Skip to content

A fast, helpful, and open-source document parser

feat: unit tests (pdfjs)

Clelia (Astra) Bertelli332ca3e31748
diff --git a/src/engines/pdf/pdfjs.test.ts b/src/engines/pdf/pdfjs.test.ts
new file mode 100644
index 0000000..e230ffc
--- /dev/null
+++ b/src/engines/pdf/pdfjs.test.ts
@@ -0,0 +1,266 @@
+import { dirname, join } from "path";
+import { fileURLToPath } from "url";
+import { vi, describe, it, expect } from "vitest";
+
+vi.mock("node:fs/promises", async () => {
+ const actual = await vi.importActual<typeof import("node:fs/promises")>("node:fs/promises");
+ return {
+ default: {
+ ...actual,
+ readFile: vi.fn(async () => {
+ return Buffer.from("hello world");
+ }),
+ },
+ };
+});
+
+vi.mock("./pdfjsImporter.js", async () => {
+ const actual = await vi.importActual<typeof import("./pdfjsImporter.js")>("./pdfjsImporter.js");
+ return {
+ ...actual,
+ importPdfJs: vi.fn().mockImplementation(async () => {
+ const __filename = fileURLToPath(import.meta.url);
+ const __dirname = dirname(__filename);
+ // From dist/src/engines/pdf/ we need to go up to dist/src/vendor/pdfjs
+ const PDFJS_DIR = join(__dirname, "../../vendor/pdfjs");
+
+ const mockDocument = {
+ getMetadata: vi
+ .fn()
+ .mockResolvedValue({ numPages: 10, size: 1024 * 20, encoding: "utf-8" }),
+ numPages: 10,
+ };
+
+ const mockGetDocument = vi.fn().mockReturnValue({
+ promise: Promise.resolve(mockDocument),
+ });
+
+ return { fn: mockGetDocument, dir: PDFJS_DIR };
+ }),
+ };
+});
+
+vi.mock("./pdfium-renderer.js", async () => {
+ const actual = await vi.importActual<typeof import("./pdfjsImporter.js")>("./pdfjsImporter.js");
+ return {
+ ...actual,
+ PdfiumRenderer: vi.fn(
+ class {
+ constructor() {}
+
+ renderPageToBuffer = vi.fn().mockResolvedValue(Buffer.from("this is a page"));
+ }
+ ),
+ };
+});
+
+import { PdfJsEngine } from "./pdfjs.js";
+
+function getTestData() {
+ const mockViewport = {
+ width: 612,
+ height: 792,
+ transform: [1, 0, 0, 1, 0, 0],
+ };
+
+ const mockTextContent = {
+ items: [
+ {
+ str: "Hello, World!",
+ transform: [12, 0, 0, 12, 50, 700],
+ width: 100,
+ height: 12,
+ fontName: "Helvetica",
+ },
+ {
+ str: "Second line of text",
+ transform: [10, 0, 0, 10, 50, 680],
+ width: 150,
+ height: 10,
+ fontName: "Times-Roman",
+ },
+ ],
+ };
+ const mockPage = {
+ getViewport: vi.fn().mockReturnValue(mockViewport),
+ getTextContent: vi.fn().mockResolvedValue(mockTextContent),
+ cleanup: vi.fn(async () => {}),
+ };
+ const mockDocument = {
+ getPage: vi.fn().mockResolvedValue(mockPage),
+ numPages: 10,
+ };
+
+ const doc = {
+ numPages: 10,
+ data: new Uint8Array(Buffer.from("hello world")),
+ metadata: { numPages: 10, size: 1024 * 20, encoding: "utf-8" },
+ _pdfDocument: mockDocument,
+ };
+
+ return doc;
+}
+
+function getExpectedResults() {
+ return [
+ {
+ str: "Hello, World!",
+ x: 50,
+ y: 700,
+ width: 100,
+ height: 12,
+ w: 100,
+ h: 12,
+ r: 0,
+ fontName: "Helvetica",
+ fontSize: 12,
+ },
+ {
+ str: "Second line of text",
+ x: 50,
+ y: 680,
+ width: 150,
+ height: 10,
+ w: 150,
+ h: 10,
+ r: 0,
+ fontName: "Times-Roman",
+ fontSize: 10,
+ },
+ ];
+}
+
+describe("test PdfJS methods", () => {
+ it("test loadDocument", async () => {
+ const engine = new PdfJsEngine();
+ const result = await engine.loadDocument("test.pdf");
+ expect(result.data).toStrictEqual(new Uint8Array(Buffer.from("hello world")));
+ expect(result.metadata).toBeDefined();
+ expect("numPages" in (result.metadata as object)).toBeTruthy();
+ expect("size" in (result.metadata as object)).toBeTruthy();
+ expect("encoding" in (result.metadata as object)).toBeTruthy();
+ expect(result.numPages).toBe(10);
+ });
+
+ it("test extractPage", async () => {
+ const doc = getTestData();
+ const expectedTextItems = getExpectedResults();
+
+ const engine = new PdfJsEngine();
+ const result = await engine.extractPage(doc, 1);
+ expect(result.pageNum).toBe(1);
+ expect(result.width).toBe(612);
+ expect(result.height).toBe(792);
+ expect(result.images.length).toBe(0);
+ expect(result.annotations?.length).toBe(0);
+ expect(result.textItems).toStrictEqual(expectedTextItems);
+ expect(result.garbledTextRegions).toBeUndefined();
+ });
+
+ it("test extractAllPages (all pages)", async () => {
+ const doc = getTestData();
+ const expectedTextItems = getExpectedResults();
+
+ const engine = new PdfJsEngine();
+ const results = await engine.extractAllPages(doc);
+ expect(results.length).toBe(10);
+ let counter = 1;
+ for (const result of results) {
+ expect(result.pageNum).toBe(counter);
+ expect(result.width).toBe(612);
+ expect(result.height).toBe(792);
+ expect(result.images.length).toBe(0);
+ expect(result.annotations?.length).toBe(0);
+ expect(result.textItems).toStrictEqual(expectedTextItems);
+ expect(result.garbledTextRegions).toBeUndefined();
+ counter++;
+ }
+ });
+
+ it("test extractAllPages (with maxPages)", async () => {
+ const doc = getTestData();
+ const expectedTextItems = getExpectedResults();
+
+ const engine = new PdfJsEngine();
+ const results = await engine.extractAllPages(doc, 5);
+ expect(results.length).toBe(5);
+ let counter = 1;
+ for (const result of results) {
+ expect(result.pageNum).toBe(counter);
+ expect(result.width).toBe(612);
+ expect(result.height).toBe(792);
+ expect(result.images.length).toBe(0);
+ expect(result.annotations?.length).toBe(0);
+ expect(result.textItems).toStrictEqual(expectedTextItems);
+ expect(result.garbledTextRegions).toBeUndefined();
+ counter++;
+ }
+ });
+
+ it("test extractAllPages (with targetPages)", async () => {
+ const doc = getTestData();
+ const expectedTextItems = getExpectedResults();
+ const expectedPages = [1, 2, 3, 5];
+
+ const engine = new PdfJsEngine();
+ const results = await engine.extractAllPages(doc, undefined, "1,2,3,5");
+ expect(results.length).toBe(4);
+ for (let i = 0; i < results.length; i++) {
+ expect(results[i].pageNum).toBe(expectedPages[i]);
+ expect(results[i].width).toBe(612);
+ expect(results[i].height).toBe(792);
+ expect(results[i].images.length).toBe(0);
+ expect(results[i].annotations?.length).toBe(0);
+ expect(results[i].textItems).toStrictEqual(expectedTextItems);
+ expect(results[i].garbledTextRegions).toBeUndefined();
+ }
+ });
+
+ it("test extractAllPages (with targetPages and maxPages)", async () => {
+ const doc = getTestData();
+ const expectedTextItems = getExpectedResults();
+ const expectedPages = [1, 2, 3];
+
+ const engine = new PdfJsEngine();
+ const results = await engine.extractAllPages(doc, 4, "1,2,3,5,8");
+ expect(results.length).toBe(3);
+ for (let i = 0; i < results.length; i++) {
+ expect(results[i].pageNum).toBe(expectedPages[i]);
+ expect(results[i].width).toBe(612);
+ expect(results[i].height).toBe(792);
+ expect(results[i].images.length).toBe(0);
+ expect(results[i].annotations?.length).toBe(0);
+ expect(results[i].textItems).toStrictEqual(expectedTextItems);
+ expect(results[i].garbledTextRegions).toBeUndefined();
+ }
+ });
+
+ it("test extractAllPages (with targetPages and maxPages)", async () => {
+ const doc = getTestData();
+ const expectedTextItems = getExpectedResults();
+ const expectedPages = [1, 2, 3];
+
+ const engine = new PdfJsEngine();
+ const results = await engine.extractAllPages(doc, 4, "1,2,3,5,8");
+ expect(results.length).toBe(3);
+ for (let i = 0; i < results.length; i++) {
+ expect(results[i].pageNum).toBe(expectedPages[i]);
+ expect(results[i].width).toBe(612);
+ expect(results[i].height).toBe(792);
+ expect(results[i].images.length).toBe(0);
+ expect(results[i].annotations?.length).toBe(0);
+ expect(results[i].textItems).toStrictEqual(expectedTextItems);
+ expect(results[i].garbledTextRegions).toBeUndefined();
+ }
+ });
+
+ it("test renderPageImage", async () => {
+ const doc = getTestData();
+ const engine = new PdfJsEngine();
+ // ensure current PDF path is set
+ await engine.loadDocument("test.pdf");
+
+ const buf = await engine.renderPageImage(doc, 1, 20);
+ expect(buf).toStrictEqual(Buffer.from("this is a page"));
+ });
+});
diff --git a/src/engines/pdf/pdfjs.ts b/src/engines/pdf/pdfjs.ts
index cf14895..b07b9fc 100644
--- a/src/engines/pdf/pdfjs.ts
+++ b/src/engines/pdf/pdfjs.ts
@@ -1,9 +1,8 @@
import fs from "node:fs/promises";
-import { fileURLToPath } from "node:url";
-import { dirname, join } from "node:path";
import { PdfEngine, PdfDocument, PageData, Image, Annotation, BoundingBox } from "./interface.js";
import { TextItem } from "../../core/types.js";
import { PdfiumRenderer } from "./pdfium-renderer.js";
+import { importPdfJs } from "./pdfjsImporter.js";
/** PDF.js internal document type - opaque to our code */
interface PdfJsDocument {
@@ -47,15 +46,7 @@ interface PdfJsExtendedDocument extends PdfDocument {
}
// Dynamic import of PDF.js
-const __filename = fileURLToPath(import.meta.url);
-const __dirname = dirname(__filename);
-// From dist/src/engines/pdf/ we need to go up to dist/src/vendor/pdfjs
-const PDFJS_DIR = join(__dirname, "../../vendor/pdfjs");
-
-// Import PDF.js dynamically
-await import(`${PDFJS_DIR}/pdf.mjs`);
-const pdfjs = await import(`${PDFJS_DIR}/pdf.mjs`);
-const { getDocument } = pdfjs;
+const { fn: getDocument, dir: PDFJS_DIR } = await importPdfJs();
const CMAP_URL = `${PDFJS_DIR}/cmaps/`;
const STANDARD_FONT_DATA_URL = `${PDFJS_DIR}/standard_fonts/`;
diff --git a/src/engines/pdf/pdfjsImporter.ts b/src/engines/pdf/pdfjsImporter.ts
new file mode 100644
index 0000000..905aa46
--- /dev/null
+++ b/src/engines/pdf/pdfjsImporter.ts
@@ -0,0 +1,17 @@
+import { fileURLToPath } from "node:url";
+import { dirname, join } from "node:path";
+
+export async function importPdfJs() {
+ // Dynamic import of PDF.js
+ const __filename = fileURLToPath(import.meta.url);
+ const __dirname = dirname(__filename);
+ // From dist/src/engines/pdf/ we need to go up to dist/src/vendor/pdfjs
+ const PDFJS_DIR = join(__dirname, "../../vendor/pdfjs");
+
+ // Import PDF.js dynamically
+ await import(`${PDFJS_DIR}/pdf.mjs`);
+ const pdfjs = await import(`${PDFJS_DIR}/pdf.mjs`);
+ const { getDocument } = pdfjs;
+
+ return { fn: getDocument, dir: PDFJS_DIR };
+}