Skip to content

Interface: DiscoveredSortDto

@kortexya/reasoninglayer


@kortexya/reasoninglayer / DiscoveredSortDto

Defined in: src/types/sources.ts:166

A sort/type discovered from a data source.

confidence: number

Defined in: src/types/sources.ts:168

Confidence score (0.0 - 1.0).


optional description?: string | null

Defined in: src/types/sources.ts:170

Human-readable description.


features: DiscoveredFeatureDto[]

Defined in: src/types/sources.ts:172

Features/fields discovered.


name: string

Defined in: src/types/sources.ts:186

The type’s identity in the source, verbatim: the CSV file stem, the SQL table, the Mongo collection, the Neo4j label.

This is the string that must go back to the source — the typeFilter of an ingest, the tableName of a binding. It is not the name the sort will be created under; read sortName for that.

Identical to sourceTypeKey, which is the unambiguous name for it. Retained so a caller that has not migrated keeps addressing the source correctly rather than silently selecting zero rows.


sortName: string

Defined in: src/types/sources.ts:203

The sort name this type will be created under — derived from sourceTypeKey so the RLQL parser can read it bare.

Usually identical to the source key (customer → customer); it differs exactly when the source names a type with a character the grammar does not admit (demo-lab-results → demo_lab_results). Show this to the operator as the sort that will be created, and send this — not name — to bulk sort creation, which refuses a name it cannot read.


sourceName: string

Defined in: src/types/sources.ts:205

Source name where this was discovered.


sourceTypeKey: string

Defined in: src/types/sources.ts:191

The type’s identity in the source. The unambiguous name for name; prefer it in new code.


optional suggestedParents?: string[]

Defined in: src/types/sources.ts:207

Suggested parent sorts.