WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: PDFlib and Link tags


From: Duff Johnson
Date: Oct 1, 2021 9:14AM

> I've read that the PDFlib product cannot generate Link tags (although it does generate Link Annotations).

This is incorrect. (NOTE: I do not work for PDFlib; these opinions are my own.)

> Their websites sample pdf file "link_annotations.pdf" has no tags at all, remarkably. (I checked in Acrobat and in PAC3.)
> Yet both JAWS and NVDA see the links, can navigate to them and activate them.

That example file derives from a section in the PDFlib Cookbook pertaining to untagged documents, so this is perfectly normal and expected.

As with other such software, PDFlib is a toolbox which does many things; if you use it to create untagged PDF it will do so.

The company also provides accessible examples in the "PDF/UA" category of the PDFlib Cookbook, including examples with links such as:

https://www.pdflib.com/pdflib-cookbook/pdfua/image_with_link_pdfua1/ <https://www.pdflib.com/pdflib-cookbook/pdfua/image_with_link_pdfua1/>

https://www.pdflib.com/pdflib-cookbook/pdfua/starter_pdfua1/ <https://www.pdflib.com/pdflib-cookbook/pdfua/starter_pdfua1/>

https://www.pdflib.com/pdflib-cookbook/pdfua/table_of_contents_pdfua1/ <https://www.pdflib.com/pdflib-cookbook/pdfua/table_of_contents_pdfua1/>

> Anyone who can explain why the screenreaders seem unfazed by the absence of tags in general and Link tags in particular?

The links exist in the file whether they are tagged or not. Just as the text is often also (mostly) readable by screen-readers irrespective of tagging, so to the links.

The tags provide structure and semantics that the raw content (text and links) cannot themselves provide.