DOCX accessibility testing API for publishing workflows

The DOCX accessibility testing API preflights Word documents before they enter a portal, knowledge base, or conversion pipeline. Direct OOXML inspection points structural findings back to document parts and source concepts.

For document platforms and teams publishing Word files at repeatable volume

Illustrative DOCX findingIllustrative
Table does not identify a header rowmajor · deterministic
Part
word/document.xml
Locator
table 3 · row 1
Observed
tblHeader property absent
Repair
Repeat Header Rows in Word
Source
Safe OOXML package inspection
Checks
Styles · language · images · tables · links
Page count
Estimated or rendered method stays explicit

Inspect the structures authors actually edit.

The DOCX worker treats Word semantics as the primary source and uses rendering only as supplementary evidence.

Document properties
Title, language, and package metadata
Paragraph styles
Semantic headings and outline structure
Lists and tables
Native markup, headers, and complexity risks
Images
Drawing relationships and alternative-text fields
Hyperlinks
Text purpose and relationship targets
Layout risks
Floating objects, text boxes, and merged cells

Choose the technical profile behind this capability.

DOCX structure can support WCAG and document-policy evidence, but the chosen profile determines the exact mappings and open human checks.

Keep DOCX findings connected to the authoring structure.

Each evidence layer narrows the route from normalized result back to the OOXML part an authoring system can change.

  1. Result

    Normalized finding

    Severity, confidence, profile, and repair guidance

  2. Part

    OOXML relationship

    Document part, node path, and linked resource

  3. Author

    Source-facing repair

    Paragraph, style, table, image, or property

Do not turn a Word page estimate into false precision.

DOCX pagination changes with fonts, renderer, and layout environment. Reports preserve whether page count was estimated or rendered.

Stable source evidence

  • OOXML part and structural node
  • Style, relationship, and property state
  • Document content hash
  • Pinned parser and rule versions

Context to review

  • Renderer and available fonts
  • Estimated versus rendered page count
  • Visual reading order and reflow
  • Meaning of alternatives and link language

Catch structural DOCX risks before the file leaves the workflow.

Scan one Word document for free and send its source-aware findings back to the author before the next export or handoff.

Check a website freeCreate an account

Automated preflight is not an accessibility certification or legal determination.