lsm/PdfPig
1
0
Fork 0
mirror of https://github.com/UglyToad/PdfPig.git synced 2026-03-10 00:23:29 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Fix related to page sizes / rotation / coordinate transformations (issue 560):

Browse Source 
The initial transformation matrix was incorrect, as it translated by the cropbox width/height
instead of by the cropbox left/bottom offsets. Also, it did not translate the results back into
the 1st quadrant so that (0,0) would (again) be the lower left corner origin for the cropped area.
Added unit tests in new file ContentStreamProcessorTests.

EFFECTIVE CHANGES:

- The coordinates used for letters etc. are different now for rotated and/or cropped pages,
  but as those were not very consistent anyway this is probably OK.

- The Page Size (A4, A3, Custom, etc.), Width and Height are now determined by the CropBox,
  not by the MediaBox; the CropBox ultimately determines what you see on screen and is printable.
  If no cropbox is defined in the PDF, it is set to the MediaBox; so in that case it is
  backwards compatible with the old code.

- The Page MediaBox and CropBox properties are no longer rotated according to Page.Rotation.
  The Page Width and Height do take rotation into account (kept it backward compatible).
This commit is contained in:
mvantzet
2023-03-09 16:42:09 +01:00
parent 3a0a6e1411
commit 0413f3f1bf
4 changed files with 225 additions and 66 deletions
Download Patch File Download Diff File Expand all files Collapse all files

103
src/UglyToad.PdfPig/Graphics/ContentStreamProcessor.cs
View File

@@ -19,6 +19,8 @@
using Util;
using XObjects;
using static PdfPig.Core.PdfSubpath;
using System.Drawing;
using System.Numerics;
internal class ContentStreamProcessor : IOperationContext
{
@@ -48,7 +50,6 @@
private readonly IPdfTokenScanner pdfScanner;
private readonly IPageContentParser pageContentParser;
private readonly ILookupFilterProvider filterProvider;
private readonly PdfVector pageSize;
private readonly InternalParsingOptions parsingOptions;
private readonly MarkedContentStack markedContentStack = new MarkedContentStack();
@@ -84,11 +85,14 @@
{XObjectType.PostScript, new List<XObjectContentRecord>()}
};
public ContentStreamProcessor(PdfRectangle cropBox, IResourceStore resourceStore, UserSpaceUnit userSpaceUnit, PageRotationDegrees rotation,
public ContentStreamProcessor(IResourceStore resourceStore,
UserSpaceUnit userSpaceUnit,
PdfRectangle mediaBox,
PdfRectangle cropBox,
PageRotationDegrees rotation,
IPdfTokenScanner pdfScanner,
IPageContentParser pageContentParser,
ILookupFilterProvider filterProvider,
PdfVector pageSize,
InternalParsingOptions parsingOptions)
{
this.resourceStore = resourceStore;
@@ -97,7 +101,6 @@
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.pageSize = pageSize;
this.parsingOptions = parsingOptions;
// initiate CurrentClippingPath to cropBox
@@ -108,7 +111,7 @@
graphicsStack.Push(new CurrentGraphicsState()
{
CurrentTransformationMatrix = GetInitialMatrix(),
CurrentTransformationMatrix = GetInitialMatrix(userSpaceUnit, mediaBox, cropBox, rotation),
CurrentClippingPath = clippingPath
});
@@ -116,63 +119,71 @@
}
[System.Diagnostics.Contracts.Pure]
private TransformationMatrix GetInitialMatrix()
internal static TransformationMatrix GetInitialMatrix(UserSpaceUnit userSpaceUnit,
PdfRectangle mediaBox,
PdfRectangle cropBox,
PageRotationDegrees rotation)
{
// TODO: this is a bit of a hack because I don't understand matrices
// TODO: use MediaBox (i.e. pageSize) or CropBox?
// Cater for scenario where the cropbox is larger than the mediabox.
// If there is no intersection (method returns null), fall back to the cropbox.
var viewBox = mediaBox.Intersect(cropBox) ?? cropBox;
/*
* There should be a single Affine Transform we can apply to any point resulting
* from a content stream operation which will rotate the point and translate it back to
* a point where the origin is in the page's lower left corner.
*
* For example this matrix represents a (clockwise) rotation and translation:
* [ cos sin tx ]
* [ -sin cos ty ]
* [ 0 0 1 ]
* Warning: rotation is counter-clockwise here
*
* The values of tx and ty are those required to move the origin back to the expected origin (lower-left).
* The corresponding values should be:
* Rotation: 0 90 180 270
* tx: 0 0 w w
* ty: 0 h h 0
*
* Where w and h are the page width and height after rotation.
*/
if (rotation.Value == 0
&& viewBox.Left == 0
&& viewBox.Bottom == 0
&& userSpaceUnit.PointMultiples == 1)
{
return TransformationMatrix.Identity;
}
double cos, sin;
double dx = 0, dy = 0;
// Move points so that (0,0) is equal to the viewbox bottom left corner.
var t1 = TransformationMatrix.GetTranslationMatrix(-viewBox.Left, -viewBox.Bottom);
// Not implemented yet: userSpaceUnit
// if (userSpaceUnit.PointMultiples != 1)
// {
// var scale = TransformationMatrix.GetScaleMatrix(userSpaceUnit.PointMultiples,
// userSpaceUnit.PointMultiples);
// ....
// }
// After rotating around the origin, our points will have negative x/y coordinates.
// Fix this by translating them by a certain dx/dy after rotation based on the viewbox.
double dx, dy;
switch (rotation.Value)
{
case 0:
cos = 1;
sin = 0;
break;
// No need to rotate / translate after rotation, just return the initial
// translation matrix.
return t1;
case 90:
cos = 0;
sin = 1;
dy = pageSize.Y;
// Move rotated points up by our (unrotated) viewbox width
dx = 0;
dy = viewBox.Width;
break;
case 180:
cos = -1;
sin = 0;
dx = pageSize.X;
dy = pageSize.Y;
// Move rotated points up/right using the (unrotated) viewbox width/height
dx = viewBox.Width;
dy = viewBox.Height;
break;
case 270:
cos = 0;
sin = -1;
dx = pageSize.X;
// Move rotated points right using the (unrotated) viewbox height
dx = viewBox.Height;
dy = 0;
break;
default:
throw new InvalidOperationException($"Invalid value for page rotation: {rotation.Value}.");
}
return new TransformationMatrix(
cos, -sin, 0,
sin, cos, 0,
dx, dy, 1);
// GetRotationMatrix uses counter clockwise angles, whereas our page rotation
// is a clockwise angle, so flip the sign.
var r = TransformationMatrix.GetRotationMatrix(-rotation.Value);
// Fix up negative coordinates after rotation
var t2 = TransformationMatrix.GetTranslationMatrix(dx, dy);
// Now get the final combined matrix T1 > R > T2
return t1.Multiply(r.Multiply(t2));
}
public PageContent Process(int pageNumberCurrent, IReadOnlyList<IGraphicsStateOperation> operations)