diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Internal/Diagnostics.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Internal/Diagnostics.cs index 6b7b4bf4..35abe830 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Internal/Diagnostics.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Internal/Diagnostics.cs @@ -73,11 +73,11 @@ public static void HandleUnexpectedCharacter(char ch) "If you think this is a bug in PDFsharp, please send us your PDF file.", (int)ch); ThrowParserException(message); } - public static void HandleUnexpectedToken(string token) + public static void HandleUnexpectedToken(string token, int position) { string message = String.Format(CultureInfo.InvariantCulture, - "Unexpected token '{0}' in PDF stream. The file may be corrupted. " + - "If you think this is a bug in PDFsharp, please send us your PDF file.", token); + "Unexpected token '{0}' at position {1} in PDF stream. The file may be corrupted. " + + "If you think this is a bug in PDFsharp, please send us your PDF file.", token, position); ThrowParserException(message); } } diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfCrossReferenceStream.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfCrossReferenceStream.cs index 6c3de94c..67baf9e9 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfCrossReferenceStream.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfCrossReferenceStream.cs @@ -12,7 +12,7 @@ namespace PdfSharp.Pdf.Advanced sealed class PdfCrossReferenceStream : PdfTrailer // Reference: 3.4.7 Cross-Reference Streams / Page 106 { /// - /// Initializes a new instance of the class. + /// Initializes a new instance of the class. /// public PdfCrossReferenceStream(PdfDocument document) : base(document) diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfObjectStream.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfObjectStream.cs index bbe0ced5..4ecef76c 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfObjectStream.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfObjectStream.cs @@ -49,37 +49,6 @@ internal PdfObjectStream(PdfDictionary dict) #endif } - /// - /// Reads the compressed object with the specified index. - /// - 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 - { -#if DEBUG - GetType(); -#endif - } - } - } - /// /// Reads the compressed object with the specified index. /// @@ -108,7 +77,7 @@ public class Keys : PdfStream.Keys /// /// (Required) The type of PDF object that this dictionary describes; - /// must be ObjStmfor an object stream. + /// must be ObjStm for an object stream. /// [KeyInfo(KeyType.Name | KeyType.Required, FixedValue = "ObjStm")] public const string Type = "/Type"; @@ -130,7 +99,7 @@ public class Keys : PdfStream.Keys /// (Optional) A reference to an object stream, of which the current object /// stream is considered an extension. Both streams are considered part of /// a collection of object streams (see below). A given collection consists - /// of a set of streams whose Extendslinks form a directed acyclic graph. + /// of a set of streams whose Extends links form a directed acyclic graph. /// [KeyInfo(KeyType.Stream | KeyType.Optional)] public const string Extends = "/Extends"; diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfReference.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfReference.cs index 1c4ba091..b8c62596 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfReference.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Advanced/PdfReference.cs @@ -17,7 +17,7 @@ namespace PdfSharp.Pdf.Advanced /// Represents an indirect reference to a PdfObject. /// [DebuggerDisplay("iref({ObjectNumber}, {GenerationNumber})")] - public sealed class PdfReference : PdfItem + public class PdfReference : PdfItem { // About PdfReference // @@ -154,7 +154,7 @@ public int Position /// /// Gets or sets the referenced PdfObject. /// - public PdfObject Value + public virtual PdfObject Value { get => _value; set @@ -246,4 +246,93 @@ public int Compare(PdfReference? l, PdfReference? r) int _uid; #endif } + + /// + /// Represents an indirect reference to an object stored in an

+ /// The value of this object is "lazily" loaded when first accessed. + /// + public sealed class PdfReferenceToCompressedObject : PdfReference + { + private readonly int _objectStreamNumber; + private readonly int _indexInObjectStream; + + internal PdfReferenceToCompressedObject(PdfDocument doc, PdfObjectID objectID, + int objectStreamNumber, int indexInObjectStream) + : base(objectID, -1) + { + Document = doc ?? throw new ArgumentNullException(nameof(doc)); + _objectStreamNumber = objectStreamNumber; + _indexInObjectStream = indexInObjectStream; + } + + public override PdfObject Value + { + get + { + if (base.Value is null) + { + ReadValue(); + } + return base.Value!; + } + set => base.Value = value; + } + + /// + /// Reads the value of this object + /// + void ReadValue() + { + PdfObjectStream? ostm = null; + var stmObjID = new PdfObjectID(_objectStreamNumber); + // reference to object stream + var streamRef = Document.IrefTable[stmObjID]; + if (streamRef is not null) + { + if (streamRef.Value is null) + { + // object stream not yet loaded. do it now + var parser = Document.GetParser()!; + var state = parser.SaveState(); + var obj = parser.ReadObject(null, stmObjID, false, false); + if (obj is PdfDictionary ostmDict) + { + // decrypt if necessary + // must be done before type-transformation because PdfObjectStream + // tries to parse the stream-header in the constructor + Document.EffectiveSecurityHandler?.DecryptObject(ostmDict); + ostm = new PdfObjectStream(ostmDict); + } + parser.RestoreState(state); + Debug.Assert(ostm != null, "Object stream should not be null here"); + } + // already transformed ? + else if (streamRef.Value is not PdfObjectStream existingOstm) + { + if (streamRef.Value is PdfDictionary ostmDict) + { + // decrypt if necessary + Document.EffectiveSecurityHandler?.DecryptObject(ostmDict); + ostm = new PdfObjectStream(ostmDict); + } + Debug.Assert(ostm != null, "Object stream should not be null here"); + } + else + ostm = existingOstm; + + if (ostm is not null) + { + // store the loaded and decrypted object-stream + streamRef.Value = ostm; + // read the actual object we're looking for + var iref = ostm.ReadCompressedObject(_indexInObjectStream); + if (iref is not null) + { + Debug.Assert(iref.ObjectID == ObjectID, "ObjectID mismatch"); + base.Value = iref.Value; + } + } + } + } + } } diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Lexer.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Lexer.cs index b6f7c590..9909ce62 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Lexer.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Lexer.cs @@ -937,8 +937,8 @@ public double TokenToReal { get { - // ReSharper disable once CompareOfFloatsByEqualityOperator - Debug.Assert(_tokenAsReal == double.Parse(_token.ToString(), CultureInfo.InvariantCulture)); + // had several documents where the assertion failed with an equality comparision (==) + Debug.Assert(Math.Abs(_tokenAsReal - double.Parse(_token.ToString(), CultureInfo.InvariantCulture)) < 0.000000001); return _tokenAsReal; } } diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Parser.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Parser.cs index 1dc257f9..3b387e34 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Parser.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/Parser.cs @@ -42,10 +42,10 @@ public Parser(PdfDocument? document, Stream pdf) _stack = new ShiftStack(); } - public Parser(PdfDocument document) + public Parser(PdfDocument document, Lexer lexer) { - _document = document; - _lexer = document?._lexer ?? throw new ArgumentNullException(nameof(document), "Lexer not defined."); + _document = document ?? throw new ArgumentNullException(nameof(document)); + _lexer = lexer ?? throw new ArgumentNullException(nameof(lexer)); _stack = new ShiftStack(); } @@ -105,7 +105,7 @@ public PdfObject ReadObject(PdfObject? pdfObject, PdfObjectID objectID, bool inc { // Attempt to read an object that was already registered. Keep the former object. // This only happens with corrupt PDF files that have duplicate IDs. - if (iref.Value != null!) + if (iref is not PdfReferenceToCompressedObject && iref.Value != null!) { LogHost.Logger.LogWarning("Another instance of object {iref} was found. Using previously encountered object instead.", iref); // Attempt to read an object that was already read. Keep the former object. @@ -258,12 +258,12 @@ public PdfObject ReadObject(PdfObject? pdfObject, PdfObjectID objectID, bool inc case Symbol.Keyword: // Should not come here anymore. - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); break; default: // Should not come here anymore. - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); break; } symbol = ScanNextToken(); @@ -344,19 +344,10 @@ int GetStreamLength(PdfDictionary dict) { #if true object length; - if (reference.Position == -1 && reference.Value != null!) - { - if (reference.Value is not PdfIntegerObject integer) - throw new InvalidOperationException("Cannot retrieve stream length from stream object."); + if (reference.Value is not PdfIntegerObject integer) + throw new InvalidOperationException("Cannot retrieve stream length from stream object."); - length = integer; - } - else - { - ParserState state = SaveState(); - length = ReadObject(null, reference.ObjectID, false, false); - RestoreState(state); - } + length = integer; #else ParserState state = SaveState(); object length = ReadObject(null, reference.ObjectID, false, false); @@ -558,7 +549,7 @@ void ParseObject(Symbol stop) //case Symbol.StartXRef: //case Symbol.Eof: default: - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); SkipCharsUntil(stop); return; } @@ -673,7 +664,7 @@ Symbol ReadSymbol(Symbol symbol) } Symbol current = _lexer.ScanNextToken(); if (symbol != current) - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); return current; } @@ -684,7 +675,7 @@ Symbol ReadToken(string token) { Symbol current = _lexer.ScanNextToken(); if (token != _lexer.Token) - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); return current; } @@ -696,7 +687,7 @@ string ReadName() string name; Symbol symbol = ScanNextToken(out name); if (symbol != Symbol.Name) - ParserDiagnostics.HandleUnexpectedToken(name); + ParserDiagnostics.HandleUnexpectedToken(name, _lexer.Position - name.Length); return name; } @@ -773,7 +764,7 @@ int ReadInteger(bool canBeIndirect) _lexer.Position = position; return n; } - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); return 0; } @@ -823,115 +814,9 @@ int ReadInteger() // return parser.ReadObject(pdfObject, objectID, false); // } - /// - /// Reads an object from the PDF input stream using the default parser. - /// - public static PdfObject ReadObject(PdfDocument owner, PdfObjectID objectID) - { - if (owner == null) - throw new ArgumentNullException(nameof(owner)); - - Parser parser = new Parser(owner); - return parser.ReadObject(null, objectID, false, false); - } - - /// - /// Reads the irefs from the compressed object with the specified index in the object stream - /// of the object with the specified object id. - /// - internal void ReadIRefsFromCompressedObject(PdfObjectID objectID) - { - Debug.Assert(_document.IrefTable.ObjectTable.ContainsKey(objectID)); - if (!_document.IrefTable.ObjectTable.TryGetValue(objectID, out var 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"); - } - - Debug.Assert(iref.Value != null, "The object shall be read in by PdfReader.ReadIndirectObjectsFromIrefTable() before accessing it."); - - if (iref.Value is not PdfObjectStream objectStreamStream) - { - 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); - } - - /// - /// Reads the compressed object with the specified index in the object stream - /// of the object with the specified object ID. - /// - internal PdfReference ReadCompressedObject(PdfObjectID objectID, int index) - { -#if true - Debug.Assert(_document.IrefTable.ObjectTable.ContainsKey(objectID)); - if (!_document.IrefTable.ObjectTable.TryGetValue(objectID, out var 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 - - Debug.Assert(iref.Value != null, "The object shall be read in by PdfReader.ReadIndirectObjectsFromIrefTable() before accessing it."); - - var 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); - } - /// /// Reads the compressed object with the specified number at the given offset. - /// The parser must be initialized with the stream an object stream object. + /// The parser must be initialized with the stream of an object stream object. /// internal PdfReference ReadCompressedObject(int objectNumber, int offset) { @@ -1111,7 +996,7 @@ internal PdfTrailer ReadTrailer() return trailer; } else - ParserDiagnostics.HandleUnexpectedToken(_lexer.Token); + ParserDiagnostics.HandleUnexpectedToken(_lexer.Token, _lexer.Position - _lexer.Token.Length); } } // ReSharper disable once RedundantIfElseBlock because of code readability. @@ -1345,7 +1230,6 @@ PdfTrailer ReadXRefStream(PdfCrossReferenceTable xrefTable) #endif // Add iref for all uncompressed objects. xrefTable.Add(new PdfReference(objectID, position)); - } #if DEBUG else @@ -1356,7 +1240,14 @@ PdfTrailer ReadXRefStream(PdfCrossReferenceTable xrefTable) break; case 2: - // Nothing to do yet. + // object-stream number / index in object-stream + // collect irefs for objects stored in object streams + objectID = new PdfObjectID(subsections[ssc][0] + idx, 0); + if (!xrefTable.Contains(objectID)) + { + xrefTable.Add(new PdfReferenceToCompressedObject(_document, objectID, + (int)item.Field2, (int)item.Field3)); + } break; } } @@ -1735,7 +1626,7 @@ public static object Read(PdfObject o, string key) } */ - ParserState SaveState() + internal ParserState SaveState() { var state = new ParserState { @@ -1745,13 +1636,13 @@ ParserState SaveState() return state; } - void RestoreState(ParserState state) + internal void RestoreState(ParserState state) { _lexer.Position = state.Position; _lexer.Symbol = state.Symbol; } - class ParserState + internal class ParserState { public int Position; public Symbol Symbol; diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/PdfReader.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/PdfReader.cs index 01162dee..f36a14b2 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/PdfReader.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.IO/PdfReader.cs @@ -259,7 +259,7 @@ public static PdfDocument Open(Stream stream, string? password, PdfDocumentOpenM // After reading all objects, all documents placeholder references get replaced by references knowing their objects in FinishReferences(), // which finally sets IsUnderConstruction to false. document.IrefTable.IsUnderConstruction = true; - var parser = new Parser(document); + var parser = document.GetParser()!; // Read all trailers or cross-reference streams, but no objects. document.Trailer = parser.ReadTrailer(); if (document.Trailer == null!) @@ -324,16 +324,10 @@ public static PdfDocument Open(Stream stream, string? password, PdfDocumentOpenM } } - // Read all file level indirect objects and decrypt them. - // Objects stored in object streams are not yet included and must not be decrypted as they are not encrypted. + // Read all indirect objects and decrypt them. + // This includes objects stored in object-streams ReadIndirectObjectsFromIrefTable(document, parser, true); - // Read all indirect objects stored in object streams. - ReadCompressedObjects(document, parser); - - // Read all not yet known indirect objects (the ones that were stored in object streams) and don't decrypt them, as they are not encrypted. - ReadIndirectObjectsFromIrefTable(document, parser, false); - // Reset encryption so that it must be redefined to save the document encrypted. effectiveSecurityHandler?.SetEncryptionToNoneAndResetPasswords(); @@ -412,7 +406,8 @@ static void ReadIndirectObjectsFromIrefTable(PdfDocument document, Parser parser } // Decrypt object, if needed. - if (decrypt && iref.Value is { } pdfObject2) + // skip objects stored in object streams (iref.Position = -1) + if (decrypt && iref.Position >= 0 && iref.Value is { } pdfObject2) effectiveSecurityHandler?.DecryptObject(pdfObject2); // Set maximum object number. @@ -420,57 +415,6 @@ static void ReadIndirectObjectsFromIrefTable(PdfDocument document, Parser parser } } - static void ReadCompressedObjects(PdfDocument document, Parser parser) - { - // The PDF Reference 1.7 states in chapter 7.5.6 (Incremental Updates): - // "When a conforming reader reads the file, - // it shall build its cross-reference information in such a way that the - // most recent copy of each object shall be the one accessed from the file." - - // IrefTable.AllReferences is sorted by ObjectId which gives older objects preference - // (as they typically have lower ObjectIds). - // For xref-streams, we revert the order, so that the most current object is read first. - // This is because Parser.ReadObject does not overwrite an object that was already collected. - - // Collect xref streams. - var xrefStreams = new List(); - foreach (var iref in document.IrefTable.AllReferences) - { - if (iref.Value is PdfCrossReferenceStream xrefStream) - xrefStreams.Add(xrefStream); - } - // Sort them so the last xref stream is read first. - // TODO: Is this always sufficient? (haven't found any issues so far testing with ~1300 PDFs...) - xrefStreams.Sort((a, b) => (b.Reference?.Position ?? 0) - (a.Reference?.Position ?? 0)); - - - - Dictionary objectStreams = new(); - foreach (var xrefStream in xrefStreams) - { - foreach (var item in xrefStream.Entries) - { - // Is type xref to compressed object? - if (item.Type == 2) - { - int objectNumber = (int)item.Field2; - - if (!objectStreams.ContainsKey(objectNumber)) - { - objectStreams.Add(objectNumber, null); - var objectID = new PdfObjectID((int)item.Field2); - parser.ReadIRefsFromCompressedObject(objectID); - } - - PdfReference irefNew = parser.ReadCompressedObject(new PdfObjectID((int)item.Field2), - (int)item.Field3); - Debug.Assert(document.IrefTable.Contains(xrefStream.ObjectID)); - Debug.Assert(document.IrefTable.Contains(irefNew.ObjectID)); - } - } - } - } - static void FinishReferences(PdfDocument document) { Debug.Assert(document.IrefTable.IsUnderConstruction); @@ -555,6 +499,24 @@ static bool FinishReference(PdfReference currentReference, PdfDocument document, var isChanged = false; PdfItem? reference = currentReference; + if (reference is PdfReferenceToCompressedObject cref) + { + // replace with regular reference + // TODO: necessary ? does it make a difference ? + if (cref.Value is not null) + { + var newIref = new PdfReference(currentReference.ObjectID, -1) + { + Document = document, + Value = cref.Value + }; + document.IrefTable.Remove(currentReference); + document.IrefTable.Add(newIref); + reference = newIref; + isChanged = true; + } + } + // The value of the reference may be null. // If a file level PdfObject refers object stream level PdfObjects, that were not yet decompressed when reading it, // placeholder references are used. diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Security/PdfStandardSecurityHandler.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Security/PdfStandardSecurityHandler.cs index 74d48c7d..69fe28b2 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Security/PdfStandardSecurityHandler.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Security/PdfStandardSecurityHandler.cs @@ -267,7 +267,15 @@ public void DecryptObject(PdfObject value) switch (value) { case PdfDictionary vDict: - DecryptDictionary(vDict); + // this check is required for object-streams + // which may be "lazily" loaded and decrypted in PdfReferenceToCompressedObject.ReadValue + // alternatively we could populate a list of already decrypted objects, + // but this would probably required more memory + if (!vDict.AlreadyDecrypted) + { + DecryptDictionary(vDict); + vDict.AlreadyDecrypted = true; + } break; case PdfArray vArray: DecryptArray(vArray); diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDictionary.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDictionary.cs index 171d99dc..fd8d02a3 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDictionary.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDictionary.cs @@ -44,6 +44,12 @@ public class PdfDictionary : PdfObject, IEnumerable + /// Determines, whether this instance was already decrypted.

+ /// (in case the document is protected) + /// + internal bool AlreadyDecrypted { get; set; } + /// /// Initializes a new instance of the class. /// diff --git a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDocument.cs b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDocument.cs index 6022d85d..3974c58b 100644 --- a/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDocument.cs +++ b/src/foundation/src/PDFsharp/src/PdfSharp/Pdf/PdfDocument.cs @@ -120,6 +120,23 @@ void Initialize() Trailer.CreateNewDocumentIDs(); } + /// + /// Gets a for an imported document.

+ /// Returns null, if the document was not imported.

+ /// If this method is called multiple times for the same document, + /// the same parser-instance is returned each time. + /// + /// The parser-instance for imported documents or null when the document was not imported + internal Parser? GetParser() + { + if (_parser == null) + { + if (_lexer != null) + _parser = new Parser(this, _lexer); + } + return _parser; + } + //~PdfDocument() //{ // Dispose(false); @@ -813,6 +830,7 @@ public void Flatten() // Imported Document. internal Lexer? _lexer; + internal Parser? _parser; internal DateTime _creation;