenceStream.Keys.Size);
PdfArray index = xrefStream.Elements.GetValue(PdfCrossReferenceStream.Keys.Index) as PdfArray;
int prev = xrefStream.Elements.GetInteger(PdfCrossReferenceStream.Keys.Prev);
PdfArray w = (PdfArray)xrefStream.Elements.GetValue(PdfCrossReferenceStream.Keys.W);
// E.g.: W[1 2 1] ¤ Index[7 12] ¤ Size 19
// Setup subsections.
int subsectionCount;
int[][] subsections = null;
int subsectionEntryCount = 0;
if (index == null)
{
// Setup with default values.
subsectionCount = 1;
subsections = new int[subsectionCount][];
subsections[0] = new int[] { 0, size }; // HACK: What is size? Contratiction in PDF reference.
subsectionEntryCount = size;
}
else
{
// Read subsections from array.
Debug.Assert(index.Elements.Count % 2 == 0);
subsectionCount = index.Elements.Count / 2;
subsections = new int[subsectionCount][];
for (int idx = 0; idx < subsectionCount; idx++)
{
subsections[idx] = new int[] { index.Elements.GetInteger(2 * idx), index.Elements.GetInteger(2 * idx + 1) };
subsectionEntryCount += subsections[idx][1];
}
}
// W key.
Debug.Assert(w.Elements.Count == 3);
int[] wsize = { w.Elements.GetInteger(0), w.Elements.GetInteger(1), w.Elements.GetInteger(2) };
int wsum = StreamHelper.WSize(wsize);
if (wsum * subsectionEntryCount != bytes.Length)
GetType();
Debug.Assert(wsum * subsectionEntryCount == bytes.Length, "Check implementation here.");
int testcount = subsections[0][1];
int[] currentSubsection = subsections[0];
#if DEBUG && CORE
if (PdfDiagnostics.TraceXrefStreams)
{
for (int idx = 0; idx < testcount; idx++)
{
uint field1 = StreamHelper.ReadBytes(bytes, idx * wsum, wsize[0]);
uint field2 = StreamHelper.ReadBytes(bytes, idx * wsum + wsize[0], wsize[1]);
uint field3 = StreamHelper.ReadBytes(bytes, idx * wsum + wsize[0] + wsize[1], wsize[2]);
string res = String.Format("{0,2:00}: {1} {2,5} {3} // ", idx, field1, field2, field3);
switch (field1)
{
case 0:
res += "Fee list: object number, generation number";
break;
case 1:
res += "Not compresed: offset, generation number";
break;
case 2:
res += "Compressed: object stream object number, index in stream";
break;
default:
res += "??? Type undefined";
break;
}
Debug.WriteLine(res);
}
}
#endif
int index2 = -1;
for (int ssc = 0; ssc < subsectionCount; ssc++)
{
int abc = subsections[ssc][1];
for (int idx = 0; idx < abc; idx++)
{
index2++;
PdfCrossReferenceStream.CrossReferenceStreamEntry item =
new PdfCrossReferenceStream.CrossReferenceStreamEntry();
item.Type = StreamHelper.ReadBytes(bytes, index2 * wsum, wsize[0]);
item.Field2 = StreamHelper.ReadBytes(bytes, index2 * wsum + wsize[0], wsize[1]);
item.Field3 = StreamHelper.ReadBytes(bytes, index2 * wsum + wsize[0] + wsize[1], wsize[2]);
xrefStream.Entries.Add(item);
switch (item.Type)
{
case 0:
// Nothing to do, not needed.
break;
case 1: // offset / generation number
//// Even it is restricted, an object can exists in more than one subsection.
//// (PDF Reference Implementation Notes 15).
int position = (int)item.Field2;
objectID = ReadObjectNumber(position);
#if DEBUG
if (objectID.ObjectNumber == 1074)
GetType();
#endif
Debug.Assert(objectID.GenerationNumber == item.Field3);
//// Ignore the latter one.
if (!xrefTable.Contains(objectID))
{
#if DEBUG
GetType();
#endif
// Add iref for all uncrompressed objects.
xrefTable.Add(new PdfReference(objectID, position));
}
break;
case 2:
// Nothing to do yet.
break;
}
}
}
return xrefStream;
}
/// <summary>
/// Indicates whether the specified object identifier is in the table.
/// </summary>
public bool Contains(PdfObjectID objectID)
{
return(ObjectTable.ContainsKey(objectID));
}
/// <summary>
/// Gets a cross reference entry from an object identifier.
/// Returns null if no object with the specified ID exists in the object table.
/// </summary>
public PdfReference this[PdfObjectID objectID]
{
get
{
PdfReference iref;
ObjectTable.TryGetValue(objectID, out iref);
return iref;
}
}
