Text objects in example article are not tagged about accessibility HOT 12 OPEN

Although when tagging is on and the tags are visible in Acrobat, we get this error in PAC "Tagged content and artifacts" for the very exact content that is tagged.

from accessibility.

I think I have one more hint on this. When I open my minimal example in Acrobat, open the tags tab, click on a content container the correct paragraph text is highlighted with a blue frame:

However, in a normal document you would see the content. See the same pdf again after I disable accessibility and add the tags automatically in Acrobat:

Does that ring a bell?

from accessibility.

@josephreed2600 - thanks for the bug report and offering to help. I'm just getting back in to this project after some distractions and will start putting together a roadmap soon. I'll get back to you when I see how this fits and what might be required to ship something useful.

from accessibility.

I do not see any tags generated on a simple dummy file:

\documentclass{scrreprt}
\usepackage{accessibility}

\begin{document}
text.
\end{document}

Although it runs without errors or warnings in pdftex.

So it's not just an issue with the example file. It's not working at all.

from accessibility.

from accessibility.

I hadn't because I read the manual p. 5/6:

Gibt man keine Optionen an, so wird ein PDF mit den Standardoptionen erzeugt. D. h. es wird Tagged PDF mit einer geschachtelten Struktur erzeugt.

Indeed, when I use \usepackage[tagged]{accessibility} I get a PDF with a tag. I think the documentation should win here. But even this minimal example does not produce accessible pdf:

from accessibility.

Hm. So with a comparable document from another source (e.g. MS Word), does the error still occur / get flagged by PAC?

I'm interested in whether this is a problem from latex or something else.

from accessibility.

In short: no
A MS word created pdf with the content "Text." and file property title<> empty validates in PAC except from the PDF/UA metadata. No other errors.
Text.pdf
Text.docx

from accessibility.

Ok, thanks. Could you upload the word document (attach to the comment) for comparison, please? Thanks!

from accessibility.

So... looking at the MWE from@viktoriasee, I see two things:

1. In the MWE generated using latex there is a highest-level "Document" branch in the PDF that shouldn't be there.
2. In the MWE generated using latex there is no content in the <p> container.

This gives us some places to look.

from accessibility.

I agree with 2. But the «Document» master tag is fine. It's one of the few things where accessibility does a better job than Acrobat.
PDF/UA checker PAC3 complains if there is no such master tag.

from accessibility.

Yes, please keep the top-level Document tag for PDF/UA checking.

from accessibility.