Details
-
Type:
Bug
-
Status: Closed
-
Priority:
Major
-
Resolution: Fixed
-
Affects Version/s: 4.1.2
-
Fix Version/s: 4.1.3
-
Component/s: Core/Parsing
-
Labels:None
-
Environment:any
Description
A client sent in a PDF document that had an image that should have had some content clipped from an image. After closer inspection of the PDF it turned out that the image being rendered was in an Xobject. When an Xobject is painted we must first convert an xobject's bbox into the correct coordinate space and then create a new clip that is the intersection of the current clip with the bbox clip. Closer inspection of the code that does this shows that we where getting the current clips bounds instead of the clips actual shape. In the PDF in question the clips is actually pretty complex shape and the bounds incorrectly define what actually needs to be clipped.
Activity
- All
- Comments
- History
- Activity
- Remote Attachments
- Subversion
Fixed a xobject clipping issue where the clips bounds where used instead of the clips actual shape. The bounds method is faster then using the clip shape but the bounds but there is a risk of not applying the correct clip.