/// <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.
/// </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>
/// Constructs the PdfTrailer from a document.
/// </summary>
/// <param name="parser">the parser used to read the file.</param>
internal void ConstructFromDocument(Parser parser)
{
// TODO - May need to also search for encryption related trailer info
PdfCrossReferenceTable xrefTable = _document._irefTable;
Elements.SetInteger(Keys.Size, xrefTable.ObjectTable.Count);
// find the root.
PdfDictionary rootToUse = null;
foreach (var reference in xrefTable.AllReferences)
{
PdfObject obj = parser.ReadObject(null, reference.ObjectID, false, false);
if (obj is PdfDictionary dObj)
{
if (dObj.Elements[PdfCatalog.Keys.Type] as PdfName == "/Catalog")
{
if (rootToUse == null)
{
rootToUse = dObj;
}
else if (dObj.ObjectID.GenerationNumber > rootToUse.ObjectID.GenerationNumber)
{
rootToUse = dObj;
}
}
}
}
if (rootToUse != null)
{
Elements.SetReference(Keys.Root, rootToUse);
}
}
/// <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>
/// 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;
}
void Initialize()
{
//_info = new PdfInfo(this);
_fontTable = new PdfFontTable(this);
_imageTable = new PdfImageTable(this);
_trailer = new PdfTrailer(this);
_irefTable = new PdfCrossReferenceTable(this);
_trailer.CreateNewDocumentIDs();
}
internal PdfDocument(Lexer lexer)
{
//PdfDocument.Gob.AttatchDocument(Handle);
_creation = DateTime.Now;
_state = DocumentState.Imported;
//_info = new PdfInfo(this);
//_pages = new PdfPages(this);
//_fontTable = new PdfFontTable();
//_catalog = new PdfCatalog(this);
////_font = new PdfFont();
//_objects = new PdfObjectTable(this);
//_trailer = new PdfTrailer(this);
_irefTable = new PdfCrossReferenceTable(this);
_lexer = lexer;
}