Horst Rutter

Adobe Acrobat possibly but this is not smth that will help here because the tool will write back the xreftable and the resulting file will probably validate fine afterwards.. -...

I added a useful log message. If you rerun doc5.pdf with the latest commit we may know more what's going on here.

``` Fatal: dict=rootDict required entry=Type missing github.com/pdfcpu/pdfcpu/pkg/pdfcpu.Dict.Entry ``` We can fix this, but before just so we know how much further this gets us during validation can you do me...

Have you run the command, I think it is hard to miss with all that output

Grep on: ParseObject: begin, obj#63

There must be smth like: ``` READ: 2021/12/12 23:25:08 dereferenceObject: begin, dereferencing object 63 READ: 2021/12/12 23:25:08 in use object 63 READ: 2021/12/12 23:25:08 dereferenceObject: dereferencing object 63 READ: 2021/12/12...

Yes the grepping subject are READ log lines, but they are followed by the actual parsing so this definitely is helping. As soon as you have log lines prefixed with...

No, that's ok actually. The catalog is a dictionary and will be pretty logged generally, in this case though we fail during parsing. According to the hexdump the catalog looks...

Hi! Resizing is not yet part of pdfcpu but it sounds like a useful feature. What's your preferred approach to handling existing page orientation in the output PDF?

OK. I see 3 variants for resizing and here the pdfcpu papersize constants including orientation prove useful: For example: 1) "A4": resize to A4 and keeps the current page orientation...