/**
 * Table extraction from PDF pages using text fragment positioning.
 *
 * Detects tabular structures by analyzing the spatial layout of text fragments.
 * Since PDF content streams typically render tables as positioned text (with or
 * without drawn grid lines), this module uses a text-only heuristic:
 *
 * 1. Group fragments into lines by Y proximity
 * 2. Detect column boundaries from consistent X-position clusters
 * 3. Identify contiguous blocks of multi-column lines as tables
 * 4. Map fragments to cells based on column/line membership
 *
 * @see content-interpreter.ts for TextFragment extraction
 * @see text-reconstruction.ts for line grouping logic
 */
import type { TextFragment } from "./content-interpreter.js";
/**
 * A single cell in a PDF table.
 */
export interface PdfTableCell {
    /** Text content of the cell */
    text: string;
    /** X position in page coordinates (points) */
    x: number;
    /** Y position in page coordinates (points) */
    y: number;
    /** Width of the cell in points */
    width: number;
    /** Height of the cell in points */
    height: number;
    /** Number of rows this cell spans (default 1) */
    rowSpan?: number;
    /** Number of columns this cell spans (default 1) */
    colSpan?: number;
}
/**
 * A single row in a PDF table.
 */
export interface PdfTableRow {
    /** Cells in this row, ordered left-to-right */
    cells: PdfTableCell[];
}
/**
 * A table extracted from a PDF page.
 */
export interface PdfTable {
    /** Rows in this table, ordered top-to-bottom */
    rows: PdfTableRow[];
    /** X position of the table (left edge) in page coordinates */
    x: number;
    /** Y position of the table (top edge) in page coordinates */
    y: number;
    /** Width of the table in points */
    width: number;
    /** Height of the table in points */
    height: number;
}
/**
 * Extract tables from a page's text fragments.
 *
 * Uses text positioning heuristics to detect tabular structures without
 * relying on drawn lines or grid paths.
 *
 * @param fragments - Text fragments from `extractTextFromPage`
 * @param pageWidth - Page width in points
 * @param pageHeight - Page height in points
 * @returns Array of detected tables
 */
export declare function extractTables(fragments: TextFragment[], pageWidth: number, pageHeight: number): PdfTable[];