ppu-pdf/pdf-reader-common.d.ts

Easily extract text from digital PDF files with coordinate and font size included, and optionally group text by lines or render scanned pdf to canvas/png.

import { type Canvas } from "@napi-rs/canvas";
import { type CompactPageLines, type CompactPdfLine, type CompactPdfWord, type PageLines, type PageTexts, type PdfCompactLineAlgorithm, type PdfLine, type PdfScannedThreshold, type PdfWord } from "./pdf.interface";
export declare class PdfReaderCommon {
    saveCanvasToPng(canvas: Canvas, filename: string, foldername: string): Promise<void>;
    protected dumpCanvasMapCommon(canvasMap: Map<number, Canvas>, filename: string, foldername?: string, startIndex?: number): Promise<void>;
    protected sortTextContent(texts: PdfWord[]): PdfWord[];
    protected sortTextContentSimple(texts: PdfWord[]): PdfWord[];
    protected getLinesFromTextsCommon(pageTexts: PageTexts, startIndex?: number): PageLines;
    protected getLines(words?: PdfWord[]): PdfLine[];
    protected mergeLines(lines: PdfWord[][]): PdfLine[];
    protected getCompactLines(words?: CompactPdfWord[]): CompactPdfLine[];
    protected mergeCompactLines(lines: CompactPdfWord[][]): CompactPdfLine[];
    protected getCompactLinesOldAlgorithm(words?: CompactPdfWord[]): CompactPdfLine[];
    protected mergeCompactLinesOldAlgorithm(lines: CompactPdfWord[][]): CompactPdfLine[];
    getCompactLinesFromTextsCommon(pageTexts: PageTexts, algorithm?: PdfCompactLineAlgorithm, startIndex?: number): CompactPageLines;
    protected mapWordsToCompactWords(words?: PdfWord[]): CompactPdfWord[];
    protected isScannedCommon(pageTexts: PageTexts, options?: PdfScannedThreshold, startIndex?: number): boolean;
    protected normalizedText(str: string): string;
}