public IFont Get(DictionaryToken dictionary, bool isLenientParsing)
{
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Font))
{
var message = "The font dictionary did not have type 'Font'. " + dictionary;
if (isLenientParsing)
{
log?.Error(message);
}
else
{
throw new InvalidFontFormatException(message);
}
}
var subtype = dictionary.GetNameOrDefault(NameToken.Subtype);
if (handlers.TryGetValue(subtype, out var handler))
{
return(handler.Generate(dictionary, isLenientParsing));
}
throw new NotImplementedException($"Parsing not implemented for fonts of type: {subtype}, please submit a pull request or an issue.");
}
public ICidFont Generate(DictionaryToken dictionary, bool isLenientParsing)
{
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (!NameToken.Font.Equals(type))
{
throw new InvalidFontFormatException($"Expected \'Font\' dictionary but found \'{type}\'");
}
var widths = ReadWidths(dictionary);
var defaultWidth = default(double?);
if (dictionary.TryGet(NameToken.Dw, pdfScanner, out NumericToken defaultWidthToken))
{
defaultWidth = defaultWidthToken.Double;
}
var verticalWritingMetrics = ReadVerticalDisplacements(dictionary);
FontDescriptor descriptor = null;
if (TryGetFontDescriptor(dictionary, out var descriptorDictionary))
{
descriptor = descriptorFactory.Generate(descriptorDictionary, pdfScanner, isLenientParsing);
}
var fontProgram = ReadDescriptorFile(descriptor);
var baseFont = dictionary.GetNameOrDefault(NameToken.BaseFont);
var systemInfo = GetSystemInfo(dictionary);
var subType = dictionary.GetNameOrDefault(NameToken.Subtype);
if (NameToken.CidFontType0.Equals(subType))
{
return(new Type0CidFont(fontProgram, type, subType, baseFont, systemInfo, descriptor, verticalWritingMetrics, widths, defaultWidth));
}
if (NameToken.CidFontType2.Equals(subType))
{
var cidToGid = GetCharacterIdentifierToGlyphIndexMap(dictionary, isLenientParsing);
return(new Type2CidFont(type, subType, baseFont, systemInfo, descriptor, fontProgram, verticalWritingMetrics, widths, defaultWidth, cidToGid));
}
return(null);
}
private static CMap GetUcs2CMap(DictionaryToken dictionary, bool isCMapPredefined, bool usesDescendantAdobeFont)
{
if (!isCMapPredefined)
{
return(null);
}
/*
* If the font is a composite font that uses one of the predefined CMaps except Identity–H and Identity–V or whose descendant
* CIDFont uses the Adobe-GB1, Adobe-CNS1, Adobe-Japan1, or Adobe-Korea1 character collection use a UCS2 CMap.
*/
var encodingName = dictionary.GetNameOrDefault(NameToken.Encoding);
if (encodingName == null)
{
return(null);
}
var isPredefinedIdentityMap = encodingName.Equals(NameToken.IdentityH) || encodingName.Equals(NameToken.IdentityV);
if (isPredefinedIdentityMap && !usesDescendantAdobeFont)
{
return(null);
}
throw new NotSupportedException("Support for UCS2 CMaps are not implemented yet. Please raise an issue.");
}
public IFont Generate(DictionaryToken dictionary, bool isLenientParsing)
{
var baseFont = dictionary.GetNameOrDefault(NameToken.BaseFont);
var cMap = ReadEncoding(dictionary, out var isCMapPredefined);
ICidFont cidFont;
if (TryGetFirstDescendant(dictionary, out var descendantObject))
{
DictionaryToken descendantFontDictionary;
if (descendantObject is IndirectReferenceToken obj)
{
var parsed = DirectObjectFinder.Get<DictionaryToken>(obj, scanner);
descendantFontDictionary = parsed;
}
else
{
descendantFontDictionary = (DictionaryToken) descendantObject;
}
cidFont = ParseDescendant(descendantFontDictionary, isLenientParsing);
}
else
{
throw new InvalidFontFormatException("No descendant font dictionary was declared for this Type 0 font. This dictionary should contain the CIDFont for the Type 0 font. " + dictionary);
}
var (ucs2CMap, isChineseJapaneseOrKorean) = GetUcs2CMap(dictionary, isCMapPredefined, cidFont);
CMap toUnicodeCMap = null;
if (dictionary.ContainsKey(NameToken.ToUnicode))
{
var toUnicodeValue = dictionary.Data[NameToken.ToUnicode];
if (DirectObjectFinder.TryGet<StreamToken>(toUnicodeValue, scanner, out var toUnicodeStream))
{
var decodedUnicodeCMap = toUnicodeStream?.Decode(filterProvider);
if (decodedUnicodeCMap != null)
{
toUnicodeCMap = CMapCache.Parse(new ByteArrayInputBytes(decodedUnicodeCMap), isLenientParsing);
}
}
else if (DirectObjectFinder.TryGet<NameToken>(toUnicodeValue, scanner, out var toUnicodeName))
{
toUnicodeCMap = CMapCache.Get(toUnicodeName.Data);
}
else
{
throw new PdfDocumentFormatException($"Invalid type of toUnicode CMap encountered. Got: {toUnicodeValue}.");
}
}
var font = new Type0Font(baseFont, cidFont, cMap, toUnicodeCMap, ucs2CMap, isChineseJapaneseOrKorean);
return font;
}
public bool FindPage(DictionaryToken currentPageDictionary, int soughtPageNumber, List <int> pageNumbersObserved, PageTreeMembers pageTreeMembers)
{
var type = currentPageDictionary.GetNameOrDefault(NameToken.Type);
if (type?.Equals(NameToken.Page) == true)
{
var pageNumber = GetNextPageNumber(pageNumbersObserved);
bool found = pageNumber == soughtPageNumber;
locatedPages[pageNumber] = currentPageDictionary;
pageNumbersObserved.Add(pageNumber);
return(found);
}
if (type?.Equals(NameToken.Pages) != true)
{
log.Warn("Did not find the expected type (Page or Pages) in dictionary: " + currentPageDictionary);
return(false);
}
if (currentPageDictionary.TryGet(NameToken.MediaBox, out var token))
{
var mediaBox = DirectObjectFinder.Get <ArrayToken>(token, pdfScanner);
pageTreeMembers.MediaBox = new MediaBox(new PdfRectangle(mediaBox.GetNumeric(0).Data,
mediaBox.GetNumeric(1).Data,
mediaBox.GetNumeric(2).Data,
mediaBox.GetNumeric(3).Data));
}
if (!currentPageDictionary.TryGet(NameToken.Kids, out var kids) ||
!(kids is ArrayToken kidsArray))
{
return(false);
}
pageFactory.LoadResources(currentPageDictionary, isLenientParsing);
bool childFound = false;
foreach (var kid in kidsArray.Data)
{
// todo: exit early
var child = DirectObjectFinder.Get <DictionaryToken>(kid, pdfScanner);
var thisPageMatches = FindPage(child, soughtPageNumber, pageNumbersObserved, pageTreeMembers);
if (thisPageMatches)
{
childFound = true;
break;
}
}
return(childFound);
}
private static (CMap, bool isChineseJapaneseOrKorean) GetUcs2CMap(DictionaryToken dictionary, bool isCMapPredefined, ICidFont cidFont)
{
if (!isCMapPredefined)
{
return(null, false);
}
/*
* If the font is a composite font that uses one of the predefined CMaps except Identity–H and Identity–V or whose descendant
* CIDFont uses the Adobe-GB1, Adobe-CNS1, Adobe-Japan1, or Adobe-Korea1 character collection use a UCS2 CMap.
*/
var encodingName = dictionary.GetNameOrDefault(NameToken.Encoding);
if (encodingName == null)
{
return(null, false);
}
var isChineseJapaneseOrKorean = false;
if (cidFont != null && string.Equals(cidFont.SystemInfo.Registry, "Adobe", StringComparison.OrdinalIgnoreCase))
{
isChineseJapaneseOrKorean = string.Equals(cidFont.SystemInfo.Ordering, "GB1", StringComparison.OrdinalIgnoreCase) ||
string.Equals(cidFont.SystemInfo.Ordering, "CNS1", StringComparison.OrdinalIgnoreCase) ||
string.Equals(cidFont.SystemInfo.Ordering, "Japan1", StringComparison.OrdinalIgnoreCase) ||
string.Equals(cidFont.SystemInfo.Ordering, "Korea1", StringComparison.OrdinalIgnoreCase);
}
var isPredefinedIdentityMap = encodingName.Equals(NameToken.IdentityH) || encodingName.Equals(NameToken.IdentityV);
if (isPredefinedIdentityMap && !isChineseJapaneseOrKorean)
{
return(null, false);
}
if (!isChineseJapaneseOrKorean)
{
return(null, false);
}
var fullCmapName = cidFont.SystemInfo.ToString();
var nonUnicodeCMap = CMapCache.Get(fullCmapName);
if (nonUnicodeCMap == null)
{
return(null, true);
}
var unicodeCMapName = $"{nonUnicodeCMap.Info.Registry}-{nonUnicodeCMap.Info.Ordering}-UCS2";
return(CMapCache.Get(unicodeCMapName), true);
}
private ICidFont ParseDescendant(DictionaryToken dictionary, bool isLenientParsing)
{
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type?.Equals(NameToken.Font) != true)
{
throw new InvalidFontFormatException($"Expected \'Font\' dictionary but found \'{type}\'");
}
var result = cidFontFactory.Generate(dictionary, isLenientParsing);
return result;
}
public ICidFont Generate(DictionaryToken dictionary, bool isLenientParsing)
{
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (!NameToken.Font.Equals(type))
{
throw new InvalidFontFormatException($"Expected \'Font\' dictionary but found \'{type}\'");
}
var widths = ReadWidths(dictionary);
var verticalWritingMetrics = ReadVerticalDisplacements(dictionary);
FontDescriptor descriptor = null;
if (TryGetFontDescriptor(dictionary, out var descriptorDictionary))
{
descriptor = descriptorFactory.Generate(descriptorDictionary, isLenientParsing);
}
var fontProgram = ReadDescriptorFile(descriptor);
var baseFont = dictionary.GetNameOrDefault(NameToken.BaseFont);
var systemInfo = GetSystemInfo(dictionary);
var subType = dictionary.GetNameOrDefault(NameToken.Subtype);
if (NameToken.CidFontType0.Equals(subType))
{
//return new PDCIDFontType0(dictionary, parent);
}
if (NameToken.CidFontType2.Equals(subType))
{
return(new Type2CidFont(type, subType, baseFont, systemInfo, descriptor, fontProgram, verticalWritingMetrics, widths));
}
return(null);
}
public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers,
bool isLenientParsing)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Page) && !isLenientParsing)
{
throw new InvalidOperationException($"Page {number} had its type specified as {type} rather than 'Page'.");
}
var rotation = new PageRotationDegrees(pageTreeMembers.Rotation);
if (dictionary.TryGet(NameToken.Rotate, pdfScanner, out NumericToken rotateToken))
{
rotation = new PageRotationDegrees(rotateToken.Int);
}
MediaBox mediaBox = GetMediaBox(number, dictionary, pageTreeMembers);
CropBox cropBox = GetCropBox(dictionary, pageTreeMembers, mediaBox);
var stackDepth = 0;
while (pageTreeMembers.ParentResources.Count > 0)
{
var resource = pageTreeMembers.ParentResources.Dequeue();
resourceStore.LoadResourceDictionary(resource, isLenientParsing);
stackDepth++;
}
if (dictionary.TryGet(NameToken.Resources, pdfScanner, out DictionaryToken resources))
{
resourceStore.LoadResourceDictionary(resources, isLenientParsing);
stackDepth++;
}
UserSpaceUnit userSpaceUnit = GetUserSpaceUnits(dictionary);
PageContent content = default(PageContent);
if (!dictionary.TryGet(NameToken.Contents, out var contents))
{
// ignored for now, is it possible? check the spec...
}
else if (DirectObjectFinder.TryGet <ArrayToken>(contents, pdfScanner, out var array))
{
var bytes = new List <byte>();
for (var i = 0; i < array.Data.Count; i++)
{
var item = array.Data[i];
if (!(item is IndirectReferenceToken obj))
{
throw new PdfDocumentFormatException($"The contents contained something which was not an indirect reference: {item}.");
}
var contentStream = DirectObjectFinder.Get <StreamToken>(obj, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException($"Could not find the contents for object {obj}.");
}
bytes.AddRange(contentStream.Decode(filterProvider));
if (i < array.Data.Count - 1)
{
bytes.Add((byte)'\n');
}
}
content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, isLenientParsing);
}
else
{
var contentStream = DirectObjectFinder.Get <StreamToken>(contents, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException("Failed to parse the content for the page: " + number);
}
var bytes = contentStream.Decode(filterProvider);
content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, isLenientParsing);
}
var page = new Page(number, dictionary, mediaBox, cropBox, rotation, content,
new AnnotationProvider(pdfScanner, dictionary, isLenientParsing),
pdfScanner);
for (var i = 0; i < stackDepth; i++)
{
resourceStore.UnloadResourceDictionary();
}
return(page);
}
public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers, bool clipPaths)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Page))
{
log?.Error($"Page {number} had its type specified as {type} rather than 'Page'.");
}
MediaBox mediaBox = GetMediaBox(number, dictionary, pageTreeMembers);
CropBox cropBox = GetCropBox(dictionary, pageTreeMembers, mediaBox);
var rotation = new PageRotationDegrees(pageTreeMembers.Rotation);
if (dictionary.TryGet(NameToken.Rotate, pdfScanner, out NumericToken rotateToken))
{
rotation = new PageRotationDegrees(rotateToken.Int);
}
var stackDepth = 0;
while (pageTreeMembers.ParentResources.Count > 0)
{
var resource = pageTreeMembers.ParentResources.Dequeue();
resourceStore.LoadResourceDictionary(resource);
stackDepth++;
}
if (dictionary.TryGet(NameToken.Resources, pdfScanner, out DictionaryToken resources))
{
resourceStore.LoadResourceDictionary(resources);
stackDepth++;
}
// Apply rotation.
if (rotation.SwapsAxis)
{
mediaBox = new MediaBox(new PdfRectangle(mediaBox.Bounds.Bottom,
mediaBox.Bounds.Left,
mediaBox.Bounds.Top,
mediaBox.Bounds.Right));
cropBox = new CropBox(new PdfRectangle(cropBox.Bounds.Bottom,
cropBox.Bounds.Left,
cropBox.Bounds.Top,
cropBox.Bounds.Right));
}
UserSpaceUnit userSpaceUnit = GetUserSpaceUnits(dictionary);
PageContent content;
if (!dictionary.TryGet(NameToken.Contents, out var contents))
{
content = new PageContent(EmptyArray <IGraphicsStateOperation> .Instance,
EmptyArray <Letter> .Instance,
EmptyArray <PdfPath> .Instance,
EmptyArray <Union <XObjectContentRecord, InlineImage> > .Instance,
EmptyArray <MarkedContentElement> .Instance,
pdfScanner,
filterProvider,
resourceStore);
// ignored for now, is it possible? check the spec...
}
else if (DirectObjectFinder.TryGet <ArrayToken>(contents, pdfScanner, out var array))
{
var bytes = new List <byte>();
for (var i = 0; i < array.Data.Count; i++)
{
var item = array.Data[i];
if (!(item is IndirectReferenceToken obj))
{
throw new PdfDocumentFormatException($"The contents contained something which was not an indirect reference: {item}.");
}
var contentStream = DirectObjectFinder.Get <StreamToken>(obj, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException($"Could not find the contents for object {obj}.");
}
bytes.AddRange(contentStream.Decode(filterProvider, pdfScanner));
if (i < array.Data.Count - 1)
{
bytes.Add((byte)'\n');
}
}
content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, clipPaths, mediaBox);
}
else
{
var contentStream = DirectObjectFinder.Get <StreamToken>(contents, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException("Failed to parse the content for the page: " + number);
}
var bytes = contentStream.Decode(filterProvider, pdfScanner);
content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, clipPaths, mediaBox);
}
var page = new Page(number, dictionary, mediaBox, cropBox, rotation, content,
new AnnotationProvider(pdfScanner, dictionary),
pdfScanner);
for (var i = 0; i < stackDepth; i++)
{
resourceStore.UnloadResourceDictionary();
}
return(page);
}
public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers,
bool isLenientParsing)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Page) && !isLenientParsing)
{
throw new InvalidOperationException($"Page {number} had its type was specified as {type} rather than 'Page'.");
}
MediaBox mediaBox = GetMediaBox(number, dictionary, pageTreeMembers, isLenientParsing);
CropBox cropBox = GetCropBox(dictionary, pageTreeMembers, mediaBox);
UserSpaceUnit userSpaceUnit = GetUserSpaceUnits(dictionary);
LoadResources(dictionary, isLenientParsing);
PageContent content = default(PageContent);
if (!dictionary.TryGet(NameToken.Contents, out var contents))
{
// ignored for now, is it possible? check the spec...
}
else if (DirectObjectFinder.TryGet <ArrayToken>(contents, pdfScanner, out var array))
{
var bytes = new List <byte>();
foreach (var item in array.Data)
{
if (!(item is IndirectReferenceToken obj))
{
throw new PdfDocumentFormatException($"The contents contained something which was not an indirect reference: {item}.");
}
var contentStream = DirectObjectFinder.Get <StreamToken>(obj, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException($"Could not find the contents for object {obj}.");
}
bytes.AddRange(contentStream.Decode(filterProvider));
}
content = GetContent(bytes, cropBox, userSpaceUnit, isLenientParsing);
}
else
{
var contentStream = DirectObjectFinder.Get <StreamToken>(contents, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException("Failed to parse the content for the page: " + number);
}
var bytes = contentStream.Decode(filterProvider);
content = GetContent(bytes, cropBox, userSpaceUnit, isLenientParsing);
}
var page = new Page(number, mediaBox, cropBox, content);
return(page);
}