mindee 4.36.3 → 5.0.0-alpha1

package/src/parsing/index.js

@@ -1,41 +1 @@
-"use strict";
-var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    var desc = Object.getOwnPropertyDescriptor(m, k);
-    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
-      desc = { enumerable: true, get: function() { return m[k]; } };
-    }
-    Object.defineProperty(o, k2, desc);
-}) : (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    o[k2] = m[k];
-}));
-var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
-    Object.defineProperty(o, "default", { enumerable: true, value: v });
-}) : function(o, v) {
-    o["default"] = v;
-});
-var __importStar = (this && this.__importStar) || (function () {
-    var ownKeys = function(o) {
-        ownKeys = Object.getOwnPropertyNames || function (o) {
-            var ar = [];
-            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
-            return ar;
-        };
-        return ownKeys(o);
-    };
-    return function (mod) {
-        if (mod && mod.__esModule) return mod;
-        var result = {};
-        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
-        __setModuleDefault(result, mod);
-        return result;
-    };
-})();
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.v2 = exports.generated = exports.standard = exports.custom = exports.common = void 0;
-exports.common = __importStar(require("./common"));
-exports.custom = __importStar(require("./custom"));
-exports.standard = __importStar(require("./standard"));
-exports.generated = __importStar(require("./generated"));
-exports.v2 = __importStar(require("./v2"));
+export { parseDate } from "./dateParser.js";

package/src/parsing/localResponseBase.d.ts

@@ -0,0 +1,34 @@
+import { StringDict } from "../parsing/stringDict.js";
+import { Buffer } from "buffer";
+/**
+ * Local response loaded from a file.
+ * Note: Has to be initialized through init() before use.
+ */
+export declare abstract class LocalResponseBase {
+    private file;
+    private readonly inputHandle;
+    protected initialized: boolean;
+    /**
+     * Creates an instance of LocalResponse.
+     */
+    constructor(inputFile: Buffer | string);
+    init(): Promise<void>;
+    /**
+     * Returns the dictionary representation of the file.
+     * @returns A JSON-like object.
+     */
+    asDict(): Promise<StringDict>;
+    /**
+     * Returns the HMAC signature of the local response from the secret key provided.
+     * @param secretKey - Secret key, either a string or a byte/byte array.
+     * @returns The HMAC signature of the local response.
+     */
+    getHmacSignature(secretKey: string | Buffer | Uint8Array): string;
+    /**
+     * Checks if the HMAC signature of the local response is valid.
+     * @param secretKey - Secret key, either a string or a byte/byte array.
+     * @param signature - The signature to be compared with.
+     * @returns True if the HMAC signature is valid.
+     */
+    isValidHmacSignature(secretKey: string | Buffer | Uint8Array, signature: string): boolean;
+}

package/src/parsing/localResponseBase.js

@@ -0,0 +1,91 @@
+import * as crypto from "crypto";
+import * as fs from "node:fs/promises";
+import { MindeeError } from "../errors/index.js";
+import { Buffer } from "buffer";
+/**
+ * Local response loaded from a file.
+ * Note: Has to be initialized through init() before use.
+ */
+export class LocalResponseBase {
+    /**
+     * Creates an instance of LocalResponse.
+     */
+    constructor(inputFile) {
+        this.initialized = false;
+        this.file = Buffer.alloc(0);
+        this.inputHandle = inputFile;
+    }
+    async init() {
+        /**
+         * @param inputFile - The input file, which can be a Buffer, string, or PathLike.
+         */
+        if (this.initialized) {
+            return;
+        }
+        if (Buffer.isBuffer(this.inputHandle)) {
+            this.file = this.inputHandle;
+        }
+        else if (typeof this.inputHandle === "string") {
+            let fileContents;
+            try {
+                await fs.access(this.inputHandle);
+                fileContents = await fs.readFile(this.inputHandle, { encoding: "utf-8" });
+            }
+            catch {
+                fileContents = this.inputHandle;
+            }
+            this.file = Buffer.from(fileContents.replace(/\r/g, "").replace(/\n/g, ""), "utf-8");
+        }
+        else {
+            throw new MindeeError("Incompatible type for input.");
+        }
+        this.initialized = true;
+    }
+    /**
+     * Returns the dictionary representation of the file.
+     * @returns A JSON-like object.
+     */
+    async asDict() {
+        if (!this.initialized) {
+            await this.init();
+        }
+        try {
+            const content = this.file.toString("utf-8");
+            return JSON.parse(content);
+        }
+        catch {
+            throw new MindeeError("File is not a valid dictionary.");
+        }
+    }
+    /**
+     * Returns the HMAC signature of the local response from the secret key provided.
+     * @param secretKey - Secret key, either a string or a byte/byte array.
+     * @returns The HMAC signature of the local response.
+     */
+    getHmacSignature(secretKey) {
+        if (!this.initialized) {
+            throw new Error("The `init()` method must be called before calling `getHmacSignature()`.");
+        }
+        const algorithm = "sha256";
+        try {
+            const hmac = crypto.createHmac(algorithm, secretKey);
+            hmac.update(this.file);
+            return hmac.digest("hex");
+        }
+        catch {
+            throw new MindeeError("Could not get HMAC signature from payload.");
+        }
+    }
+    /**
+     * Checks if the HMAC signature of the local response is valid.
+     * @param secretKey - Secret key, either a string or a byte/byte array.
+     * @param signature - The signature to be compared with.
+     * @returns True if the HMAC signature is valid.
+     */
+    isValidHmacSignature(secretKey, signature) {
+        if (!this.initialized) {
+            throw new Error("The `init()` method must be called before calling `isValidHmacSignature()`.");
+        }
+        return signature === this.getHmacSignature(secretKey);
+    }
+}

package/src/parsing/stringDict.js

@@ -0,0 +1 @@
+export {};

package/src/pdf/index.d.ts

@@ -1,3 +1,4 @@
-export { extractPages, countPages, SplitPdf } from "./pdfOperation";
-export { compressPdf } from "./pdfCompressor";
-export { hasSourceText } from "./pdfUtils";
+export { extractPages, countPages } from "./pdfOperation.js";
+export type { SplitPdf } from "./pdfOperation.js";
+export { compressPdf } from "./pdfCompressor.js";
+export { hasSourceText } from "./pdfUtils.js";

package/src/pdf/index.js

@@ -1,10 +1,3 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.hasSourceText = exports.compressPdf = exports.countPages = exports.extractPages = void 0;
-var pdfOperation_1 = require("./pdfOperation");
-Object.defineProperty(exports, "extractPages", { enumerable: true, get: function () { return pdfOperation_1.extractPages; } });
-Object.defineProperty(exports, "countPages", { enumerable: true, get: function () { return pdfOperation_1.countPages; } });
-var pdfCompressor_1 = require("./pdfCompressor");
-Object.defineProperty(exports, "compressPdf", { enumerable: true, get: function () { return pdfCompressor_1.compressPdf; } });
-var pdfUtils_1 = require("./pdfUtils");
-Object.defineProperty(exports, "hasSourceText", { enumerable: true, get: function () { return pdfUtils_1.hasSourceText; } });
+export { extractPages, countPages } from "./pdfOperation.js";
+export { compressPdf } from "./pdfCompressor.js";
+export { hasSourceText } from "./pdfUtils.js";

package/src/pdf/pdfCompressor.js

@@ -1,49 +1,17 @@
-"use strict";
-var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    var desc = Object.getOwnPropertyDescriptor(m, k);
-    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
-      desc = { enumerable: true, get: function() { return m[k]; } };
+import { logger } from "../logger.js";
+import tmp from "tmp";
+import { extractTextFromPdf, hasSourceText } from "./pdfUtils.js";
+import * as fs from "node:fs";
+import { compressImage } from "../image/index.js";
+import { loadOptionalDependency } from "../dependency/index.js";
+let pdfLib = null;
+async function getPdfLib() {
+    if (!pdfLib) {
+        const pdfLibImport = await loadOptionalDependency("@cantoo/pdf-lib", "Text Embedding");
+        pdfLib = pdfLibImport.default || pdfLibImport;
     }
-    Object.defineProperty(o, k2, desc);
-}) : (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    o[k2] = m[k];
-}));
-var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
-    Object.defineProperty(o, "default", { enumerable: true, value: v });
-}) : function(o, v) {
-    o["default"] = v;
-});
-var __importStar = (this && this.__importStar) || (function () {
-    var ownKeys = function(o) {
-        ownKeys = Object.getOwnPropertyNames || function (o) {
-            var ar = [];
-            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
-            return ar;
-        };
-        return ownKeys(o);
-    };
-    return function (mod) {
-        if (mod && mod.__esModule) return mod;
-        var result = {};
-        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
-        __setModuleDefault(result, mod);
-        return result;
-    };
-})();
-var __importDefault = (this && this.__importDefault) || function (mod) {
-    return (mod && mod.__esModule) ? mod : { "default": mod };
-};
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.compressPdf = compressPdf;
-const logger_1 = require("../logger");
-const tmp_1 = __importDefault(require("tmp"));
-const pdfUtils_1 = require("./pdfUtils");
-const fs = __importStar(require("node:fs"));
-const node_poppler_1 = require("node-poppler");
-const pdf_lib_1 = require("@cantoo/pdf-lib");
-const imageOperations_1 = require("../imageOperations");
+    return pdfLib;
+}
 /**
  * Compresses each page of a provided PDF buffer.
  * @param pdfData The input PDF as a Buffer.
@@ -52,29 +20,29 @@ const imageOperations_1 = require("../imageOperations");
  * @param disableSourceText If true, doesn't re-apply source text to the output PDF.
  * @returns A Promise resolving to the compressed PDF as a Buffer.
  */
-async function compressPdf(pdfData, imageQuality = 85, forceSourceTextCompression = false, disableSourceText = true) {
+export async function compressPdf(pdfData, imageQuality = 85, forceSourceTextCompression = false, disableSourceText = true) {
     handleCompressionWarnings(forceSourceTextCompression, disableSourceText);
-    if (await (0, pdfUtils_1.hasSourceText)(pdfData)) {
+    if (await hasSourceText(pdfData)) {
         if (forceSourceTextCompression) {
             if (!disableSourceText) {
-                logger_1.logger.warn("Re-writing PDF source-text is an EXPERIMENTAL feature.");
+                logger.warn("Re-writing PDF source-text is an EXPERIMENTAL feature.");
             }
             else {
-                logger_1.logger.warn("Source file contains text, but disable_source_text flag. " +
+                logger.warn("Source file contains text, but disable_source_text flag. " +
                     "is set to false. Resulting file will not contain any embedded text.");
             }
         }
         else {
-            logger_1.logger.warn("Found text inside of the provided PDF file. Compression operation aborted since disableSourceText "
+            logger.warn("Found text inside of the provided PDF file. Compression operation aborted since disableSourceText "
                 + "is set to 'true'.");
             return pdfData;
         }
     }
-    const extractedText = disableSourceText ? await (0, pdfUtils_1.extractTextFromPdf)(pdfData) : null;
-    const extractedPdfInfo = await (0, pdfUtils_1.extractTextFromPdf)(pdfData);
+    const extractedText = disableSourceText ? await extractTextFromPdf(pdfData) : null;
+    const extractedPdfInfo = await extractTextFromPdf(pdfData);
     const compressedPages = await compressPdfPages(pdfData, extractedPdfInfo, imageQuality, disableSourceText, extractedText);
     if (!compressedPages) {
-        logger_1.logger.warn("Could not compress PDF to a smaller size. Returning original PDF.");
+        logger.warn("Could not compress PDF to a smaller size. Returning original PDF.");
         return pdfData;
     }
     return createNewPdfFromCompressedPages(compressedPages);
@@ -87,10 +55,10 @@ async function compressPdf(pdfData, imageQuality = 85, forceSourceTextCompressio
 function handleCompressionWarnings(forceSourceTextCompression, disableSourceText) {
     if (forceSourceTextCompression) {
         if (!disableSourceText) {
-            logger_1.logger.warn("Re-writing PDF source-text is an EXPERIMENTAL feature.");
+            logger.warn("Re-writing PDF source-text is an EXPERIMENTAL feature.");
         }
         else {
-            logger_1.logger.warn("Source file contains text, but the disable_source_text is set to false. "
+            logger.warn("Source file contains text, but the disable_source_text is set to false. "
                 + "Resulting file will not contain any embedded text.");
         }
     }
@@ -128,7 +96,8 @@ async function compressPdfPages(pdfData, extractedPdfInfo, imageQuality, disable
  * @returns A Promise resolving to an array of compressed page buffers.
  */
 async function compressPagesWithQuality(pdfData, extractedPdfInfo, imageQuality, disableSourceText, extractedText) {
-    const pdfDoc = await pdf_lib_1.PDFDocument.load(pdfData, {
+    const pdfLib = await getPdfLib();
+    const pdfDoc = await pdfLib.PDFDocument.load(pdfData, {
         ignoreEncryption: true,
         password: ""
     });
@@ -136,7 +105,7 @@ async function compressPagesWithQuality(pdfData, extractedPdfInfo, imageQuality,
     for (let i = 0; i < extractedPdfInfo.pages.length; i++) {
         const page = pdfDoc.getPages()[i];
         const rasterizedPage = await rasterizePage(pdfData, i + 1, imageQuality);
-        const compressedImage = await (0, imageOperations_1.compressImage)(Buffer.from(rasterizedPage, "binary"), imageQuality);
+        const compressedImage = await compressImage(Buffer.from(rasterizedPage, "binary"), imageQuality);
         if (!disableSourceText) {
             await addTextToPdfPage(page, extractedText);
         }
@@ -173,7 +142,8 @@ function isCompressionSuccessful(totalCompressedSize, originalSize, imageQuality
  * @returns A Promise resolving to the new PDF as a Buffer.
  */
 async function createNewPdfFromCompressedPages(compressedPages) {
-    const newPdfDoc = await pdf_lib_1.PDFDocument.create();
+    const pdfLib = await getPdfLib();
+    const newPdfDoc = await pdfLib.PDFDocument.create();
     for (const compressedPage of compressedPages) {
         const image = await newPdfDoc.embedJpg(compressedPage);
         const newPage = newPdfDoc.addPage([image.width, image.height]);
@@ -191,26 +161,29 @@ async function addTextToPdfPage(page, textInfo) {
     if (textInfo === null) {
         return;
     }
+    const pdfLib = await getPdfLib();
     for (const textPages of textInfo.pages) {
         for (const textPage of textPages.content) {
             page.drawText(textPage.str, {
                 x: textPage.x,
                 y: textPage.y,
                 size: textPage.height,
-                color: (0, pdf_lib_1.rgb)(0, 0, 0),
+                color: pdfLib.rgb(0, 0, 0),
                 font: await getFontFromName(textPage.fontName)
             });
         }
     }
 }
 async function getFontFromName(fontName) {
-    const pdfDoc = await pdf_lib_1.PDFDocument.create();
+    const pdfLib = await getPdfLib();
+    const pdfDoc = await pdfLib.PDFDocument.create();
     let font;
-    if (Object.values(pdf_lib_1.StandardFonts).map(value => value.toString()).includes(fontName)) {
+    const standardFontValues = Object.values(pdfLib.StandardFonts);
+    if (standardFontValues.includes(fontName)) {
         font = await pdfDoc.embedFont(fontName);
     }
     else {
-        font = await pdfDoc.embedFont(pdf_lib_1.StandardFonts.Helvetica);
+        font = await pdfDoc.embedFont(pdfLib.StandardFonts.Helvetica);
     }
     return font;
 }
@@ -222,8 +195,10 @@ async function getFontFromName(fontName) {
  * @param quality Quality to apply during rasterization.
  */
 async function rasterizePage(pdfData, index, quality = 85) {
-    const poppler = new node_poppler_1.Poppler();
-    const tmpPdf = tmp_1.default.fileSync();
+    const popplerImport = await loadOptionalDependency("node-poppler", "Image Processing");
+    const poppler = popplerImport.default || popplerImport;
+    const popplerInstance = new poppler.Poppler();
+    const tmpPdf = tmp.fileSync();
     const tempPdfPath = tmpPdf.name;
     const antialiasOption = "best";
     try {
@@ -236,12 +211,12 @@ async function rasterizePage(pdfData, index, quality = 85) {
             jpegOptions: `quality=${quality}`,
             singleFile: true
         };
-        const jpegBuffer = await poppler.pdfToCairo(tempPdfPath, undefined, options);
+        const jpegBuffer = await popplerInstance.pdfToCairo(tempPdfPath, undefined, options);
         await fs.promises.unlink(tempPdfPath);
         return jpegBuffer;
     }
     catch (error) {
-        logger_1.logger.error("Error rasterizing PDF:", error);
+        logger.error("Error rasterizing PDF:", error);
         throw error;
     }
     finally {

package/src/pdf/pdfOperation.d.ts

@@ -1,17 +1,17 @@
-import { PageOptions } from "../input";
+import { PageOptions } from "../input/pageOptions.js";
 export interface SplitPdf {
     file: Buffer;
     totalPagesRemoved: number;
 }
 /**
- * Cut pages from a pdf file. If pages index are out of bound, it will throw an error.
+ * Cut pages from a PDF file. If pages indexes are out of bounds, it will throw an error.
  * @param file
  * @param pageOptions
- * @returns the new cut pdf file.
+ * @returns the new cut PDF file.
  */
 export declare function extractPages(file: Buffer, pageOptions: PageOptions): Promise<SplitPdf>;
 /**
- * Count the number of pages in a pdf file.
+ * Count the number of pages in a PDF file.
  * @param file
  * @returns the number of pages in the file.
  */

package/src/pdf/pdfOperation.js

@@ -1,31 +1,36 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.extractPages = extractPages;
-exports.countPages = countPages;
-const handler_1 = require("../errors/handler");
-const pdf_lib_1 = require("@cantoo/pdf-lib");
-const input_1 = require("../input");
-const errors_1 = require("../errors");
-const logger_1 = require("../logger");
+import { errorHandler } from "../errors/handler.js";
+import { PageOptionsOperation } from "../input/pageOptions.js";
+import { MindeeError } from "../errors/index.js";
+import { logger } from "../logger.js";
+import { loadOptionalDependency } from "../dependency/index.js";
+let pdfLib = null;
+async function getPdfLib() {
+    if (!pdfLib) {
+        const pdfLibImport = await loadOptionalDependency("@cantoo/pdf-lib", "Text Embedding");
+        pdfLib = pdfLibImport.default || pdfLibImport;
+    }
+    return pdfLib;
+}
 /**
- * Cut pages from a pdf file. If pages index are out of bound, it will throw an error.
+ * Cut pages from a PDF file. If pages indexes are out of bounds, it will throw an error.
  * @param file
  * @param pageOptions
- * @returns the new cut pdf file.
+ * @returns the new cut PDF file.
  */
-async function extractPages(file, pageOptions) {
-    const currentPdf = await pdf_lib_1.PDFDocument.load(file, {
+export async function extractPages(file, pageOptions) {
+    const pdfLib = await getPdfLib();
+    const currentPdf = await pdfLib.PDFDocument.load(file, {
         ignoreEncryption: true,
         password: ""
     });
-    const newPdf = await pdf_lib_1.PDFDocument.create();
+    const newPdf = await pdfLib.PDFDocument.create();
     const pageCount = currentPdf.getPageCount();
     if (pageCount < pageOptions.onMinPages) {
-        logger_1.logger.debug(`File skipped because it had less than ${pageOptions} pages (${pageCount}).`);
+        logger.debug(`File skipped because it had less than ${pageOptions} pages (${pageCount}).`);
         return { file: file, totalPagesRemoved: 0 };
     }
     if (pageOptions.pageIndexes.length > pageCount) {
-        handler_1.errorHandler.throw(new errors_1.MindeeError("The total indexes of pages to cut is superior to the total page count of the file (" +
+        errorHandler.throw(new MindeeError("The total indexes of pages to cut is superior to the total page count of the file (" +
             pageCount +
             ")."));
     }
@@ -39,18 +44,18 @@ async function extractPages(file, pageOptions) {
         }
     });
     if (!pageIndexes.every((v) => currentPdf.getPageIndices().includes(v))) {
-        handler_1.errorHandler.throw(new errors_1.MindeeError(`Some indexes pages
+        errorHandler.throw(new MindeeError(`Some indexes pages
         (${pageIndexes.join(",")})
         don't exist in the file
         (${currentPdf.getPageIndices().join(", ")})`));
     }
-    if (pageOptions.operation === input_1.PageOptionsOperation.KeepOnly) {
+    if (pageOptions.operation === PageOptionsOperation.KeepOnly) {
         const keptPages = await newPdf.copyPages(currentPdf, pageIndexes);
         keptPages.forEach((keptPage) => {
             newPdf.addPage(keptPage);
         });
     }
-    else if (pageOptions.operation === input_1.PageOptionsOperation.Remove) {
+    else if (pageOptions.operation === PageOptionsOperation.Remove) {
         const pagesToKeep = currentPdf
             .getPageIndices()
             .filter((v) => !pageIndexes.includes(v));
@@ -67,12 +72,13 @@ async function extractPages(file, pageOptions) {
     return { file: fileBuffer, totalPagesRemoved: sumRemovedPages };
 }
 /**
- * Count the number of pages in a pdf file.
+ * Count the number of pages in a PDF file.
  * @param file
  * @returns the number of pages in the file.
  */
-async function countPages(file) {
-    const currentPdf = await pdf_lib_1.PDFDocument.load(file, {
+export async function countPages(file) {
+    const pdfLib = await getPdfLib();
+    const currentPdf = await pdfLib.PDFDocument.load(file, {
         ignoreEncryption: true,
         password: ""
     });

package/src/pdf/pdfUtils.js

@@ -1,9 +1,5 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.extractTextFromPdf = extractTextFromPdf;
-exports.hasSourceText = hasSourceText;
-const pdf_js_extract_1 = require("pdf.js-extract");
-const mindeeError_1 = require("../errors/mindeeError");
+import { MindeePdfError } from "../errors/index.js";
+import { loadOptionalDependency } from "../dependency/index.js";
 function getConcatenatedText(pages) {
     return pages.flatMap(page => page.content.map(item => item.str)).join(" ");
 }
@@ -13,22 +9,24 @@ function getConcatenatedText(pages) {
  * @returns A Promise containing the extracted text as a string.
  * @param pdfBuffer PDF handle, as a buffer.
  */
-async function extractTextFromPdf(pdfBuffer) {
-    const pdfExtract = new pdf_js_extract_1.PDFExtract();
+export async function extractTextFromPdf(pdfBuffer) {
+    const pdfJsExtract = await loadOptionalDependency("pdf.js-extract", "PDF text extraction");
+    const pdfJs = pdfJsExtract.pdfjs || pdfJsExtract;
+    const pdfExtract = new pdfJs.PDFExtract();
     const options = {};
     const pdf = await new Promise((resolve, reject) => {
         pdfExtract.extractBuffer(pdfBuffer, options, (err, result) => {
             if (err)
                 reject(err);
             if (result === undefined)
-                reject(new mindeeError_1.MindeePdfError("Couldn't process result."));
+                reject(new MindeePdfError("Couldn't process result."));
             else
                 resolve(result);
         });
     });
     const pages = pdf.pages.map((page, index) => ({
         pageNumber: index + 1,
-        content: page.content.map(item => ({
+        content: page.content.map((item) => ({
             str: item.str,
             x: item.x,
             y: item.y,
@@ -49,7 +47,7 @@ async function extractTextFromPdf(pdfBuffer) {
  *
  * @returns A Promise containing a boolean indicating if the PDF has source text.
  */
-async function hasSourceText(pdfData) {
+export async function hasSourceText(pdfData) {
     const text = await extractTextFromPdf(pdfData);
     return text.getConcatenatedText().trim().length > 0;
 }