import type { IDataQueryPayload, IRow, IViewField, IVisFilter } from '../../interfaces';
import type { IExplainContext } from './types';
/**
 * Shared workflow-DSL query builders for all explainers.
 *
 * Explainers MUST build queries through these helpers: the engine dedupes
 * structurally identical payloads, so two explainers using the same builder
 * automatically share one query execution (this is how the contributing-
 * dimension and unique-mark explainers split a single scan per candidate).
 */
/** aggregate result key for the per-group row count */
export declare const EXPLAIN_COUNT_KEY = "__explain_row_count";
/** derived-field key used when a quantitative candidate dimension is binned */
export declare const explainBinKey: (fid: string) => string;
export declare const QUANT_BIN_NUM = 10;
/** measure aggregators whose mass composition is meaningful for D/B class analysis */
export declare const ADDITIVE_SAFE_AGGS: Set<string>;
export declare const DEFAULT_CANDIDATE_LIMIT = 20;
/**
 * Candidate dimensions for composition-based explainers: every dimension
 * field not in the view, capped at ctx.candidateLimit (wide tables would
 * otherwise trigger one query per field). Temporal candidates are excluded
 * for now (drill granularity selection is an open question, worklog §6.3).
 */
export declare function complementaryDimensions(ctx: IExplainContext): IViewField[];
/** truncation info so the UI can say "N of M fields scanned" — no silent caps */
export declare function candidateTruncation(ctx: IExplainContext): {
    scanned: number;
    total: number;
    truncated: boolean;
};
/** equality filters pinning the selected mark, over the view dimensions */
export declare function markFilters(ctx: IExplainContext): IVisFilter[];
export interface IMarkLevelQuery {
    payload: IDataQueryPayload;
    /** the group-by key holding the candidate dimension's (possibly binned) value */
    candidateKey: string;
}
/**
 * The workhorse query shared by the contributing-dimension and unique-mark
 * explainers: group by [view dimensions..., candidate] over ALL marks,
 * fetching row counts plus per-measure sums. Sums and counts are additive,
 * so mark / siblings / totals are all derivable client-side from this one
 * result set.
 */
export declare function markLevelQuery(ctx: IExplainContext, candidate: IViewField): IMarkLevelQuery;
/**
 * Pre-screen aggregates for the selected mark: cheap distribution statistics
 * that decide whether fetching row-level data is worthwhile.
 */
export declare function markPrescreenQuery(ctx: IExplainContext, measureFids: string[]): IDataQueryPayload;
/** row-level values of the selected mark, for IQR outlier analysis */
export declare function markRawQuery(ctx: IExplainContext, fields: string[], limit: number): IDataQueryPayload;
/**
 * Marks-level aggregation at the view's level of detail, carrying the view
 * measures plus candidate (unvisualized) measures — the single query behind
 * the contributing-measure explainer.
 */
export declare function marksWithMeasuresQuery(ctx: IExplainContext, candidateMeasureFids: string[]): IDataQueryPayload;
/** does this aggregated row belong to the selected mark? */
export declare function isMarkRow(row: IRow, ctx: IExplainContext): boolean;
/** stable stringify with sorted object keys — the engine's dedup key */
export declare function stableStringify(value: unknown): string;