It seems that the embedded PDF images are converted to JPEG by Writer 6.1.0. 6.0.6 works fine. My OS is Windows 10.
Update 1: After some investigation, I guess it’s the way v6.1.0 saves a .odt file that contains PDF images.
- Create a document in Writer, insert a PDF into it, and save it as “saved_by_6.1.0.odt”.
- Try to export it to PDF. This hangs Writer on my machine.
- Open “saved_by_6.1.0.odt” with v6.0.6 on another machine, and save it as “resaved_by_6.0.6.odt”.
- Try to export it to PDF. This works flawlessly with both v6.0.6 and v6.1.0.
Hope this can help track down the bug. I really like the fresh new look of v6.1.0.
Update 2: Turns out that v6.1.1 has fixed this problem.
- If the document contains a v1.4 PDF image, the generated PDF is perfect. Since v6.0.6 works all OK, I think this is a regression in the code of down-converting v1.5+ PDF to v1.4 using pdfium.
- If the document contains a v1.5 PDF image, the generated PDF is corrupted (Acrobat saying: Critical parser failure: Extended graphic state resource missing).
- If the document contains a v1.5 PDF image and we export it to PDF with reference XObjects enabled, the generated PDF is valid but with a blurred (raster) image. I think this is another regression.
Changing the PDF version is boring, but at least a way out.
Updated 3: Filed a bug report here: https://bugs.documentfoundation.org/show_bug.cgi?id=119880