Skip to content

Diffusion

See also: Diffusion (system overview) · Exporting data · Glossary: diffusion

Diffusion is how catalogued work becomes published data. Dédalo keeps two database systems — the private work system a cataloguer edits, and the diffusion system a website or an integration reads — and the diffusion engine is the ontology-driven pipeline that moves a selected subset from one to the other, transforming it on the way. Work data and public data therefore never share a store, and publishing is never a read of the editing back-end.

The pages

  • The diffusion engine — the pipeline itself: how a publication is described in the diffusion ontology, what the engine produces (SQL tables, RDF / XML / Markdown / CSV / JSON files), and how tool_diffusion drives it. Start here.
  • Diffusion data flow — the conceptual half: which data leaves the work system, what happens to it in transit, and why the two systems stay separate.
  • Parser cookbook & reference — the transformation steps an institution writes per publishable field: every parser, what it takes and what it emits.
  • Markdown diffusion — the one-file-per-record Markdown output, written so that AI agents and humans can read a record comfortably.
  • Publication API — the read-only service that serves the published databases to websites, integrations and agents.

Where the control surface is

Publication is operated, not scripted: tool_diffusion issues the runs and shows their progress, and the engine's own actions (diffuse, validate, get_process_status, cancel_process, the media-index rebuild) are the dd_diffusion_api class of the JSON API.