Documentation
¶
Overview ¶
@index Pure search-document content construction from graph nodes and annotations.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func BuildContent ¶
BuildContent assembles the text indexed for one node's search document. @intent combine symbol names, path/language tokens, and annotation evidence without persistence concerns.
func BuildIntentContent ¶ added in v0.13.0
BuildIntentContent assembles the text indexed for one node's recorded reason.
This is deliberately not BuildContent with a filter. BuildContent answers "what is this called", so it mixes the identifier, its split subtokens, and the path into the same text as the annotation prose. A question like "why do we verify the webhook signature" then competes against every node whose *name* happens to contain a query word. Keeping the reason in its own index is what lets that question be scored on the reason alone.
Only @intent and @domainRule are taken. Both answer why the code exists; the remaining tags describe what it does or what it takes, which is the name index's job. @intent index the reason a node exists separately from what the node is called.
Types ¶
type Maintenance ¶
type Maintenance interface {
RefreshDocuments(ctx context.Context) (int, error)
RebuildIndex(ctx context.Context) error
}
Maintenance refreshes persisted search documents and the bound full-text index. @intent let inbound orchestration trigger one complete search rebuild without receiving database/backend handles.
Source Files