Documentation
¶
Overview ¶
Package vault is the normalized model an importer produces: a set of pages with a parent tree and wikilinks preserved inline in the body. It is pure (no cloud, no framework, no I/O) so every format normalizer — obsidian, notion, roam, evernote — maps to ONE shape that the KB import handler files as kb-page documents through the SAME framework.Ingest path a connector sync uses.
A normalizer never writes anything: it turns raw export bytes into []Page, and the KB lane files them. Wikilinks stay as literal "[[Title]]" text in Body, so the kb-page after_save hook extracts them into kb-link edges exactly as it does for a hand-authored page — the import path adds no second link-extraction path.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func Slug ¶
Slug turns a page title/name into a URL-safe kb-page slug: lowercase, with every run of non-alphanumeric characters folded to a single dash and the ends trimmed. It is the ONE slug rule every normalizer shares, so the same title yields the same slug across formats. An empty result falls back to "page" (the import handler then makes it unique within the org).
Types ¶
type File ¶
File is one entry from an unpacked archive: its archive path and raw bytes. Normalizers take a []File so they stay pure — the KB import handler owns the zip decode — and are unit-testable with a hand-built list, no real archive needed.
type Page ¶
Page is a normalized knowledge page ready to file as a kb-page. Body is a Lexical EditorState JSON string (built via clients/knowledge/lexical) with wikilinks preserved as "[[Title]]" text. Parent is the slug of the parent page, or "" for a top-level page — the import handler files parents before children so the kb-page `parent` Link resolves.
Source Files