docuglean-ocr/dist/utils.js

An SDK for intelligent document processing using State of the Art AI models.

"use strict";
var __importDefault = (this && this.__importDefault) || function (mod) {
    return (mod && mod.__esModule) ? mod : { "default": mod };
};
Object.defineProperty(exports, "__esModule", { value: true });
exports.isURL = isURL;
exports.getFileType = getFileType;
exports.encodePdf = encodePdf;
exports.encodeImage = encodeImage;
exports.handleMistralOCRResponse = handleMistralOCRResponse;
exports.getSignedMistralUrl = getSignedMistralUrl;
const fs_1 = __importDefault(require("fs"));
const path_1 = __importDefault(require("path"));
const mistralai_1 = require("@mistralai/mistralai");
function isURL(filePath) {
    return filePath.startsWith('http://') || filePath.startsWith('https://');
}
function getFileType(filePath) {
    const extension = path_1.default.extname(filePath).toLowerCase();
    switch (extension) {
        case '.jpg':
        case '.jpeg':
        case '.png':
        case '.gif':
        case '.webp':
            return 'image';
        case '.pdf':
            return 'pdf';
        default:
            return 'unknown';
    }
}
async function encodePdf(filePath) {
    try {
        // Read the PDF file as a buffer
        const pdfBuffer = fs_1.default.readFileSync(filePath);
        // Convert the buffer to a Base64-encoded string
        const base64Pdf = pdfBuffer.toString('base64');
        return `data:application/pdf;base64,${base64Pdf}`;
    }
    catch (error) {
        if (error instanceof Error) {
            throw new Error(`Failed to encode PDF: ${error.message}`);
        }
        throw new Error('Failed to encode PDF: Unknown error');
    }
}
async function encodeImage(filePath) {
    try {
        // Read the image file as a buffer
        const imageBuffer = fs_1.default.readFileSync(filePath);
        // Convert the buffer to a Base64-encoded string
        const base64Image = imageBuffer.toString('base64');
        return base64Image;
    }
    catch (error) {
        if (error instanceof Error) {
            throw new Error(`Failed to encode image: ${error.message}`);
        }
        throw new Error('Failed to encode image: Unknown error');
    }
}
function handleMistralOCRResponse(response) {
    if (!response.pages || response.pages.length === 0) {
        throw new Error('No pages found in OCR response');
    }
    // Get markdown content from all pages
    const markdownContent = response.pages
        .sort((a, b) => a.index - b.index)
        .map(page => page.markdown)
        .join('\n\n');
    // Get image information
    const images = response.pages.flatMap(page => page.images);
    return {
        markdown: markdownContent,
        images,
        rawResponse: response
    };
}
async function getSignedMistralUrl(filePath, apiKey) {
    try {
        const client = new mistralai_1.Mistral({ apiKey });
        const fileContent = fs_1.default.readFileSync(filePath);
        const fileName = path_1.default.basename(filePath);
        const uploadedFile = await client.files.upload({
            file: {
                fileName,
                content: fileContent,
            },
            purpose: "ocr"
        });
        const signedUrl = await client.files.getSignedUrl({
            fileId: uploadedFile.id,
        });
        return signedUrl.url;
    }
    catch (error) {
        if (error instanceof Error) {
            throw new Error(`Failed to get signed URL: ${error.message}`);
        }
        throw new Error('Failed to get signed URL: Unknown error');
    }
}