PDF/UA identifier missing

PDF/UA identifier missing is a result in PAC, the free PDF Accessibility Checker. It means the PDF's metadata doesn't declare that the file conforms to PDF/UA (ISO 14289-1), the international standard for accessible PDFs. The identifier is a short entry in the file's XMP metadata, the built-in record of facts about the document. PAC and other PDF/UA validators require it, because a file can't claim to be PDF/UA without it.

What this means

The identifier is a label, not a fix. PDF/UA-1 says a conforming file must identify itself in its XMP metadata using the PDF/UA identification schema, with a "part" value of 1. Inside the metadata, the entry looks like this:

<rdf:Description rdf:about="" xmlns:pdfuaid="http://www.aiim.org/pdfua/ns/id/">
  <pdfuaid:part>1</pdfuaid:part>
</rdf:Description>

Adding it doesn't change anything a screen reader user hears. It's a claim that the file meets the standard. That's why it should be the last thing you add:

• Add it only after the file passes every PAC check.
• Add it only after a person has done the checks no machine can, such as whether the reading order is logical and the alt text says what images mean.
• Don't add it to a file that still fails. A file that claims PDF/UA and doesn't meet it is making a false claim.

Many authoring tools don't add the identifier at all, which is reasonable: the claim should only be made once the file has been checked. That's why PAC reports it so often, even on files that are otherwise in good shape. The identifier lives in the same XMP metadata as the document title, so if a file has no XMP metadata at all, both are missing.

Why it matters

The identifier matters to people who check compliance, not to people reading the document. If your organization or a client checks documents with PAC, a file without the identifier can't get a clean report. Under WCAG 2.1, the identifier isn't required at all: it's purely a PDF/UA requirement. So treat this result as the last item on your list, not the first.

How to fix it in Acrobat Pro and re-check in PAC

Fix it in the source file

Most authoring tools don't add the identifier, and that's reasonable: adding it is a claim you should make after checking. LibreOffice is an exception. Its PDF export has a Universal Accessibility (PDF/UA) option that writes the identifier into the file. Even then, check the result in PAC and do the manual checks before you publish, because an export option can't confirm that the content itself is accessible.

For files from Word, InDesign and most other tools, plan to add the identifier at the end of remediation, after everything else passes.

Fix it automatically with Includoc

Includoc writes standard XMP metadata with the document title, and adds the PDF/UA identifier automatically, but only when every machine check passes. If anything still fails, we leave the identifier off and the results show you what's left to fix.

We're also careful about what the identifier means. Our conformance report lists which checks were machine-verified and which need a person, so the identifier is never presented as proof that a document is fully accessible.

Standards

Related errors

• Title – Failed: the title lives in the same metadata as the identifier.
• Fonts are embedded: another requirement that comes from PDF/UA rather than WCAG.

For the bigger picture, read PDF/UA explained, or check a PDF free to see everything that stands between your file and a clean PAC report.