Research · implementation
The Quesar research export pipeline converts validated publication records into stable JSON and static HTML, verifies and copies declared attachments, and records content digests while enforcing unique identifiers, public source URLs, and canonical links.
Before export, the pipeline parses every research record through its schema, rejects duplicate slugs, and checks public citations for HTTPS URLs that are not localhost. Public PDF paths are constrained to the research download namespace.
The export projects only stable public fields into a versioned index. Internal rendering details do not become accidental API surface, and canonical publication URLs are rooted at the configured MLAI origin.
The exporter writes machine-readable JSON and static HTML, verifies each declared attachment against its expected digest before copying it, and produces a manifest containing publication metadata plus hashes for exported files. It does not generate the PDF attachments it packages.
Digest and manifest generation strengthen provenance by connecting a public artifact to a specific validated content projection. They complement, rather than replace, the revision-pinned citations attached to each research record.
73721dba44afd14a081150a020127a7dec1dbb9e
73721dba44afd14a081150a020127a7dec1dbb9e
73721dba44afd14a081150a020127a7dec1dbb9e