unpdf 0.9.0 → 0.10.1

package/README.md

@@ -87,10 +87,11 @@ for (let i = 1; i <= doc.numPages; i++) {
 Generally speaking, you don't need to worry about the PDF.js build. `unpdf` ships with a serverless build of the latest PDF.js version. However, if you want to use the official PDF.js version or the legacy build, you can define a custom PDF.js module.
 
 ```ts
-// Before using any other methods, define the PDF.js module
-import { defineUnPDFConfig } from "unpdf";
+// Before using any other method, define the PDF.js module
+// if you need another PDF.js build
+import { configureUnPDF } from "unpdf";
 
-defineUnPDFConfig({
+configureUnPDF({
   // Use the official PDF.js build (make sure to install it first)
   pdfjs: () => import("pdfjs-dist"),
 });
@@ -118,12 +119,12 @@ interface UnPDFConfiguration {
 
 ## Methods
 
-### `defineUnPDFConfig`
+### `configureUnPDF`
 
 Define a custom PDF.js module, like the legacy build. Make sure to call this method before using any other methods.
 
 ```ts
-function defineUnPDFConfig(config: UnPDFConfiguration): Promise<void>;
+function configureUnPDF(config: UnPDFConfiguration): Promise<void>;
 ```
 
 ### `getResolvedPDFJS`
@@ -172,9 +173,9 @@ In order to use this method, you have to meet the following requirements:
 **Example**
 
 ```ts
-import { defineUnPDFConfig, renderPageAsImage } from "unpdf";
+import { configureUnPDF, renderPageAsImage } from "unpdf";
 
-defineUnPDFConfig({
+configureUnPDF({
   // Use the official PDF.js build
   pdfjs: () => import("pdfjs-dist"),
 });
@@ -193,7 +194,7 @@ await writeFile("dummy-page-1.png", Buffer.from(result));
 
 ```ts
 declare function renderPageAsImage(
-  data: BinaryData | PDFDocumentProxy,
+  data: BinaryData,
   pageNumber: number,
   options?: {
     canvas?: () => Promise<typeof import("canvas")>;

package/dist/index.cjs

@@ -8,6 +8,8 @@ async function getDocumentProxy(data, options = {}) {
   const pdf = await getDocument({
     data,
     isEvalSupported: false,
+    // See: https://github.com/mozilla/pdf.js/issues/4244#issuecomment-1479534301
+    useSystemFonts: true,
     ...options
   }).promise;
   return pdf;
@@ -24,12 +26,10 @@ async function resolvePDFJSImports(pdfjsResolver, { force = false } = {}) {
   }
   if (pdfjsResolver) {
     try {
-      const _import = await pdfjsResolver();
-      if ("resolvePDFJS" in _import) {
-        resolvedModule = await _import.resolvePDFJS();
-        return;
+      resolvedModule = await interopDefault(pdfjsResolver());
+      if (resolvedModule && "resolvePDFJS" in resolvedModule) {
+        resolvedModule = await resolvedModule.resolvePDFJS();
       }
-      resolvedModule = _import.default || _import;
       return;
     } catch {
       throw new Error(
@@ -49,6 +49,10 @@ async function resolvePDFJSImports(pdfjsResolver, { force = false } = {}) {
 function isPDFDocumentProxy(data) {
   return typeof data === "object" && data !== null && "_pdfInfo" in data;
 }
+async function interopDefault(m) {
+  const resolved = await m;
+  return resolved.default || resolved;
+}
 
 async function getMeta$1(data) {
   const pdf = isPDFDocumentProxy(data) ? data : await getDocumentProxy(data);
@@ -104,7 +108,8 @@ async function extractImages$1(data, pageNumber) {
   return images;
 }
 async function renderPageAsImage$1(data, pageNumber, options = {}) {
-  const pdf = isPDFDocumentProxy(data) ? data : await getDocumentProxy(data);
+  const canvasFactory = await createIsomorphicCanvasFactory(options.canvas);
+  const pdf = await getDocumentProxy(data, { canvasFactory });
   const page = await pdf.getPage(pageNumber);
   if (pageNumber < 1 || pageNumber > pdf.numPages) {
     throw new Error(
@@ -121,7 +126,6 @@ async function renderPageAsImage$1(data, pageNumber, options = {}) {
   if (outputScale !== 1 && outputScale > 0) {
     viewport = page.getViewport({ scale: outputScale });
   }
-  const canvasFactory = await createIsomorphicCanvasFactory(options.canvas);
   const ctx = canvasFactory.create(viewport.width, viewport.height);
   await page.render({
     canvasContext: ctx.context,
@@ -132,7 +136,7 @@ async function renderPageAsImage$1(data, pageNumber, options = {}) {
   return await response.arrayBuffer();
 }
 async function createIsomorphicCanvasFactory(canvas) {
-  const _canvas = await canvas?.();
+  const _canvas = canvas ? await interopDefault(canvas()) : void 0;
   return {
     _createCanvas(width, height) {
       if (isBrowser) {
@@ -140,7 +144,8 @@ async function createIsomorphicCanvasFactory(canvas) {
         canvas2.width = width;
         canvas2.height = height;
         return canvas2;
-      } else if (isNode) {
+      }
+      if (isNode) {
         if (!_canvas) {
           throw new Error('Failed to resolve "canvas" package.');
         }
@@ -175,13 +180,14 @@ async function createIsomorphicCanvasFactory(canvas) {
   };
 }
 
-async function defineUnPDFConfig(options) {
+async function configureUnPDF(options) {
   const { pdfjs } = { ...options };
   if (pdfjs) {
     await resolvePDFJSImports(pdfjs, { force: true });
   }
 }
 
+const defineUnPDFConfig = configureUnPDF;
 const getMeta = async (...args) => {
   await resolvePDFJSImports();
   return await getMeta$1(...args);
@@ -190,7 +196,6 @@ const extractText = async (...args) => {
   await resolvePDFJSImports();
   return await extractText$1(...args);
 };
-const extractPDFText = extractText;
 const extractImages = async (...args) => {
   await resolvePDFJSImports();
   return await extractImages$1(...args);
@@ -200,9 +205,9 @@ const renderPageAsImage = async (...args) => {
   return await renderPageAsImage$1(...args);
 };
 
+exports.configureUnPDF = configureUnPDF;
 exports.defineUnPDFConfig = defineUnPDFConfig;
 exports.extractImages = extractImages;
-exports.extractPDFText = extractPDFText;
 exports.extractText = extractText;
 exports.getDocumentProxy = getDocumentProxy;
 exports.getMeta = getMeta;

package/dist/index.d.cts

@@ -15,7 +15,7 @@ declare function extractText$1(data: BinaryData | PDFDocumentProxy, options?: {
 }>;
 
 declare function extractImages$1(data: BinaryData | PDFDocumentProxy, pageNumber: number): Promise<Uint8ClampedArray[]>;
-declare function renderPageAsImage$1(data: BinaryData | PDFDocumentProxy, pageNumber: number, options?: {
+declare function renderPageAsImage$1(data: BinaryData, pageNumber: number, options?: {
     canvas?: () => Promise<typeof canvas>;
     /** @default 1 */
     scale?: number;
@@ -37,13 +37,14 @@ interface UnPDFConfiguration {
     pdfjs?: () => Promise<any>;
 }
 
-declare function defineUnPDFConfig(options: UnPDFConfiguration): Promise<void>;
+declare function configureUnPDF(options: UnPDFConfiguration): Promise<void>;
 
 /**
  * Returns a PDFDocumentProxy instance from a given binary data.
  *
  * Applies the following defaults:
  * - `isEvalSupported: false`
+ * - `useSystemFonts: true`
  */
 declare function getDocumentProxy(data: BinaryData, options?: DocumentInitParameters): Promise<PDFDocumentProxy>;
 declare function getResolvedPDFJS(): Promise<PDFJS>;
@@ -51,11 +52,12 @@ declare function resolvePDFJSImports(pdfjsResolver?: () => Promise<any>, { force
     force?: boolean | undefined;
 }): Promise<void>;
 
+/** @deprecated Use `configureUnPDF` instead */
+declare const defineUnPDFConfig: typeof configureUnPDF;
+
 declare const getMeta: typeof getMeta$1;
 declare const extractText: typeof extractText$1;
-/** @deprecated Use `extractText` instead. */
-declare const extractPDFText: typeof extractText$1;
 declare const extractImages: typeof extractImages$1;
 declare const renderPageAsImage: typeof renderPageAsImage$1;
 
-export { defineUnPDFConfig, extractImages, extractPDFText, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };
+export { configureUnPDF, defineUnPDFConfig, extractImages, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };

package/dist/index.d.mts

@@ -15,7 +15,7 @@ declare function extractText$1(data: BinaryData | PDFDocumentProxy, options?: {
 }>;
 
 declare function extractImages$1(data: BinaryData | PDFDocumentProxy, pageNumber: number): Promise<Uint8ClampedArray[]>;
-declare function renderPageAsImage$1(data: BinaryData | PDFDocumentProxy, pageNumber: number, options?: {
+declare function renderPageAsImage$1(data: BinaryData, pageNumber: number, options?: {
     canvas?: () => Promise<typeof canvas>;
     /** @default 1 */
     scale?: number;
@@ -37,13 +37,14 @@ interface UnPDFConfiguration {
     pdfjs?: () => Promise<any>;
 }
 
-declare function defineUnPDFConfig(options: UnPDFConfiguration): Promise<void>;
+declare function configureUnPDF(options: UnPDFConfiguration): Promise<void>;
 
 /**
  * Returns a PDFDocumentProxy instance from a given binary data.
  *
  * Applies the following defaults:
  * - `isEvalSupported: false`
+ * - `useSystemFonts: true`
  */
 declare function getDocumentProxy(data: BinaryData, options?: DocumentInitParameters): Promise<PDFDocumentProxy>;
 declare function getResolvedPDFJS(): Promise<PDFJS>;
@@ -51,11 +52,12 @@ declare function resolvePDFJSImports(pdfjsResolver?: () => Promise<any>, { force
     force?: boolean | undefined;
 }): Promise<void>;
 
+/** @deprecated Use `configureUnPDF` instead */
+declare const defineUnPDFConfig: typeof configureUnPDF;
+
 declare const getMeta: typeof getMeta$1;
 declare const extractText: typeof extractText$1;
-/** @deprecated Use `extractText` instead. */
-declare const extractPDFText: typeof extractText$1;
 declare const extractImages: typeof extractImages$1;
 declare const renderPageAsImage: typeof renderPageAsImage$1;
 
-export { defineUnPDFConfig, extractImages, extractPDFText, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };
+export { configureUnPDF, defineUnPDFConfig, extractImages, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };

package/dist/index.d.ts

@@ -15,7 +15,7 @@ declare function extractText$1(data: BinaryData | PDFDocumentProxy, options?: {
 }>;
 
 declare function extractImages$1(data: BinaryData | PDFDocumentProxy, pageNumber: number): Promise<Uint8ClampedArray[]>;
-declare function renderPageAsImage$1(data: BinaryData | PDFDocumentProxy, pageNumber: number, options?: {
+declare function renderPageAsImage$1(data: BinaryData, pageNumber: number, options?: {
     canvas?: () => Promise<typeof canvas>;
     /** @default 1 */
     scale?: number;
@@ -37,13 +37,14 @@ interface UnPDFConfiguration {
     pdfjs?: () => Promise<any>;
 }
 
-declare function defineUnPDFConfig(options: UnPDFConfiguration): Promise<void>;
+declare function configureUnPDF(options: UnPDFConfiguration): Promise<void>;
 
 /**
  * Returns a PDFDocumentProxy instance from a given binary data.
  *
  * Applies the following defaults:
  * - `isEvalSupported: false`
+ * - `useSystemFonts: true`
  */
 declare function getDocumentProxy(data: BinaryData, options?: DocumentInitParameters): Promise<PDFDocumentProxy>;
 declare function getResolvedPDFJS(): Promise<PDFJS>;
@@ -51,11 +52,12 @@ declare function resolvePDFJSImports(pdfjsResolver?: () => Promise<any>, { force
     force?: boolean | undefined;
 }): Promise<void>;
 
+/** @deprecated Use `configureUnPDF` instead */
+declare const defineUnPDFConfig: typeof configureUnPDF;
+
 declare const getMeta: typeof getMeta$1;
 declare const extractText: typeof extractText$1;
-/** @deprecated Use `extractText` instead. */
-declare const extractPDFText: typeof extractText$1;
 declare const extractImages: typeof extractImages$1;
 declare const renderPageAsImage: typeof renderPageAsImage$1;
 
-export { defineUnPDFConfig, extractImages, extractPDFText, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };
+export { configureUnPDF, defineUnPDFConfig, extractImages, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };

package/dist/index.mjs

@@ -6,6 +6,8 @@ async function getDocumentProxy(data, options = {}) {
   const pdf = await getDocument({
     data,
     isEvalSupported: false,
+    // See: https://github.com/mozilla/pdf.js/issues/4244#issuecomment-1479534301
+    useSystemFonts: true,
     ...options
   }).promise;
   return pdf;
@@ -22,12 +24,10 @@ async function resolvePDFJSImports(pdfjsResolver, { force = false } = {}) {
   }
   if (pdfjsResolver) {
     try {
-      const _import = await pdfjsResolver();
-      if ("resolvePDFJS" in _import) {
-        resolvedModule = await _import.resolvePDFJS();
-        return;
+      resolvedModule = await interopDefault(pdfjsResolver());
+      if (resolvedModule && "resolvePDFJS" in resolvedModule) {
+        resolvedModule = await resolvedModule.resolvePDFJS();
       }
-      resolvedModule = _import.default || _import;
       return;
     } catch {
       throw new Error(
@@ -47,6 +47,10 @@ async function resolvePDFJSImports(pdfjsResolver, { force = false } = {}) {
 function isPDFDocumentProxy(data) {
   return typeof data === "object" && data !== null && "_pdfInfo" in data;
 }
+async function interopDefault(m) {
+  const resolved = await m;
+  return resolved.default || resolved;
+}
 
 async function getMeta$1(data) {
   const pdf = isPDFDocumentProxy(data) ? data : await getDocumentProxy(data);
@@ -102,7 +106,8 @@ async function extractImages$1(data, pageNumber) {
   return images;
 }
 async function renderPageAsImage$1(data, pageNumber, options = {}) {
-  const pdf = isPDFDocumentProxy(data) ? data : await getDocumentProxy(data);
+  const canvasFactory = await createIsomorphicCanvasFactory(options.canvas);
+  const pdf = await getDocumentProxy(data, { canvasFactory });
   const page = await pdf.getPage(pageNumber);
   if (pageNumber < 1 || pageNumber > pdf.numPages) {
     throw new Error(
@@ -119,7 +124,6 @@ async function renderPageAsImage$1(data, pageNumber, options = {}) {
   if (outputScale !== 1 && outputScale > 0) {
     viewport = page.getViewport({ scale: outputScale });
   }
-  const canvasFactory = await createIsomorphicCanvasFactory(options.canvas);
   const ctx = canvasFactory.create(viewport.width, viewport.height);
   await page.render({
     canvasContext: ctx.context,
@@ -130,7 +134,7 @@ async function renderPageAsImage$1(data, pageNumber, options = {}) {
   return await response.arrayBuffer();
 }
 async function createIsomorphicCanvasFactory(canvas) {
-  const _canvas = await canvas?.();
+  const _canvas = canvas ? await interopDefault(canvas()) : void 0;
   return {
     _createCanvas(width, height) {
       if (isBrowser) {
@@ -138,7 +142,8 @@ async function createIsomorphicCanvasFactory(canvas) {
         canvas2.width = width;
         canvas2.height = height;
         return canvas2;
-      } else if (isNode) {
+      }
+      if (isNode) {
         if (!_canvas) {
           throw new Error('Failed to resolve "canvas" package.');
         }
@@ -173,13 +178,14 @@ async function createIsomorphicCanvasFactory(canvas) {
   };
 }
 
-async function defineUnPDFConfig(options) {
+async function configureUnPDF(options) {
   const { pdfjs } = { ...options };
   if (pdfjs) {
     await resolvePDFJSImports(pdfjs, { force: true });
   }
 }
 
+const defineUnPDFConfig = configureUnPDF;
 const getMeta = async (...args) => {
   await resolvePDFJSImports();
   return await getMeta$1(...args);
@@ -188,7 +194,6 @@ const extractText = async (...args) => {
   await resolvePDFJSImports();
   return await extractText$1(...args);
 };
-const extractPDFText = extractText;
 const extractImages = async (...args) => {
   await resolvePDFJSImports();
   return await extractImages$1(...args);
@@ -198,4 +203,4 @@ const renderPageAsImage = async (...args) => {
   return await renderPageAsImage$1(...args);
 };
 
-export { defineUnPDFConfig, extractImages, extractPDFText, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };
+export { configureUnPDF, defineUnPDFConfig, extractImages, extractText, getDocumentProxy, getMeta, getResolvedPDFJS, renderPageAsImage, resolvePDFJSImports };