pdf-ocr-cli/dist/splitPdf.js

A CLI tool for OCR processing of PDF files using Mistral API with optional LLM verification

"use strict";
Object.defineProperty(exports, "__esModule", { value: true });
exports.splitPdf = splitPdf;
const pdf_lib_1 = require("pdf-lib");
/**
 * Splits a PDF buffer into individual pages
 * @param pdfBuffer - Buffer containing the PDF data
 * @param maxPages - Maximum number of pages to extract (optional)
 * @returns Array of Buffers, each containing a single page PDF
 * @throws Error if the PDF is invalid
 */
async function splitPdf(pdfBuffer, maxPages) {
    try {
        // Load the PDF document
        const pdfDoc = await pdf_lib_1.PDFDocument.load(pdfBuffer);
        // Get the total number of pages
        const pageCount = pdfDoc.getPageCount();
        // Determine how many pages to process
        const pagesToProcess = maxPages ? Math.min(pageCount, maxPages) : pageCount;
        // Create an array to store the individual page buffers
        const pageBuffers = [];
        // Process each page
        for (let i = 0; i < pagesToProcess; i++) {
            // Create a new document for this page
            const newPdfDoc = await pdf_lib_1.PDFDocument.create();
            // Copy the page from the original document
            const [copiedPage] = await newPdfDoc.copyPages(pdfDoc, [i]);
            newPdfDoc.addPage(copiedPage);
            // Save the new document to a buffer
            const newPdfBytes = await newPdfDoc.save();
            // Convert to Buffer and add to the array
            pageBuffers.push(Buffer.from(newPdfBytes));
        }
        return pageBuffers;
    }
    catch (error) {
        // Handle errors
        if (error instanceof Error) {
            throw new Error(`Invalid PDF: ${error.message}`);
        }
        else {
            throw new Error('Invalid PDF: Unknown error');
        }
    }
}