LaTeX Tagging Project

Using the LaTeX prototype for accessible PDF (as of 2024/06)

The new code can be used with pdfLaTeX or the Unicode engine luaLaTeX. The latter is the preferred engine recommended for new documents.

The interface descriptions below are all temporary and will be replaced with proper interfaces over time. E.g., the \DocumentMetadata will remain, but the key testphase and its value will eventually vanish.

Basic setup

To enable tagging in supported documents it is enough to add at the very beginning of the document, i.e., before \documentclass, a \DocumentMetadata declaration. It takes one argument, a key-value list:

\DocumentMetadata{
  lang        = de,
  pdfversion  = 2.0,
  pdfstandard = ua-2,
  pdfstandard = a-4f, %or a-4
  testphase   = 
   {phase-III,
    title,
    table,
    math,
    firstaid}  
}
\documentclass{....}

The first four keys in the example set important document metadata, like the language, the requested PDF version and the standards the document should comply with.

Tagging is then enabled by loading various modules through the testphase key.

The phase-III module activates tagging and loads support for most standard document elements like paragraphs, lists, sectioning, table of contents, graphics, floats, links and more. You should always use this key first.
The title module redefines the \title, \author and \maketitle command to make them tagging aware. It also takes care that the title and author are added to the XMP metadata.
The table module adds tagging support to tables like tabular, tabularx, tabulary and longtable.
The math module enables basic tagging of equations.
Finally, the firstaid contains small fixes of external packages to make them compatible with the tagging code, e.g., it fixes clashes with the cleveref and the booktabs packages.

Handling graphics in the document

While most tagging is done automatically by adding the \DocumentMetadata command, there remain a few things the author has to add manually. At first every graphic must be marked either as artifact or get an alternative text describing its meaning in the context of the document. This can be done with additional keys that have been added to the \includegraphics command:

\includegraphics[alt={this shows a yellow duck}]{duckimage}
\includegraphics[artifact]{dekoration}

Handling tables (`tabular`) in the document

If the document contains data tables, the author has to identify the header rows of the tables so that they can be tagged as <TH> cells. By default all cells are considered data cells. This can be done

\tagpdfsetup{table/header-rows={1,2}}

generally (in the preamble) or on a table by table basis before the table.

If a table should not be tagged as table, for example, because it is merely used to ensure that the content is properly aligned, it should be turned into a presentation table with

\tagpdfsetup{table/tagging=presentation}

Extended math support

To improve the accessibility of math equations the code tries to embed MathML representations of all equations. These MathML representations are currently not created automatically but must be provided by the author.

For this the document should be compiled once with the command

\tagpdfsetup{math/mathml/write-dummy}

This will write a file <file>-mathml-dummy.html which has a prepared section for every equation found in the document. The section shows the LaTeX-source and a hash value and an empty <math></math> tag.

<div>
<h2>\mml 1</h2>
<p>$x=1$</p>
<p>8FD5BB0EEAA8887F6A312C99359A3B93</p>
<math></math>
</div>