The PDF/UA Standard

The PDF/UA-1 Standard for Universal Accessibility

In the early days PDF documents didn’t have a good reputation regarding basic accessibility principles. While PDF does a great job at faithfully preserving the visual appearance of a document, the logical structure of the content is generally not preserved in PDF, mostly because the format is rooted in PostScript printing technology and not in structured document formats such as XML. This shortcoming has been addressed with the introduction of Tagged PDF in PDF 1.4 (Acrobat 5, released in 2001), but adoption has been slow during the first ten years. The PDF/UA standard ISO 14289, published in 2012, is built on these structured document capabilities and further expands Tagged PDF. In the words of the standard:

»The primary purpose of ISO 14289 (known as PDF/UA) is to define how to represent electronic documents in the PDF format in a manner that allows the file to be accessible.«

Code and PDF Samples for PDF/UA-1

The PDFlib Cookbook contains sample PDF/UA documents along with the corresponding PDFlib 9 application code. The following Cookbook topics are available:

  • The clone_pdfua topic clones PDF/A, PDF/UA and PDF/X standard documents;
  • The form_fields_pdfua1topic creates various kinds of accessible form fields;
  • The invoice_pdfua1topic creates a combined PDF/UA-1 and PDF/A-2a invoice with typical invoice components;
  • The list_pdfua1topic creates an accessible list;
  • The merge_and_stamp_pdfua1 topic merges PDF/UA documents including the document structure trees. Additional text is stamped onto each page and tagged as Artifact;
  • The parallel_columns_pdfua1 topic demonstrates how to tag multiple columns in parallel with activate_item();
  • The scan_with_ocr_pdfua1 topic demonstrates tagging for invisible OCR text which accompanies a scanned page;
  • The starter_pdfua1 topic creates a PDF/UA-1 document with various content types including structure elements, artifacts, and interactive elements;
  • The table_of_contents_pdfua1topic creates a linked table of contents;
  • The table_pdfua1 topic demonstrates automatic table tagging.
  • The tag_out_of_order_pdfua1 topic demonstrates how to create Tagged PDF out of order, i.e. content creation does not follow logical reading order;
  • The textflow_pdfua1 sample creates a Tagged PDF containing a Textflow with appropriate structure elements on several pages.

PDF/UA-2

PDF/UA-2, to be defined in ISO 14289-2, is the successor of PDF/UA-1 and will be based on PDF 2.0. This means that the new PDF 2.0 tags as well as structure namespaces can be used, and that the strict PDF 2.0 tag nesting rules apply to PDF/UA-2. While PDF/UA-1 includes some requirements regarding the use of heading levels, these requirements will be dropped from PDF/UA-2. Also of interest: features which are deprecated in PDF 2.0 are not allowed in PDF/UA-2.

All navigation targets inside a document, e.g. link destinations, must be expressed with the PDF 2.0 method of structure destinations.

PDF/UA-2 introduces the concept of Additional Accessibility Declarations (AAD) which describe the document’s compliance with other accessibility specifications. AADs are stored as properties in the XMP metadata.

PDF/UA Resources

ISO 14289-1:2014: Document management applications - Electronic document file format enhancement for accessibility - Part 1: Use of ISO 32000-1 (PDF/UA-1) (note that the ISO standard document itself conforms to PDF/UA-1)

Matterhorn Protocol with PDF/UA checkpoints

PDF/UA Reference Suite with sample documents

Free PDF Accessibility Checker (PAC) for checking PDF/UA conformance

Web Content Accessibility Guidelines (WCAG) 2.0

Summary of PDF/UA File Format Requirements

PDF/UA committee Chair Duff Johnson interviews PDFlib's Managing Director Thomas Merz about PDF/UA support in PDFlib