unpdf 0.2.1 → 0.3.0

package/README.md

@@ -1,13 +1,15 @@
 # unpdf
 
-A collection of utilities to work with PDFs.
+A collection of utilities to work with PDFs. Uses Mozilla's [PDF.js](https://github.com/mozilla/pdf.js) under the hood.
 
-`unpdf` takes advantage of [export conditions](https://nodejs.org/api/packages.html#packages_conditional_exports) to provide a minimal bundle size for the browser. As of now, the available methods are only supported in Node contexts.
+`unpdf` takes advantage of [export conditions](https://nodejs.org/api/packages.html#packages_conditional_exports) to circumvent build issues in serverless environments. PDF.js depends on the optional `canvas` module, which [doesn't work inside worker threads](https://github.com/Automattic/node-canvas/issues/1394).
 
-**Why this package then?**
+This library is also intended as a modern alternative to the unmaintained [`pdf-parse`](https://www.npmjs.com/package/pdf-parse).
 
-- To circumvent build issues in serverless environments, where the `canvas` package used by `PDF.js` is not supported.
-- WIP and more to come.
+## Features
+
+- 🏗️ Conditional exports for Browser, Node and worker environments
+- 💬 Extract text from PDFs
 
 ## Installation
 

package/dist/index.node.cjs

@@ -1,67 +1,15 @@
 'use strict';
 
-const pdfjsDist = require('pdfjs-dist');
-
-async function getDocumentProxy(data) {
-  const { getDocument } = await getPDFJSImports();
-  const pdf = await getDocument({
-    data,
-    useWorkerFetch: false,
-    useSystemFonts: true,
-    isEvalSupported: false
-  }).promise;
-  return pdf;
-}
-async function getPDFJSImports() {
-  try {
-    const { default: mod } = await import('pdfjs-dist/legacy/build/pdf');
-    const { getDocument, version } = mod;
-    return { getDocument, version };
-  } catch (e) {
-    console.error(e);
-    throw new Error(
-      "PDF.js is not available. Please run `pnpm add -D pdfjs-dist` and try again."
-    );
-  }
-}
-
-async function decodePDFText(data, { mergePages = false } = {}) {
-  const pdf = await getDocumentProxy(data);
-  const meta = await pdf.getMetadata().catch(() => null);
-  const texts = await Promise.all(
-    Array.from({ length: pdf.numPages }, (_, i) => getPageText(pdf, i + 1))
-  );
-  const filteredTexts = texts.filter(Boolean);
-  return {
-    totalPages: pdf.numPages,
-    info: meta?.info,
-    metadata: meta?.metadata?.getAll(),
-    text: mergePages ? filteredTexts.join("\n\n") : filteredTexts
-  };
-}
-async function getPageText(pdf, pageNumber) {
-  const page = await pdf.getPage(pageNumber);
-  const content = await page.getTextContent();
-  const items = content.items;
-  if (items.length === 0)
-    return;
-  return items.map(({ str }) => str).join("\n");
-}
-
-async function getImagesFromPage(data, pageNumber) {
-  const pdf = await getDocumentProxy(data);
-  const page = await pdf.getPage(pageNumber);
-  const operatorList = await page.getOperatorList();
-  const images = [];
-  for (const op of operatorList.fnArray) {
-    if (op !== pdfjsDist.OPS.paintImageXObject)
-      continue;
-    const image = await page.objs.get(operatorList.argsArray[op][0]);
-    if (image.data)
-      images.push(image.data.buffer);
-  }
-  return images;
-}
+const image = require('./shared/unpdf.970262b7.cjs');
+
+const decodePDFText = async (...args) => {
+  await image.resolvePDFJSNodeImports();
+  return await image.decodePDFText(...args);
+};
+const getImagesFromPage = async (...args) => {
+  await image.resolvePDFJSNodeImports();
+  return await image.getImagesFromPage(...args);
+};
 
 exports.decodePDFText = decodePDFText;
 exports.getImagesFromPage = getImagesFromPage;

package/dist/index.node.d.cts

@@ -1,14 +1,6 @@
-interface PDFContent {
-    totalPages: number;
-    info?: Record<string, any>;
-    metadata?: any;
-    text: string | string[];
-}
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.cjs';
 
-declare function decodePDFText(data: ArrayBuffer, { mergePages }?: {
-    mergePages?: boolean;
-}): Promise<PDFContent>;
-
-declare function getImagesFromPage(data: ArrayBuffer, pageNumber: number): Promise<ArrayBuffer[]>;
+declare const decodePDFText: typeof decodePDFText$1;
+declare const getImagesFromPage: typeof getImagesFromPage$1;
 
 export { decodePDFText, getImagesFromPage };

package/dist/index.node.d.mts

@@ -1,14 +1,6 @@
-interface PDFContent {
-    totalPages: number;
-    info?: Record<string, any>;
-    metadata?: any;
-    text: string | string[];
-}
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.mjs';
 
-declare function decodePDFText(data: ArrayBuffer, { mergePages }?: {
-    mergePages?: boolean;
-}): Promise<PDFContent>;
-
-declare function getImagesFromPage(data: ArrayBuffer, pageNumber: number): Promise<ArrayBuffer[]>;
+declare const decodePDFText: typeof decodePDFText$1;
+declare const getImagesFromPage: typeof getImagesFromPage$1;
 
 export { decodePDFText, getImagesFromPage };

package/dist/index.node.d.ts

@@ -1,14 +1,6 @@
-interface PDFContent {
-    totalPages: number;
-    info?: Record<string, any>;
-    metadata?: any;
-    text: string | string[];
-}
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.js';
 
-declare function decodePDFText(data: ArrayBuffer, { mergePages }?: {
-    mergePages?: boolean;
-}): Promise<PDFContent>;
-
-declare function getImagesFromPage(data: ArrayBuffer, pageNumber: number): Promise<ArrayBuffer[]>;
+declare const decodePDFText: typeof decodePDFText$1;
+declare const getImagesFromPage: typeof getImagesFromPage$1;
 
 export { decodePDFText, getImagesFromPage };

package/dist/index.node.mjs

@@ -1,64 +1,12 @@
-import { OPS } from 'pdfjs-dist';
-
-async function getDocumentProxy(data) {
-  const { getDocument } = await getPDFJSImports();
-  const pdf = await getDocument({
-    data,
-    useWorkerFetch: false,
-    useSystemFonts: true,
-    isEvalSupported: false
-  }).promise;
-  return pdf;
-}
-async function getPDFJSImports() {
-  try {
-    const { default: mod } = await import('pdfjs-dist/legacy/build/pdf');
-    const { getDocument, version } = mod;
-    return { getDocument, version };
-  } catch (e) {
-    console.error(e);
-    throw new Error(
-      "PDF.js is not available. Please run `pnpm add -D pdfjs-dist` and try again."
-    );
-  }
-}
-
-async function decodePDFText(data, { mergePages = false } = {}) {
-  const pdf = await getDocumentProxy(data);
-  const meta = await pdf.getMetadata().catch(() => null);
-  const texts = await Promise.all(
-    Array.from({ length: pdf.numPages }, (_, i) => getPageText(pdf, i + 1))
-  );
-  const filteredTexts = texts.filter(Boolean);
-  return {
-    totalPages: pdf.numPages,
-    info: meta?.info,
-    metadata: meta?.metadata?.getAll(),
-    text: mergePages ? filteredTexts.join("\n\n") : filteredTexts
-  };
-}
-async function getPageText(pdf, pageNumber) {
-  const page = await pdf.getPage(pageNumber);
-  const content = await page.getTextContent();
-  const items = content.items;
-  if (items.length === 0)
-    return;
-  return items.map(({ str }) => str).join("\n");
-}
-
-async function getImagesFromPage(data, pageNumber) {
-  const pdf = await getDocumentProxy(data);
-  const page = await pdf.getPage(pageNumber);
-  const operatorList = await page.getOperatorList();
-  const images = [];
-  for (const op of operatorList.fnArray) {
-    if (op !== OPS.paintImageXObject)
-      continue;
-    const image = await page.objs.get(operatorList.argsArray[op][0]);
-    if (image.data)
-      images.push(image.data.buffer);
-  }
-  return images;
-}
+import { a as resolvePDFJSNodeImports, d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.72a42ec0.mjs';
+
+const decodePDFText = async (...args) => {
+  await resolvePDFJSNodeImports();
+  return await decodePDFText$1(...args);
+};
+const getImagesFromPage = async (...args) => {
+  await resolvePDFJSNodeImports();
+  return await getImagesFromPage$1(...args);
+};
 
 export { decodePDFText, getImagesFromPage };

package/dist/index.web.cjs

@@ -1,10 +1,14 @@
 'use strict';
 
-const decodePDFText = async () => {
-  throw new Error("Not supported in browser context yet");
+const image = require('./shared/unpdf.970262b7.cjs');
+
+const decodePDFText = async (...args) => {
+  await image.resolvePDFJSWebImports();
+  return await image.decodePDFText(...args);
 };
-const getImagesFromPage = async () => {
-  throw new Error("Not supported in browser context yet");
+const getImagesFromPage = async (...args) => {
+  await image.resolvePDFJSWebImports();
+  return await image.getImagesFromPage(...args);
 };
 
 exports.decodePDFText = decodePDFText;

package/dist/index.web.d.cts

@@ -1,4 +1,4 @@
-import { decodePDFText as decodePDFText$1, getImagesFromPage as getImagesFromPage$1 } from './index.node.cjs';
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.cjs';
 
 declare const decodePDFText: typeof decodePDFText$1;
 declare const getImagesFromPage: typeof getImagesFromPage$1;

package/dist/index.web.d.mts

@@ -1,4 +1,4 @@
-import { decodePDFText as decodePDFText$1, getImagesFromPage as getImagesFromPage$1 } from './index.node.mjs';
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.mjs';
 
 declare const decodePDFText: typeof decodePDFText$1;
 declare const getImagesFromPage: typeof getImagesFromPage$1;

package/dist/index.web.d.ts

@@ -1,4 +1,4 @@
-import { decodePDFText as decodePDFText$1, getImagesFromPage as getImagesFromPage$1 } from './index.node.js';
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.js';
 
 declare const decodePDFText: typeof decodePDFText$1;
 declare const getImagesFromPage: typeof getImagesFromPage$1;

package/dist/index.web.mjs

@@ -1,8 +1,12 @@
-const decodePDFText = async () => {
-  throw new Error("Not supported in browser context yet");
+import { r as resolvePDFJSWebImports, d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.72a42ec0.mjs';
+
+const decodePDFText = async (...args) => {
+  await resolvePDFJSWebImports();
+  return await decodePDFText$1(...args);
 };
-const getImagesFromPage = async () => {
-  throw new Error("Not supported in browser context yet");
+const getImagesFromPage = async (...args) => {
+  await resolvePDFJSWebImports();
+  return await getImagesFromPage$1(...args);
 };
 
 export { decodePDFText, getImagesFromPage };

package/dist/index.worker.cjs

@@ -0,0 +1,11 @@
+'use strict';
+
+const decodePDFText = async () => {
+  throw new Error("Not implemented in worker context yet");
+};
+const getImagesFromPage = async () => {
+  throw new Error("Not implemented in worker context yet");
+};
+
+exports.decodePDFText = decodePDFText;
+exports.getImagesFromPage = getImagesFromPage;

package/dist/index.worker.d.cts

@@ -0,0 +1,6 @@
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.cjs';
+
+declare const decodePDFText: typeof decodePDFText$1;
+declare const getImagesFromPage: typeof getImagesFromPage$1;
+
+export { decodePDFText, getImagesFromPage };

package/dist/index.worker.d.mts

@@ -0,0 +1,6 @@
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.mjs';
+
+declare const decodePDFText: typeof decodePDFText$1;
+declare const getImagesFromPage: typeof getImagesFromPage$1;
+
+export { decodePDFText, getImagesFromPage };

package/dist/index.worker.d.ts

@@ -0,0 +1,6 @@
+import { d as decodePDFText$1, g as getImagesFromPage$1 } from './shared/unpdf.5c39ffe2.js';
+
+declare const decodePDFText: typeof decodePDFText$1;
+declare const getImagesFromPage: typeof getImagesFromPage$1;
+
+export { decodePDFText, getImagesFromPage };

package/dist/index.worker.mjs

@@ -0,0 +1,8 @@
+const decodePDFText = async () => {
+  throw new Error("Not implemented in worker context yet");
+};
+const getImagesFromPage = async () => {
+  throw new Error("Not implemented in worker context yet");
+};
+
+export { decodePDFText, getImagesFromPage };

package/dist/shared/unpdf.5c39ffe2.d.cts

@@ -0,0 +1,14 @@
+interface PDFContent {
+    totalPages: number;
+    info?: Record<string, any>;
+    metadata?: any;
+    text: string | string[];
+}
+
+declare function decodePDFText(data: ArrayBuffer, { mergePages }?: {
+    mergePages?: boolean;
+}): Promise<PDFContent>;
+
+declare function getImagesFromPage(data: ArrayBuffer, pageNumber: number): Promise<ArrayBuffer[]>;
+
+export { decodePDFText as d, getImagesFromPage as g };

package/dist/shared/unpdf.5c39ffe2.d.mts

@@ -0,0 +1,14 @@
+interface PDFContent {
+    totalPages: number;
+    info?: Record<string, any>;
+    metadata?: any;
+    text: string | string[];
+}
+
+declare function decodePDFText(data: ArrayBuffer, { mergePages }?: {
+    mergePages?: boolean;
+}): Promise<PDFContent>;
+
+declare function getImagesFromPage(data: ArrayBuffer, pageNumber: number): Promise<ArrayBuffer[]>;
+
+export { decodePDFText as d, getImagesFromPage as g };

package/dist/shared/unpdf.5c39ffe2.d.ts

@@ -0,0 +1,14 @@
+interface PDFContent {
+    totalPages: number;
+    info?: Record<string, any>;
+    metadata?: any;
+    text: string | string[];
+}
+
+declare function decodePDFText(data: ArrayBuffer, { mergePages }?: {
+    mergePages?: boolean;
+}): Promise<PDFContent>;
+
+declare function getImagesFromPage(data: ArrayBuffer, pageNumber: number): Promise<ArrayBuffer[]>;
+
+export { decodePDFText as d, getImagesFromPage as g };

package/dist/shared/unpdf.72a42ec0.mjs

@@ -0,0 +1,79 @@
+let instance;
+async function getDocumentProxy(data) {
+  const { getDocument } = instance;
+  const pdf = await getDocument({
+    data,
+    useWorkerFetch: false,
+    useSystemFonts: true,
+    isEvalSupported: false
+  }).promise;
+  return pdf;
+}
+function getResolvedPDFJSInstance() {
+  return instance;
+}
+async function resolvePDFJSWebImports() {
+  if (instance)
+    return;
+  try {
+    const { default: mod } = await import('pdfjs-dist');
+    instance = mod;
+  } catch (error) {
+    throw new Error(
+      "PDF.js is not available. Please add the package as a dependency."
+    );
+  }
+}
+async function resolvePDFJSNodeImports() {
+  if (instance)
+    return;
+  try {
+    const { default: mod } = await import('pdfjs-dist/legacy/build/pdf');
+    instance = mod;
+  } catch (error) {
+    throw new Error(
+      "PDF.js is not available. Please add the package as a dependency."
+    );
+  }
+}
+
+async function decodePDFText(data, { mergePages = false } = {}) {
+  const pdf = await getDocumentProxy(data);
+  const meta = await pdf.getMetadata().catch(() => null);
+  const texts = await Promise.all(
+    Array.from({ length: pdf.numPages }, (_, i) => getPageText(pdf, i + 1))
+  );
+  const filteredTexts = texts.filter(Boolean);
+  return {
+    totalPages: pdf.numPages,
+    info: meta?.info,
+    metadata: meta?.metadata?.getAll(),
+    text: mergePages ? filteredTexts.join("\n\n") : filteredTexts
+  };
+}
+async function getPageText(pdf, pageNumber) {
+  const page = await pdf.getPage(pageNumber);
+  const content = await page.getTextContent();
+  const items = content.items;
+  if (items.length === 0)
+    return;
+  return items.map(({ str }) => str).join("\n");
+}
+
+async function getImagesFromPage(data, pageNumber) {
+  const pdf = await getDocumentProxy(data);
+  const page = await pdf.getPage(pageNumber);
+  const operatorList = await page.getOperatorList();
+  const { OPS } = getResolvedPDFJSInstance();
+  const images = [];
+  for (const op of operatorList.fnArray) {
+    if (op !== OPS.paintImageXObject)
+      continue;
+    const image = await page.objs.get(operatorList.argsArray[op][0]);
+    if (image.data)
+      images.push(image.data.buffer);
+  }
+  return images;
+}
+
+export { resolvePDFJSNodeImports as a, decodePDFText as d, getImagesFromPage as g, resolvePDFJSWebImports as r };

package/dist/shared/unpdf.970262b7.cjs

@@ -0,0 +1,84 @@
+'use strict';
+
+let instance;
+async function getDocumentProxy(data) {
+  const { getDocument } = instance;
+  const pdf = await getDocument({
+    data,
+    useWorkerFetch: false,
+    useSystemFonts: true,
+    isEvalSupported: false
+  }).promise;
+  return pdf;
+}
+function getResolvedPDFJSInstance() {
+  return instance;
+}
+async function resolvePDFJSWebImports() {
+  if (instance)
+    return;
+  try {
+    const { default: mod } = await import('pdfjs-dist');
+    instance = mod;
+  } catch (error) {
+    throw new Error(
+      "PDF.js is not available. Please add the package as a dependency."
+    );
+  }
+}
+async function resolvePDFJSNodeImports() {
+  if (instance)
+    return;
+  try {
+    const { default: mod } = await import('pdfjs-dist/legacy/build/pdf');
+    instance = mod;
+  } catch (error) {
+    throw new Error(
+      "PDF.js is not available. Please add the package as a dependency."
+    );
+  }
+}
+
+async function decodePDFText(data, { mergePages = false } = {}) {
+  const pdf = await getDocumentProxy(data);
+  const meta = await pdf.getMetadata().catch(() => null);
+  const texts = await Promise.all(
+    Array.from({ length: pdf.numPages }, (_, i) => getPageText(pdf, i + 1))
+  );
+  const filteredTexts = texts.filter(Boolean);
+  return {
+    totalPages: pdf.numPages,
+    info: meta?.info,
+    metadata: meta?.metadata?.getAll(),
+    text: mergePages ? filteredTexts.join("\n\n") : filteredTexts
+  };
+}
+async function getPageText(pdf, pageNumber) {
+  const page = await pdf.getPage(pageNumber);
+  const content = await page.getTextContent();
+  const items = content.items;
+  if (items.length === 0)
+    return;
+  return items.map(({ str }) => str).join("\n");
+}
+
+async function getImagesFromPage(data, pageNumber) {
+  const pdf = await getDocumentProxy(data);
+  const page = await pdf.getPage(pageNumber);
+  const operatorList = await page.getOperatorList();
+  const { OPS } = getResolvedPDFJSInstance();
+  const images = [];
+  for (const op of operatorList.fnArray) {
+    if (op !== OPS.paintImageXObject)
+      continue;
+    const image = await page.objs.get(operatorList.argsArray[op][0]);
+    if (image.data)
+      images.push(image.data.buffer);
+  }
+  return images;
+}
+
+exports.decodePDFText = decodePDFText;
+exports.getImagesFromPage = getImagesFromPage;
+exports.resolvePDFJSNodeImports = resolvePDFJSNodeImports;
+exports.resolvePDFJSWebImports = resolvePDFJSWebImports;

package/package.json

@@ -1,9 +1,9 @@
 {
   "name": "unpdf",
   "type": "module",
-  "version": "0.2.1",
+  "version": "0.3.0",
   "packageManager": "pnpm@8.6.12",
-  "description": "Utilities to work with PDFs",
+  "description": "Utilities to work with PDFs, like extracting text",
   "author": "Johann Schopplich <pkg@johannschopplich.com>",
   "license": "MIT",
   "homepage": "https://github.com/johannschopplich/unpdf#readme",
@@ -21,16 +21,16 @@
   "sideEffects": false,
   "exports": {
     "browser": "./dist/index.web.mjs",
-    "bun": "./dist/index.web.mjs",
-    "deno": "./dist/index.web.mjs",
-    "edge-light": "./dist/index.web.mjs",
-    "edge-routine": "./dist/index.web.mjs",
-    "lagon": "./dist/index.web.mjs",
-    "netlify": "./dist/index.web.mjs",
-    "react-native": "./dist/index.web.mjs",
-    "wintercg": "./dist/index.web.mjs",
-    "worker": "./dist/index.web.mjs",
-    "workerd": "./dist/index.web.mjs",
+    "bun": "./dist/index.worker.mjs",
+    "deno": "./dist/index.worker.mjs",
+    "edge-light": "./dist/index.worker.mjs",
+    "edge-routine": "./dist/index.worker.mjs",
+    "lagon": "./dist/index.worker.mjs",
+    "netlify": "./dist/index.worker.mjs",
+    "react-native": "./dist/index.worker.mjs",
+    "wintercg": "./dist/index.worker.mjs",
+    "worker": "./dist/index.worker.mjs",
+    "workerd": "./dist/index.worker.mjs",
     "node": {
       "types": "./dist/index.node.d.ts",
       "import": "./dist/index.node.mjs",