Skip to content
Keryx/ docs

Publish a PDF

keryx publish converts a stored draft version to PDF. It takes a draft ID, not a path to arbitrary HTML.

Terminal window
keryx publish --id <draft-id> --output ./report.pdf

Without --version, Keryx uses the latest version. Pin a version for a fixed deliverable:

Terminal window
keryx publish --id <draft-id> --version 2 --output ./report-v2.pdf

The server renders the PDF and returns its bytes without storing a PDF or creating a new document version. The CLI writes the result atomically and refuses to overwrite an existing output file.

Structure the HTML for print

Section titled “Structure the HTML for print”
  • Use one body-level header for the cover.
  • Put the document’s main sections in top-level section elements. Each begins on a new page.
  • Add class="keryx-page-flow" to a section that should continue in normal flow.
  • Use class="keryx-page-break" to start another element on a new page.
  • Add data-keryx-print="stack" to a custom multi-column layout that should become one column on A4.

Keryx adds a title/version header and publication-date/page footer to a render-only copy. An empty element with data-keryx-version is filled with the selected version in that copy. The original HTML stays unchanged.

Supported content

Section titled “Supported content”

Semantic HTML, paginated tables, inline SVG diagrams and base64-embedded PNG, JPEG and GIF images are supported. Script-generated content, canvas, CSS imports, external images and CSS image URLs are rejected for deterministic output.

A document that works in a browser may need changes for PDF publication. Keep reading order meaningful and embed images rather than depending on remote resources.