Readable documentation

Documentation tables: escaping, captions and readable headers

A useful table explains the relationship between headings and values. Code generation should preserve those values while making the structure explicit.

Content updated: · Maintained by Hung Chi-Hsueh · corrections

Treat imported cells as text

Keep leading zeros, quoted delimiters and multiline CSV fields as text. A cell containing <script> or Markdown syntax is not permission to execute or interpret it. The HTML export escapes cell content rather than injecting imported markup.

GFM tables are a Markdown extension, not a universal format. Pipes and line breaks need deliberate handling; another renderer may display the same output differently. Inspect the published document, especially when cells contain code-like characters or multiple lines.

Label both axes in a plan comparison

For a CSV header of Plan,Seats and a record of Basic,3, enable both header options. Seats is a column header and Basic is a row header. A reader navigating to 3 can identify both the plan and the measure.

Keep Plan when it names the first column. If you leave that corner empty, the generator exports a plain empty cell. Give every data record a first-column label.

Provide a caption and meaningful headings

Use a caption to identify the table and th elements with column scope for a simple header row. A visible title outside a table does not always provide the same programmatic relationship. Avoid blank or misleading column names.

This generator targets simple data tables, not page layout or complex spanning headers. Keep wide tables in an accessible scrolling region and verify keyboard, zoom and assistive-technology behavior in the destination document. Escaped HTML alone is not complete accessibility certification.

Data provenance

Sources