Remediation of Complex Tables in PDF

Scheduled at 11:30 am in Colorado F on Friday, November 15.

#39775

Speaker(s)

  • William Kilian, Software Architect, PDF/UA, TargetStream Technologies

Session Details

  • Length of Session: 1-hr
  • Format: Lecture
  • Expertise Level: Intermediate
  • Type of session: General Conference

Summary

PDF documents can contain tables that span hundreds of table, use complex visual layouts, or that exhibit both properties. This presentation will show what tags, attributes, and other settings are necessary when remediating documents.

Abstract

In order for Assistive Technology to accurately present complex tables, PDF documents generally need to properly set ID values in tags for table header cells and specify those ID values on the appropriate data cells. When remediating documents for PDF output, the first task is recognizing when this extra work is necessary. The PDF 2.0 specification outlines an algorithm that PDF viewers should use for identifying header cells for a given table cell. If that algorithm does not yield the correct header cells, then setting IDs is necessary. The algorithm will be explained.

If ID values are necessary to identify header cells, correctly setting those values is the next task. This task is so difficult that best practice guides and examples on standards websites are sometimes incorrect. Common errors, misunderstandings, and the respective corrections will be shown.

Tables in PDF documents can span hundreds or thousands of pages. Such long tables can actually represent multiple sections of a document. Although most document technologies do not allow for headings inside tables, PDF does. Because Word and HTML do not support document headings within tables, either special software or manual tagging is necessary to achieve the proper structure.

Although some software provides direct support for complex or extremely long tables, the presentation will be for users of any tool. Thus, only manual tagging procedures and proper tagging results will be shown.

Keypoints

  1. Remediators need to know how to recognize simple vs complex tables in PDF.
  2. Complex tables require explicit settings to associate data cells with their header cells.
  3. Long tables in PDF can have document headings inside the table.

Disability Areas

Vision

Topic Areas

Alternate Format, Assistive Technology, Uncategorized, Web/Media/App Access

Speaker Bio(s)

William Kilian

William Kilian is a Software Architect at TargetStream Technologies. He has over a decade of experience as TargetStream's senior Software Architect for PDF/UA. In that role, he has created solutions for PDF/UA remediation. He works with clients large and small all around the world to help them remediate documents with complex dynamic layouts to create accessible PDF/UA output.