public void ReaderEscapesUnexpectedObject()
{
const string s = @"%PDF-1.7
abcd
1 0 obj
<< /Type /Any >>
endobj
%AZ 0 obj
11 0 obj
769
endobj
%%EOF";
var bytes = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(s));
var locations = BruteForceSearcher.GetObjectLocations(bytes);
Assert.Equal(2, locations.Count);
var expectedLocations = new long[]
{
s.IndexOf("1 0 obj", StringComparison.OrdinalIgnoreCase),
s.IndexOf("11 0 obj", StringComparison.OrdinalIgnoreCase)
};
Assert.Equal(expectedLocations, locations.Values);
}
public void ReaderEscapesUnexpectedGenerationNumber()
{
const string s = @"%PDF-2.0
abcdefghijklmnop
1 0 obj
256
endobj
16-0 obj
5 0 obj
<< /IsEmpty false >>
endobj";
var bytes = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(s));
var locations = BruteForceSearcher.GetObjectLocations(bytes);
Assert.Equal(2, locations.Count);
var expectedLocations = new long[]
{
s.IndexOf("1 0 obj", StringComparison.OrdinalIgnoreCase),
s.IndexOf("5 0 obj", StringComparison.OrdinalIgnoreCase)
};
Assert.Equal(expectedLocations, locations.Values);
}
private static void AddAdobeFontMetrics(string fontName, string afmName, Standard14Font?type = null)
{
Standard14Names.Add(fontName);
Standard14Mapping.Add(fontName, afmName);
if (Standard14AfmMap.TryGetValue(afmName, out var metrics))
{
Standard14AfmMap[fontName] = metrics;
}
try
{
var assembly = typeof(Standard14).Assembly;
var name = $"UglyToad.PdfPig.Resources.AdobeFontMetrics.{afmName}.afm";
IInputBytes bytes;
using (var memory = new MemoryStream())
using (var resource = assembly.GetManifestResourceStream(name))
{
resource.CopyTo(memory);
bytes = new ByteArrayInputBytes(memory.ToArray());
}
Standard14AfmMap[fontName] = Parser.Parse(bytes, true);
if (type.HasValue)
{
Standard14AfmTypeMap[type.Value] = Standard14AfmMap[fontName];
}
}
catch (Exception ex)
{
throw new InvalidOperationException($"Could not load {fontName} from the AFM files.", ex);
}
}
public void ReadsStreamWithoutBreakBeforeEndstream()
{
const string s = @"
1 0 obj
12
endobj
7 0 obj
<< /Length 288
/Filter /FlateDecode >>
stream
xœ]‘ËjÃ0E÷ÿÃ,ÓEð#NÒ€1¤N^ôA~€-]A- YYøï+Ï4¡t#qfî FWQY* Dïv5:è”–§ñjB‹½Òa¤ •p7¤K ƒÈûëyr8Tº!Ïà úð‚ÉÙVG9¶ø@Å7+Ñ*ÝÃ곬¹T_ùƵƒ8 Š$vË̗Ƽ6BDöu%½B¹yí$—Ù ¤\Hx71JœL#Ð6ºÇ0È㸀ü|. µüßõÏ""WÛ‰¯Æ.êÄ«ã8; ¤iL°!Ø %É`K°ßì¸ÃöÜáÜ) [‚#CFðÄ°#(yƒg^ÿ¶æò
ÿž“¸Zë#¢?¢h–P”Æû?šÑï÷ø¯‰Šendstream
endobj
9 0 obj
16
endobj";
var inputBytes = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(s));
var scanner = new PdfTokenScanner(inputBytes, new TestObjectLocationProvider(), new TestFilterProvider());
var token = ReadToEnd(scanner)[1];
Assert.Equal(7, token.Number.ObjectNumber);
}
internal static (CoreTokenScanner scanner, IInputBytes bytes) Scanner(string s)
{
var inputBytes = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(s));
var result = new CoreTokenScanner(inputBytes);
return(result, inputBytes);
}
public void BruteForceSearcherFileOffsetsCorrectOpenOffice()
{
var bytes = new ByteArrayInputBytes(File.ReadAllBytes(IntegrationHelpers.GetDocumentPath("Single Page Simple - from open office.pdf")));
var locations = BruteForceSearcher.GetObjectLocations(bytes);
Assert.Equal(13, locations.Count);
Assert.Equal(17, locations[new IndirectReference(1, 0)]);
Assert.Equal(249, locations[new IndirectReference(2, 0)]);
Assert.Equal(14291, locations[new IndirectReference(3, 0)]);
Assert.Equal(275, locations[new IndirectReference(4, 0)]);
Assert.Equal(382, locations[new IndirectReference(5, 0)]);
Assert.Equal(13283, locations[new IndirectReference(6, 0)]);
Assert.Equal(13309, locations[new IndirectReference(7, 0)]);
Assert.Equal(13556, locations[new IndirectReference(8, 0)]);
Assert.Equal(13926, locations[new IndirectReference(9, 0)]);
Assert.Equal(14183, locations[new IndirectReference(10, 0)]);
Assert.Equal(14224, locations[new IndirectReference(11, 0)]);
Assert.Equal(14428, locations[new IndirectReference(12, 0)]);
Assert.Equal(14488, locations[new IndirectReference(13, 0)]);
var s = GetStringAt(bytes, locations[new IndirectReference(12, 0)]);
Assert.StartsWith("12 0 obj", s);
}
public void DoesNotStartWithOpenBracket_ReturnsFalse(char firstByte)
{
var input = new ByteArrayInputBytes(new[] {(byte)firstByte});
var result = tokenizer.TryTokenize((byte)firstByte, input, out var token);
Assert.False(result);
Assert.Null(token);
}
public void ArrayAndStreamBehaveTheSame()
{
var bytes = OtherEncodings.StringAsLatin1Bytes(TestData);
var array = new ByteArrayInputBytes(bytes);
using (var memoryStream = new MemoryStream(bytes))
{
var stream = new StreamInputBytes(memoryStream);
Assert.Equal(bytes.Length, array.Length);
Assert.Equal(bytes.Length, stream.Length);
Assert.Equal(0, array.CurrentOffset);
Assert.Equal(0, stream.CurrentOffset);
array.Seek(5);
stream.Seek(5);
Assert.Equal(array.CurrentOffset, stream.CurrentOffset);
Assert.Equal((byte)'5', array.CurrentByte);
Assert.Equal(array.CurrentByte, stream.CurrentByte);
Assert.Equal(array.Peek(), stream.Peek());
array.Seek(0);
stream.Seek(0);
Assert.Equal(0, array.CurrentByte);
Assert.Equal(array.CurrentByte, stream.CurrentByte);
array.Seek(7);
stream.Seek(7);
var arrayString = string.Empty;
var streamString = string.Empty;
while (array.MoveNext())
{
arrayString += (char)array.CurrentByte;
}
while (stream.MoveNext())
{
streamString += (char)stream.CurrentByte;
}
Assert.Equal("89", streamString);
Assert.Equal(arrayString, streamString);
Assert.True(stream.IsAtEnd());
Assert.True(array.IsAtEnd());
}
}
public void BruteForceSearcherCorrectlyFindsAllObjectsWhenOffset()
{
var input = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(TestData));
input.Seek(593);
var locations = BruteForceSearcher.GetObjectLocations(input);
Assert.Equal(TestDataOffsets, locations.Values);
}
public void CanParseHelveticaAfmFile()
{
var helvetica = GetResourceBytes("UglyToad.PdfPig.Fonts.Resources.AdobeFontMetrics.Helvetica.afm");
var input = new ByteArrayInputBytes(helvetica);
var metrics = AdobeFontMetricsParser.Parse(input, false);
Assert.NotNull(metrics);
}
public void CanParseAllPredefinedCMaps(string resourceName)
{
Debug.WriteLine("Parsing: " + resourceName);
var input = new ByteArrayInputBytes(ReadResourceBytes(resourceName));
var cmap = cMapParser.Parse(input, false);
Assert.NotNull(cmap);
}
public void SearcherFindsCorrectObjects()
{
var input = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(TestData));
var locations = BruteForceSearcher.GetObjectLocations(input);
Assert.Equal(4, locations.Count);
Assert.Equal(TestDataOffsets, locations.Values);
}
public void Issue334()
{
var input = OtherEncodings.StringAsLatin1Bytes("%PDF-1.7\r\n%âãÏÓ\r\n1 0 obj\r\n<</Lang(en-US)>>\r\nendobj");
var bytes = new ByteArrayInputBytes(input);
var scanner = new CoreTokenScanner(bytes, ScannerScope.None);
var result = FileHeaderParser.Parse(scanner, bytes, false, log);
Assert.Equal(1.7m, result.Version);
}
/// <summary>
/// Merge the set of PDF documents.
/// </summary>
public static byte[] Merge(IReadOnlyList <byte[]> files, IReadOnlyList <IReadOnlyList <int> > pagesBundle = null)
{
if (files == null)
{
throw new ArgumentNullException(nameof(files));
}
const bool isLenientParsing = false;
var documentBuilder = new DocumentMerger();
foreach (var fileIndex in Enumerable.Range(0, files.Count))
{
var file = files[fileIndex];
IReadOnlyList <int> pages = null;
if (pagesBundle != null && fileIndex < pagesBundle.Count)
{
pages = pagesBundle[fileIndex];
}
var inputBytes = new ByteArrayInputBytes(file);
var coreScanner = new CoreTokenScanner(inputBytes);
var version = FileHeaderParser.Parse(coreScanner, isLenientParsing, Log);
var crossReferenceParser = new CrossReferenceParser(Log, new XrefOffsetValidator(Log),
new Parser.Parts.CrossReference.CrossReferenceStreamParser(FilterProvider));
CrossReferenceTable crossReference = null;
// ReSharper disable once AccessToModifiedClosure
var locationProvider = new ObjectLocationProvider(() => crossReference, inputBytes);
var pdfScanner = new PdfTokenScanner(inputBytes, locationProvider, FilterProvider, NoOpEncryptionHandler.Instance);
var crossReferenceOffset = FileTrailerParser.GetFirstCrossReferenceOffset(inputBytes, coreScanner, isLenientParsing);
crossReference = crossReferenceParser.Parse(inputBytes, isLenientParsing, crossReferenceOffset, version.OffsetInFile, pdfScanner, coreScanner);
var catalogDictionaryToken = ParseCatalog(crossReference, pdfScanner, out var encryptionDictionary);
if (encryptionDictionary != null)
{
throw new PdfDocumentEncryptedException("Unable to merge document with password");
}
var documentCatalog = CatalogFactory.Create(crossReference.Trailer.Root, catalogDictionaryToken, pdfScanner, isLenientParsing);
documentBuilder.AppendDocument(documentCatalog, version.Version, pdfScanner, pages);
}
return(documentBuilder.Build());
}
public void HandlesUtf16BigEndianStrings()
{
var input = new ByteArrayInputBytes(new byte[]
{
0xFF, 0xFE, 0x4D, 0x00, 0x69, 0x00, 0x63,
0x00, 0x29
});
var result = tokenizer.TryTokenize(0x28, input, out var token);
Assert.True(result);
Assert.Equal(@"Mic", AssertStringToken(token).Data);
}
private void Run(byte[] bytes, bool checkHeaderChecksum, bool checkWholeFileChecksum)
{
var inputBytes = new ByteArrayInputBytes(bytes);
var font = TrueTypeFontParser.Parse(new TrueTypeDataBytes(inputBytes));
inputBytes = new ByteArrayInputBytes(bytes);
foreach (var header in font.TableHeaders)
{
// Acts as the whole table checksum
if (header.Key == "head")
{
if (checkHeaderChecksum)
{
var headerChecksum = TrueTypeChecksumCalculator.Calculate(inputBytes, header.Value);
Assert.Equal(header.Value.CheckSum, headerChecksum);
}
continue;
}
var input = bytes.Skip((int)header.Value.Offset).Take((int)header.Value.Length);
var checksum = TrueTypeChecksumCalculator.Calculate(input);
Assert.Equal(header.Value.CheckSum, checksum);
var checksumByTable = TrueTypeChecksumCalculator.Calculate(inputBytes, header.Value);
Assert.Equal(header.Value.CheckSum, checksumByTable);
}
if (checkWholeFileChecksum)
{
var headerActual = font.TableHeaders["head"];
var wholeFontChecksum = TrueTypeChecksumCalculator.CalculateWholeFontChecksum(inputBytes, headerActual);
var adjustment = 0xB1B0AFBA - wholeFontChecksum;
var adjustmentRecorded = font.TableRegister.HeaderTable.CheckSumAdjustment;
Assert.Equal(adjustmentRecorded, adjustment);
var expectedWholeFontChecksum = 0xB1B0AFBA - adjustmentRecorded;
Assert.Equal(expectedWholeFontChecksum, wholeFontChecksum);
}
}
public static PdfDocument Open(byte[] fileBytes, ParsingOptions options = null)
{
var container = Bootstrapper.GenerateContainer(options?.Logger);
var isLenientParsing = options?.UseLenientParsing ?? true;
var reader = new RandomAccessBuffer(fileBytes);
var inputBytes = new ByteArrayInputBytes(fileBytes);
var tokenScanner = new CoreTokenScanner(inputBytes);
var document = OpenDocument(reader, inputBytes, tokenScanner, container, isLenientParsing);
return(document);
}
public void ReadUnsignedInt()
{
var input = new ByteArrayInputBytes(new byte[]
{
220,
43,
250,
6
});
var data = new TrueTypeDataBytes(input);
var result = data.ReadUnsignedInt();
Assert.Equal(3693869574L, result);
}
public void SearcherFindsCorrectObjects()
{
var input = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes(TestData));
var searcher = new BruteForceSearcher(input);
var locations = searcher.GetObjectLocations();
Assert.Equal(4, locations.Count);
Assert.Equal(locations.Values, new long[]
{
TestData.IndexOf("2 17 obj", StringComparison.OrdinalIgnoreCase),
TestData.IndexOf("3 0 obj", StringComparison.OrdinalIgnoreCase),
TestData.IndexOf("4 0 obj", StringComparison.OrdinalIgnoreCase),
TestData.IndexOf("5 0 obj", StringComparison.OrdinalIgnoreCase)
});
}
public static Result Convert(string s, bool readFirst = true)
{
var input = new ByteArrayInputBytes(Encoding.UTF8.GetBytes(s));
byte initialByte = 0;
if (readFirst)
{
input.MoveNext();
initialByte = input.CurrentByte;
}
return(new Result
{
First = initialByte,
Bytes = input
});
}
public void CanParseIdentityHorizontalCMap()
{
var input = new ByteArrayInputBytes(ReadResourceBytes("UglyToad.Pdf.Resources.CMap.Identity-H"));
var cmap = cMapParser.Parse(input, false);
Assert.Equal(1, cmap.CodespaceRanges.Count);
var range = cmap.CodespaceRanges[0];
Assert.Equal(0, range.StartInt);
Assert.Equal(65535, range.EndInt);
Assert.Equal(2, range.CodeLength);
Assert.Equal(256, cmap.CidRanges.Count);
Assert.Equal("10.003", cmap.Version);
}
public void BruteForceSearcherBytesFileOffsetsCorrect()
{
var bytes = new ByteArrayInputBytes(File.ReadAllBytes(IntegrationHelpers.GetDocumentPath("Single Page Simple - from inkscape.pdf")));
var locations = BruteForceSearcher.GetObjectLocations(bytes);
Assert.Equal(13, locations.Count);
Assert.Equal(6183, locations[new IndirectReference(1, 0)]);
Assert.Equal(244, locations[new IndirectReference(2, 0)]);
Assert.Equal(15, locations[new IndirectReference(3, 0)]);
Assert.Equal(222, locations[new IndirectReference(4, 0)]);
Assert.Equal(5766, locations[new IndirectReference(5, 0)]);
Assert.Equal(353, locations[new IndirectReference(6, 0)]);
Assert.Equal(581, locations[new IndirectReference(7, 0)]);
Assert.Equal(5068, locations[new IndirectReference(8, 0)]);
Assert.Equal(5091, locations[new IndirectReference(9, 0)]);
var s = GetStringAt(bytes, locations[new IndirectReference(3, 0)]);
Assert.StartsWith("3 0 obj", s);
}
public void ColorspaceParserError()
{
var parser = new CodespaceRangeParser();
var byteArrayInput = new ByteArrayInputBytes(OtherEncodings.StringAsLatin1Bytes("1 begincodespacerange\nendcodespacerange"));
var tokenScanner = new CoreTokenScanner(byteArrayInput);
Assert.True(tokenScanner.MoveNext());
Assert.True(tokenScanner.CurrentToken is NumericToken);
var numeric = (NumericToken)tokenScanner.CurrentToken;
Assert.True(tokenScanner.MoveNext());
Assert.True(tokenScanner.CurrentToken is OperatorToken);
var opToken = (OperatorToken)tokenScanner.CurrentToken;
Assert.Equal("begincodespacerange", opToken.Data);
var cmapBuilder = new CharacterMapBuilder();
parser.Parse(numeric, tokenScanner, cmapBuilder);
Assert.Empty(cmapBuilder.CodespaceRanges);
}
/// <summary>
/// Parses an embedded Adobe Type 1 font file.
/// </summary>
/// <param name="inputBytes">The bytes of the font program.</param>
/// <param name="length1">The length in bytes of the clear text portion of the font program.</param>
/// <param name="length2">The length in bytes of the encrypted portion of the font program.</param>
/// <returns>The parsed type 1 font.</returns>
public Type1FontProgram Parse(IInputBytes inputBytes, int length1, int length2)
{
// Sometimes the entire PFB file including the header bytes can be included which prevents parsing in the normal way.
var isEntirePfbFile = inputBytes.Peek() == PfbFileIndicator;
IReadOnlyList <byte> eexecPortion = new byte[0];
if (isEntirePfbFile)
{
var(ascii, binary) = ReadPfbHeader(inputBytes);
eexecPortion = binary;
inputBytes = new ByteArrayInputBytes(ascii);
}
var scanner = new CoreTokenScanner(inputBytes);
if (!scanner.TryReadToken(out CommentToken comment) || !comment.Data.StartsWith("!"))
{
throw new InvalidFontFormatException("The Type1 program did not start with '%!'.");
}
string name;
var parts = comment.Data.Split(new[] { " " }, StringSplitOptions.RemoveEmptyEntries);
if (parts.Length == 3)
{
name = parts[1];
}
else
{
name = "Unknown";
}
var comments = new List <string>();
while (scanner.MoveNext() && scanner.CurrentToken is CommentToken commentToken)
{
comments.Add(commentToken.Data);
}
var dictionaries = new List <DictionaryToken>();
// Override arrays and names since type 1 handles these differently.
var arrayTokenizer = new Type1ArrayTokenizer();
var nameTokenizer = new Type1NameTokenizer();
scanner.RegisterCustomTokenizer((byte)'{', arrayTokenizer);
scanner.RegisterCustomTokenizer((byte)'/', nameTokenizer);
try
{
var tempEexecPortion = new List <byte>();
var tokenSet = new PreviousTokenSet();
tokenSet.Add(scanner.CurrentToken);
while (scanner.MoveNext())
{
if (scanner.CurrentToken is OperatorToken operatorToken)
{
if (Equals(scanner.CurrentToken, OperatorToken.Eexec))
{
int offset = 0;
while (inputBytes.MoveNext())
{
if (inputBytes.CurrentByte == (byte)ClearToMark[offset])
{
offset++;
}
else
{
if (offset > 0)
{
for (int i = 0; i < offset; i++)
{
tempEexecPortion.Add((byte)ClearToMark[i]);
}
}
offset = 0;
}
if (offset == ClearToMark.Length)
{
break;
}
if (offset > 0)
{
continue;
}
tempEexecPortion.Add(inputBytes.CurrentByte);
}
}
else
{
HandleOperator(operatorToken, scanner, tokenSet, dictionaries);
}
}
tokenSet.Add(scanner.CurrentToken);
}
if (!isEntirePfbFile)
{
eexecPortion = tempEexecPortion;
}
}
finally
{
scanner.DeregisterCustomTokenizer(arrayTokenizer);
scanner.DeregisterCustomTokenizer(nameTokenizer);
}
var encoding = GetEncoding(dictionaries);
var matrix = GetFontMatrix(dictionaries);
var boundingBox = GetBoundingBox(dictionaries);
var(privateDictionary, charStrings) = encryptedPortionParser.Parse(eexecPortion, false);
return(new Type1FontProgram(name, encoding, matrix, boundingBox ?? new PdfRectangle(), privateDictionary, charStrings));
}
/// <inheritdoc />
public void Run(IOperationContext operationContext)
{
var input = new ByteArrayInputBytes(Text != null ? OtherEncodings.StringAsLatin1Bytes(Text) : Bytes);
operationContext.ShowText(input);
}
public static PdfDocument Open(byte[] fileBytes, ParsingOptions options = null)
{
var inputBytes = new ByteArrayInputBytes(fileBytes);
return(Open(inputBytes, options));
}
