public ContentStreamProcessor(PdfRectangle cropBox, IResourceStore resourceStore, UserSpaceUnit userSpaceUnit, PageRotationDegrees rotation,
IPdfTokenScanner pdfScanner,
IPageContentParser pageContentParser,
IFilterProvider filterProvider,
ILog log,
bool clipPaths)
{
this.resourceStore = resourceStore;
this.userSpaceUnit = userSpaceUnit;
this.rotation = rotation;
this.pdfScanner = pdfScanner ?? throw new ArgumentNullException(nameof(pdfScanner));
this.pageContentParser = pageContentParser ?? throw new ArgumentNullException(nameof(pageContentParser));
this.filterProvider = filterProvider ?? throw new ArgumentNullException(nameof(filterProvider));
this.log = log;
this.clipPaths = clipPaths;
// initiate CurrentClippingPath to cropBox
var clippingSubpath = new PdfSubpath();
clippingSubpath.Rectangle(cropBox.BottomLeft.X, cropBox.BottomLeft.Y, cropBox.Width, cropBox.Height);
var clippingPath = new PdfPath()
{
clippingSubpath
};
clippingPath.SetClipping(FillingRule.NonZeroWinding);
graphicsStack.Push(new CurrentGraphicsState()
{
CurrentClippingPath = clippingPath
});
ColorSpaceContext = new ColorSpaceContext(GetCurrentState, resourceStore);
}
public ContentStreamProcessor(PdfRectangle cropBox, IResourceStore resourceStore, UserSpaceUnit userSpaceUnit, bool isLenientParsing)
{
this.resourceStore = resourceStore;
this.userSpaceUnit = userSpaceUnit;
this.isLenientParsing = isLenientParsing;
graphicsStack.Push(new CurrentGraphicsState());
}
private static UserSpaceUnit GetUserSpaceUnits(DictionaryToken dictionary)
{
var spaceUnits = UserSpaceUnit.Default;
if (dictionary.TryGet(NameToken.UserUnit, out var userUnitBase) && userUnitBase is NumericToken userUnitNumber)
{
spaceUnits = new UserSpaceUnit(userUnitNumber.Int);
}
return(spaceUnits);
}
private static UserSpaceUnit GetUserSpaceUnits(PdfDictionary dictionary)
{
var spaceUnits = UserSpaceUnit.Default;
if (dictionary.TryGetValue(CosName.USER_UNIT, out var userUnitCosBase) && userUnitCosBase is ICosNumber userUnitNumber)
{
spaceUnits = new UserSpaceUnit(userUnitNumber.AsInt());
}
return(spaceUnits);
}
public ContentStreamProcessor(PdfRectangle cropBox, IResourceStore resourceStore, UserSpaceUnit userSpaceUnit, bool isLenientParsing,
IPdfTokenScanner pdfScanner,
XObjectFactory xObjectFactory)
{
this.resourceStore = resourceStore;
this.userSpaceUnit = userSpaceUnit;
this.isLenientParsing = isLenientParsing;
this.pdfScanner = pdfScanner;
this.xObjectFactory = xObjectFactory;
graphicsStack.Push(new CurrentGraphicsState());
}
public Page Create(int number, PdfDictionary dictionary, PageTreeMembers pageTreeMembers, IRandomAccessRead reader,
bool isLenientParsing)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetName(CosName.TYPE);
if (type != null && !type.Equals(CosName.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, reader, isLenientParsing);
PageContent content = default(PageContent);
var contentObject = dictionary.GetItemOrDefault(CosName.CONTENTS) as CosObject;
if (contentObject != null)
{
var contentStream = pdfObjectParser.Parse(contentObject.ToIndirectReference(), reader, false) as PdfRawStream;
if (contentStream == null)
{
throw new InvalidOperationException("Failed to parse the content for the page: " + number);
}
var contents = contentStream.Decode(filterProvider);
var txt = OtherEncodings.BytesAsLatin1String(contents);
var operations = pageContentParser.Parse(new ByteArrayInputBytes(contents));
var context = new ContentStreamProcessor(cropBox.Bounds, resourceStore, userSpaceUnit);
content = context.Process(operations);
}
var page = new Page(number, mediaBox, cropBox, content);
return(page);
}
public ContentStreamProcessor(PdfRectangle cropBox, IResourceStore resourceStore, UserSpaceUnit userSpaceUnit, PageRotationDegrees rotation, bool isLenientParsing,
IPdfTokenScanner pdfScanner,
XObjectFactory xObjectFactory,
ILog log)
{
this.resourceStore = resourceStore;
this.userSpaceUnit = userSpaceUnit;
this.rotation = rotation;
this.isLenientParsing = isLenientParsing;
this.pdfScanner = pdfScanner;
this.xObjectFactory = xObjectFactory;
this.log = log;
graphicsStack.Push(new CurrentGraphicsState());
ColorSpaceContext = new ColorSpaceContext(GetCurrentState);
}
public ContentStreamProcessor(PdfRectangle cropBox, IResourceStore resourceStore, UserSpaceUnit userSpaceUnit, PageRotationDegrees rotation,
IPdfTokenScanner pdfScanner,
IPageContentParser pageContentParser,
IFilterProvider filterProvider,
ILog log)
{
this.resourceStore = resourceStore;
this.userSpaceUnit = userSpaceUnit;
this.rotation = rotation;
this.pdfScanner = pdfScanner ?? throw new ArgumentNullException(nameof(pdfScanner));
this.pageContentParser = pageContentParser ?? throw new ArgumentNullException(nameof(pageContentParser));
this.filterProvider = filterProvider ?? throw new ArgumentNullException(nameof(filterProvider));
this.log = log;
graphicsStack.Push(new CurrentGraphicsState());
ColorSpaceContext = new ColorSpaceContext(GetCurrentState, resourceStore);
}
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);
}
private PageContent GetContent(int pageNumber, IReadOnlyList <byte> contentBytes, CropBox cropBox, UserSpaceUnit userSpaceUnit,
PageRotationDegrees rotation,
bool isLenientParsing)
{
var operations = pageContentParser.Parse(pageNumber, new ByteArrayInputBytes(contentBytes),
log);
var context = new ContentStreamProcessor(cropBox.Bounds, resourceStore, userSpaceUnit, rotation, isLenientParsing, pdfScanner,
pageContentParser,
filterProvider,
log);
return(context.Process(pageNumber, operations));
}
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($"Could not find the contents for object {obj}.");
}
bytes.AddRange(contentStream.Decode(filterProvider, pdfScanner));
if (i < array.Data.Count - 1)
{
bytes.Add((byte)'\n');
}
}
content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, clipPaths, mediaBox);
}
else
{
var contentStream = DirectObjectFinder.Get <StreamToken>(contents, pdfScanner);
if (contentStream == null)
{
throw new InvalidOperationException("Failed to parse the content for the page: " + number);
}
var bytes = contentStream.Decode(filterProvider, pdfScanner);
content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, clipPaths, mediaBox);
}
var page = new Page(number, dictionary, mediaBox, cropBox, rotation, content,
new AnnotationProvider(pdfScanner, dictionary),
pdfScanner);
for (var i = 0; i < stackDepth; i++)
{
resourceStore.UnloadResourceDictionary();
}
return(page);
}
private PageContent GetContent(int pageNumber, IReadOnlyList <byte> contentBytes, CropBox cropBox, UserSpaceUnit userSpaceUnit,
PageRotationDegrees rotation, bool clipPaths, MediaBox mediaBox)
{
var operations = pageContentParser.Parse(pageNumber, new ByteArrayInputBytes(contentBytes),
log);
var context = new ContentStreamProcessor(cropBox.Bounds, resourceStore, userSpaceUnit, rotation, pdfScanner,
pageContentParser,
filterProvider,
log,
clipPaths,
new PdfVector(mediaBox.Bounds.Width, mediaBox.Bounds.Height));
return(context.Process(pageNumber, operations));
}
public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers,
bool isLenientParsing)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Page) && !isLenientParsing)
{
throw new InvalidOperationException($"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);
}
private PageContent GetContent(IReadOnlyList <byte> contentBytes, CropBox cropBox, UserSpaceUnit userSpaceUnit, bool isLenientParsing)
{
var operations = pageContentParser.Parse(new ByteArrayInputBytes(contentBytes));
var context = new ContentStreamProcessor(cropBox.Bounds, resourceStore, userSpaceUnit, isLenientParsing, pdfScanner, xObjectFactory);
return(context.Process(operations));
}
private PageContent GetContent(IReadOnlyList <byte> contentBytes, CropBox cropBox, UserSpaceUnit userSpaceUnit, bool isLenientParsing)
{
if (Debugger.IsAttached)
{
var txt = OtherEncodings.BytesAsLatin1String(contentBytes.ToArray());
}
var operations = pageContentParser.Parse(new ByteArrayInputBytes(contentBytes));
var context = new ContentStreamProcessor(cropBox.Bounds, resourceStore, userSpaceUnit, isLenientParsing);
return(context.Process(operations));
}
