private MediaBox GetMediaBox(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers, bool isLenientParsing)
{
MediaBox mediaBox;
if (dictionary.TryGet(NameToken.MediaBox, out var mediaboxObject) &&
DirectObjectFinder.TryGet(mediaboxObject, pdfScanner, out ArrayToken mediaboxArray))
{
if (mediaboxArray.Length != 4 && isLenientParsing)
{
log.Error($"The MediaBox was the wrong length in the dictionary: {dictionary}. Array was: {mediaboxArray}.");
mediaBox = MediaBox.A4;
return(mediaBox);
}
mediaBox = new MediaBox(mediaboxArray.ToIntRectangle());
}
else
{
mediaBox = pageTreeMembers.MediaBox;
if (mediaBox == null)
{
if (isLenientParsing)
{
mediaBox = MediaBox.A4;
}
else
{
throw new InvalidOperationException("No mediabox was present for page: " + number);
}
}
}
return(mediaBox);
}
public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers,
bool isLenientParsing)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Page) && !isLenientParsing)
{
throw new InvalidOperationException($"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 CropBox GetCropBox(DictionaryToken dictionary, PageTreeMembers pageTreeMembers, MediaBox mediaBox)
{
CropBox cropBox;
if (dictionary.TryGet(NameToken.CropBox, out var cropBoxObject) &&
DirectObjectFinder.TryGet(cropBoxObject, pdfScanner, out ArrayToken cropBoxArray))
{
if (cropBoxArray.Length != 4)
{
log.Error($"The CropBox was the wrong length in the dictionary: {dictionary}. Array was: {cropBoxArray}. Using MediaBox.");
cropBox = new CropBox(mediaBox.Bounds);
return(cropBox);
}
cropBox = new CropBox(cropBoxArray.ToIntRectangle(pdfScanner));
}
else
{
cropBox = pageTreeMembers.GetCropBox() ?? new CropBox(mediaBox.Bounds);
}
return(cropBox);
}
public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers, bool clipPaths)
{
if (dictionary == null)
{
throw new ArgumentNullException(nameof(dictionary));
}
var type = dictionary.GetNameOrDefault(NameToken.Type);
if (type != null && !type.Equals(NameToken.Page))
{
log?.Error($"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);
}
private static CropBox GetCropBox(DictionaryToken dictionary, PageTreeMembers pageTreeMembers, MediaBox mediaBox)
{
CropBox cropBox;
if (dictionary.TryGet(NameToken.CropBox, out var cropBoxObject) && cropBoxObject is ArrayToken cropBoxArray)
{
var x1 = cropBoxArray.GetNumeric(0).Int;
var y1 = cropBoxArray.GetNumeric(1).Int;
var x2 = cropBoxArray.GetNumeric(2).Int;
var y2 = cropBoxArray.GetNumeric(3).Int;
cropBox = new CropBox(new PdfRectangle(x1, y1, x2, y2));
}
else
{
cropBox = pageTreeMembers.GetCropBox() ?? new CropBox(mediaBox.Bounds);
}
return(cropBox);
}
private static CropBox GetCropBox(PdfDictionary dictionary, PageTreeMembers pageTreeMembers, MediaBox mediaBox)
{
CropBox cropBox;
if (dictionary.TryGetItemOfType(CosName.CROP_BOX, out COSArray cropBoxArray))
{
var x1 = cropBoxArray.getInt(0);
var y1 = cropBoxArray.getInt(1);
var x2 = cropBoxArray.getInt(2);
var y2 = cropBoxArray.getInt(3);
cropBox = new CropBox(new PdfRectangle(x1, y1, x2, y2));
}
else
{
cropBox = pageTreeMembers.GetCropBox() ?? new CropBox(mediaBox.Bounds);
}
return(cropBox);
}