One particular PDF has a jbig2 compressed image as an xobject that fills an entire page. I get, approximately, the first third of the image but the rest of it is chopped. Acrobat seems to do ok with rendering and extracting the image but Ghostscript does not. I have tried release 8.63 and the *bleeding edge* build and I see no difference.
Created attachment 4784 [details] PDF page with jbig2 xobject
Created attachment 4785 [details] converted tiffg4 lacking the last part of the image
Created attachment 4786 [details] foo.pdf - simplified sample file
I confirm that the file has a single image and Ghostscript shows only a part of it. My simplified file has only 7 objects and makes this obvious. Reassigning to the owner of JBIG2 decoder.
Hi Nathan. You have a knack with finding these files! Looks a bug in our arith decoder or symbol dictionary parser.
Don't blame me Ralph, I'm only the messenger. I have an infinite supply of crappy PDF files at my disposal and they keep on coming too! N.
Accidently marked fixed when fixing another bug
Hello folks, is there any news related to this issue? Maybe you could give me a hint where to look in the sources, so that I can try to provide a fix. Thanks, Benjamin
Bug still reproducible in Ghostscript 9.03
Reportedly this is fixed in the latest code, if not please reopen.