unpdf 0.5.4 → 0.6.0

package/README.md

@@ -1,16 +1,18 @@
 # unpdf
 
-A collection of utilities to work with PDFs. Uses Mozilla's [PDF.js](https://github.com/mozilla/pdf.js) under the hood and lazily initializes the library.
+A collection of utilities to work with PDFs. Designed specifically for Deno, workers and other nodeless environments.
 
-`unpdf` takes advantage of [export conditions](https://nodejs.org/api/packages.html#packages_conditional_exports) to circumvent build issues in serverless environments. For example, PDF.js depends on the optional `canvas` module, which [doesn't work inside worker threads](https://github.com/Automattic/node-canvas/issues/1394).
+`unpdf` ships with a redistribution of Mozilla's [PDF.js](https://github.com/mozilla/pdf.js) for serverless environments. Apart from some string replacements and mocks, [`unenv`](https://github.com/unjs/unenv) does the heavy lifting by converting Node.js specific code to be platform-agnostic.
 
 This library is also intended as a modern alternative to the unmaintained but still popular [`pdf-parse`](https://www.npmjs.com/package/pdf-parse).
 
 ## Features
 
-- 🏗️ Conditional exports for Browser, Node and worker environments
+- 🏗️ Works in Node.js, browser and workers
+- 🪭 Includes serverless build of PDF.js ([`unpdf/pdfjs`](./package.json#L45))
 - 💬 Extract text and images from PDFs
 - 🧱 Opt-in to legacy PDF.js build
+- 💨 Zero dependencies
 
 ## Installation
 
@@ -29,32 +31,37 @@ yarn add unpdf
 
 ## Usage
 
+### Extract Text From PDF
+
 ```ts
-import { extractPDFText } from 'unpdf'
+import { extractPDFText } from "unpdf";
 
 // Fetch a PDF file from the web
-const pdf = await fetch('https://www.w3.org/WAI/ER/tests/xhtml/testfiles/resources/pdf/dummy.pdf')
-  .then(res => res.arrayBuffer())
+const pdf = await fetch(
+  "https://www.w3.org/WAI/ER/tests/xhtml/testfiles/resources/pdf/dummy.pdf",
+).then((res) => res.arrayBuffer());
 
 // Or load it from the filesystem
-const pdf = await readFile('./dummy.pdf')
+const pdf = await readFile("./dummy.pdf");
 
 // Pass the PDF buffer to the relevant method
-const { totalPages, text } = await extractPDFText(
-  new Uint8Array(pdf), { mergePages: true }
-)
+const { totalPages, text } = await extractPDFText(new Uint8Array(pdf), {
+  mergePages: true,
+});
 ```
 
 ### Use Legacy Or Custom PDF.js Build
 
+Typically, 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 an older 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'
+import { defineUnPDFConfig } from "unpdf";
 
-// Use the legacy build
 defineUnPDFConfig({
-  pdfjs: () => import('pdfjs-dist/legacy/build/pdf.js')
-})
+  // Use the legacy build
+  pdfjs: () => import("pdfjs-dist/legacy/build/pdf.js"),
+});
 
 // Now, you can use the other methods
 // …
@@ -62,10 +69,32 @@ defineUnPDFConfig({
 
 ### Access the PDF.js Module
 
+This will return the resolved PDF.js module. If no build is defined, the serverless build bundled with `unpdf` will be initialized.
+
+```ts
+import { getResolvedPDFJS } from "unpdf";
+
+const { version } = await getResolvedPDFJS();
+```
+
+### Use Serverless PDF.js Build In 🦕 Deno
+
+Instead of using the methods provided by `unpdf`, you can directly import the serverless PDF.js build in Deno. This is useful if you want to use the PDF.js API directly.
+
 ```ts
-import { getResolvedPDFJS } from 'unpdf'
+import { getDocument } from "https://esm.sh/unpdf/pdfjs";
+
+const data = Deno.readFileSync("dummy.pdf");
+const doc = await getDocument(data).promise;
 
-const { version } = await getResolvedPDFJS()
+console.log(await doc.getMetadata());
+
+for (let i = 1; i <= doc.numPages; i++) {
+  const page = await doc.getPage(i);
+  const textContent = await page.getTextContent();
+  const contents = textContent.items.map((item) => item.str).join(" ");
+  console.log(contents);
+}
 ```
 
 ## Config
@@ -80,7 +109,7 @@ interface UnPDFConfiguration {
    * @example
    * () => import('pdfjs-dist/legacy/build/pdf.js')
    */
-  pdfjs?: () => Promise<typeof PDFJS>
+  pdfjs?: () => Promise<typeof PDFJS>;
 }
 ```
 
@@ -91,7 +120,7 @@ interface UnPDFConfiguration {
 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 defineUnPDFConfig(config: UnPDFConfiguration): Promise<void>;
 ```
 
 ### `getResolvedPDFJS`
@@ -99,18 +128,16 @@ function defineUnPDFConfig(config: UnPDFConfiguration): Promise<void>
 Returns the resolved PDF.js module. If no build is defined, the latest version will be initialized.
 
 ```ts
-function getResolvedPDFJS(): Promise<typeof import('pdfjs-dist')>
+function getResolvedPDFJS(): Promise<typeof import("pdfjs-dist")>;
 ```
 
 ### `getPDFMeta`
 
 ```ts
-function getPDFMeta(
-  data: BinaryData | PDFDocumentProxy
-): Promise<{
-  info: Record<string, any>
-  metadata: Record<string, any>
-}>
+function getPDFMeta(data: BinaryData | PDFDocumentProxy): Promise<{
+  info: Record<string, any>;
+  metadata: Record<string, any>;
+}>;
 ```
 
 ### `extractPDFText`
@@ -118,11 +145,11 @@ function getPDFMeta(
 ```ts
 function extractPDFText(
   data: BinaryData | PDFDocumentProxy,
-  { mergePages }?: { mergePages?: boolean }
+  { mergePages }?: { mergePages?: boolean },
 ): Promise<{
-  totalPages: number
-  text: string | string[]
-}>
+  totalPages: number;
+  text: string | string[];
+}>;
 ```
 
 ### `getImagesFromPage`
@@ -130,10 +157,16 @@ function extractPDFText(
 ```ts
 function getImagesFromPage(
   data: BinaryData | PDFDocumentProxy,
-  pageNumber: number
-): Promise<ArrayBuffer[]>
+  pageNumber: number,
+): Promise<ArrayBuffer[]>;
 ```
 
+## FAQ
+
+### What About Canvas?
+
+The official PDF.js library depends on the optional `canvas` module, which [doesn't work inside worker threads](https://github.com/Automattic/node-canvas/issues/1394). That's why `unpdf` ships with a serverless build of PDF.js that mocks the `canvas` module.
+
 ## License
 
 [MIT](./LICENSE) License © 2023-PRESENT [Johann Schopplich](https://github.com/johannschopplich)

package/dist/index.node.cjs

@@ -1,25 +1,24 @@
 'use strict';
 
-const image = require('./shared/unpdf.eba9a3dc.cjs');
-const utils = require('./shared/unpdf.42987ff1.cjs');
+const image = require('./shared/unpdf.6c0ca5b5.cjs');
 
 const getPDFMeta = async (...args) => {
-  await utils.resolvePDFJSImports();
+  await image.resolvePDFJSImports();
   return await image.getPDFMeta(...args);
 };
 const extractPDFText = async (...args) => {
-  await utils.resolvePDFJSImports();
+  await image.resolvePDFJSImports();
   return await image.extractPDFText(...args);
 };
 const getImagesFromPage = async (...args) => {
-  await utils.resolvePDFJSImports();
+  await image.resolvePDFJSImports();
   return await image.getImagesFromPage(...args);
 };
 
-exports.defineUnPDFConfig = utils.defineUnPDFConfig;
-exports.getDocumentProxy = utils.getDocumentProxy;
-exports.getResolvedPDFJS = utils.getResolvedPDFJS;
-exports.resolvePDFJSImports = utils.resolvePDFJSImports;
+exports.defineUnPDFConfig = image.defineUnPDFConfig;
+exports.getDocumentProxy = image.getDocumentProxy;
+exports.getResolvedPDFJS = image.getResolvedPDFJS;
+exports.resolvePDFJSImports = image.resolvePDFJSImports;
 exports.extractPDFText = extractPDFText;
 exports.getImagesFromPage = getImagesFromPage;
 exports.getPDFMeta = getPDFMeta;

package/dist/index.node.d.cts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.cjs';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.cjs';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.cjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.cjs';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.node.d.mts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.mjs';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.mjs';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.mjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.mjs';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.node.d.ts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.js';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.js';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.js';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.js';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.node.mjs

@@ -1,6 +1,5 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3dcacb97.mjs';
-import { r as resolvePDFJSImports } from './shared/unpdf.ab6a2535.mjs';
-export { d as defineUnPDFConfig, g as getDocumentProxy, a as getResolvedPDFJS } from './shared/unpdf.ab6a2535.mjs';
+import { r as resolvePDFJSImports, g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.ce219a40.mjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS } from './shared/unpdf.ce219a40.mjs';
 
 const getPDFMeta = async (...args) => {
   await resolvePDFJSImports();

package/dist/index.web.cjs

@@ -1,25 +1,24 @@
 'use strict';
 
-const image = require('./shared/unpdf.eba9a3dc.cjs');
-const utils = require('./shared/unpdf.42987ff1.cjs');
+const image = require('./shared/unpdf.6c0ca5b5.cjs');
 
 const getPDFMeta = async (...args) => {
-  await utils.resolvePDFJSImports();
+  await image.resolvePDFJSImports();
   return await image.getPDFMeta(...args);
 };
 const extractPDFText = async (...args) => {
-  await utils.resolvePDFJSImports();
+  await image.resolvePDFJSImports();
   return await image.extractPDFText(...args);
 };
 const getImagesFromPage = async (...args) => {
-  await utils.resolvePDFJSImports();
+  await image.resolvePDFJSImports();
   return await image.getImagesFromPage(...args);
 };
 
-exports.defineUnPDFConfig = utils.defineUnPDFConfig;
-exports.getDocumentProxy = utils.getDocumentProxy;
-exports.getResolvedPDFJS = utils.getResolvedPDFJS;
-exports.resolvePDFJSImports = utils.resolvePDFJSImports;
+exports.defineUnPDFConfig = image.defineUnPDFConfig;
+exports.getDocumentProxy = image.getDocumentProxy;
+exports.getResolvedPDFJS = image.getResolvedPDFJS;
+exports.resolvePDFJSImports = image.resolvePDFJSImports;
 exports.extractPDFText = extractPDFText;
 exports.getImagesFromPage = getImagesFromPage;
 exports.getPDFMeta = getPDFMeta;

package/dist/index.web.d.cts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.cjs';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.cjs';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.cjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.cjs';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.web.d.mts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.mjs';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.mjs';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.mjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.mjs';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.web.d.ts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.js';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.js';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.js';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.js';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.web.mjs

@@ -1,6 +1,5 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3dcacb97.mjs';
-import { r as resolvePDFJSImports } from './shared/unpdf.ab6a2535.mjs';
-export { d as defineUnPDFConfig, g as getDocumentProxy, a as getResolvedPDFJS } from './shared/unpdf.ab6a2535.mjs';
+import { r as resolvePDFJSImports, g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.ce219a40.mjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS } from './shared/unpdf.ce219a40.mjs';
 
 const getPDFMeta = async (...args) => {
   await resolvePDFJSImports();

package/dist/index.worker.cjs

@@ -1,21 +1,24 @@
 'use strict';
 
-const utils = require('./shared/unpdf.42987ff1.cjs');
+const image = require('./shared/unpdf.6c0ca5b5.cjs');
 
-const getPDFMeta = async () => {
-  throw new Error("Not implemented in worker context yet");
+const getPDFMeta = async (...args) => {
+  await image.resolvePDFJSImports();
+  return await image.getPDFMeta(...args);
 };
-const extractPDFText = async () => {
-  throw new Error("Not implemented in worker context yet");
+const extractPDFText = async (...args) => {
+  await image.resolvePDFJSImports();
+  return await image.extractPDFText(...args);
 };
-const getImagesFromPage = async () => {
-  throw new Error("Not implemented in worker context yet");
+const getImagesFromPage = async (...args) => {
+  await image.resolvePDFJSImports();
+  return await image.getImagesFromPage(...args);
 };
 
-exports.defineUnPDFConfig = utils.defineUnPDFConfig;
-exports.getDocumentProxy = utils.getDocumentProxy;
-exports.getResolvedPDFJS = utils.getResolvedPDFJS;
-exports.resolvePDFJSImports = utils.resolvePDFJSImports;
+exports.defineUnPDFConfig = image.defineUnPDFConfig;
+exports.getDocumentProxy = image.getDocumentProxy;
+exports.getResolvedPDFJS = image.getResolvedPDFJS;
+exports.resolvePDFJSImports = image.resolvePDFJSImports;
 exports.extractPDFText = extractPDFText;
 exports.getImagesFromPage = getImagesFromPage;
 exports.getPDFMeta = getPDFMeta;

package/dist/index.worker.d.cts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.cjs';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.cjs';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.cjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.cjs';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.worker.d.mts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.mjs';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.mjs';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.mjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.mjs';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.worker.d.ts

@@ -1,7 +1,7 @@
-import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.6363738c.js';
-export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.6363738c.js';
-import 'pdfjs-dist/types/src/display/api';
-import 'pdfjs-dist';
+import { g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.3991886a.js';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.3991886a.js';
+import 'types/src/display/api';
+import 'types/src/pdf';
 
 declare const getPDFMeta: typeof getPDFMeta$1;
 declare const extractPDFText: typeof extractPDFText$1;

package/dist/index.worker.mjs

@@ -1,13 +1,17 @@
-export { d as defineUnPDFConfig, g as getDocumentProxy, a as getResolvedPDFJS, r as resolvePDFJSImports } from './shared/unpdf.ab6a2535.mjs';
+import { r as resolvePDFJSImports, g as getPDFMeta$1, e as extractPDFText$1, a as getImagesFromPage$1 } from './shared/unpdf.ce219a40.mjs';
+export { d as defineUnPDFConfig, b as getDocumentProxy, c as getResolvedPDFJS } from './shared/unpdf.ce219a40.mjs';
 
-const getPDFMeta = async () => {
-  throw new Error("Not implemented in worker context yet");
+const getPDFMeta = async (...args) => {
+  await resolvePDFJSImports();
+  return await getPDFMeta$1(...args);
 };
-const extractPDFText = async () => {
-  throw new Error("Not implemented in worker context yet");
+const extractPDFText = async (...args) => {
+  await resolvePDFJSImports();
+  return await extractPDFText$1(...args);
 };
-const getImagesFromPage = async () => {
-  throw new Error("Not implemented in worker context yet");
+const getImagesFromPage = async (...args) => {
+  await resolvePDFJSImports();
+  return await getImagesFromPage$1(...args);
 };
 
-export { extractPDFText, getImagesFromPage, getPDFMeta };
+export { extractPDFText, getImagesFromPage, getPDFMeta, resolvePDFJSImports };

package/dist/pdfjs.d.mts

@@ -0,0 +1 @@
+export * from './types/src/pdf.d.ts'

package/dist/pdfjs.d.ts

@@ -0,0 +1 @@
+export * from './types/src/pdf.d.ts'