WeasyPrint 57.1 "Hello World"HTML targeted to PDF/A-1b produces output that fails PDF/A-1b validation - "file trailer dictionary shall contain the ID keyword"

[paulbarron1]

Hi.
I was excited to see the latest versions of WeasyPrint support option to produce PDF output for a variety of PDF/A standards.

However, I can't get the resulting PDF/A output to validate successfully, and have tried several PDF/A validators with the same validation error reported.

The execution trace reports that WeasyPrint's PDF/A feature option is an experimental new feature and asks that any issues be reported thus ;
WARNING: PDF/A support is experimental, generated PDF files are not guaranteed to be valid. Please open an issue if you have problems generating PDF/A files.

I have tried several inputs , all with the same validation error reported , including on a "Hello World" HTML input.
hello-world-htmlinput-pdfa1b-output.zip

Input HTML and PDF/A-1b output in referenced file"hello-world-htmlinput-pdfa1b-output.zip"

I used WeasyPrint 57.1 , requesting PDF/A-1b thus ;
weasyprint.exe --pdf-variant pdf/a-1b hello-world.html hello-world.pdf

Resulting PDF was interpreted correctly as PDF/A-1b , but failed validation on PDF/A Validators
https://demo.verapdf.org/
https://tools.pdfforge.org/en/validate-pdfa

Both reported the same PDF/A-1b validation error
Specification ISO 19005-1:2005
Clause 6.1.3
Description
The file trailer dictionary shall contain the ID keyword. The file trailer referred to is either the last trailer dictionary in a PDF file, as described in PDF Reference 3.4.4 and 3.4.5, or the first page trailer in a linearized PDF file, as described in PDF Reference F.2

[liZe]

Hi!

Thanks for opening the issue.

You can add an identifier using the --pdf-identifier option, this should fix your problem!

We know that we should definitely improve WeasyPrint to check many things that are mandatory/forbidden with PDF/A, and set missing required values (like the ID) if needed. We can discuss that in #1661.

[paulbarron1]

Hi liZe

Many thanks for your quick reply.

I can confirm that adding weasyprint option

--pdf-identifier NameOfIdentifier

..to the --pdf-variant pdf/a-1b request, does indeed produce PDF/A1-b compliant output.

I must admit I hadn't linked the two options , but can see that this should be covered by existing "set missing required PDF/A values" on #1661.

Thanks again for your help.