Minor edit

.eslintrc

@@ -1,6 +1,5 @@
 {
   "extends": "airbnb-base",
-  "parser": "babel-eslint",
   "env": {
     "browser": true,
     "node": true,
@@ -11,7 +10,6 @@
     "no-underscore-dangle": 0,
     "no-console": 0,
     "global-require": 0,
-    "camelcase": 0,
-    "no-control-regex": 0
+    "camelcase": 0
   }
 }

.github/workflows/node.js.yml

@@ -16,7 +16,7 @@ jobs:
 
     strategy:
       matrix:
-        node-version: [14.x, 16.x]
+        node-version: [10.x, 12.x, 14.x]
 
     steps:
     - uses: actions/checkout@v2

README.md

@@ -12,6 +12,13 @@
 [![Downloads Total](https://img.shields.io/npm/dt/tesseract.js.svg)](https://www.npmjs.com/package/tesseract.js)
 [![Downloads Month](https://img.shields.io/npm/dm/tesseract.js.svg)](https://www.npmjs.com/package/tesseract.js)
 
+<h3 align="center">
+  Version 2 is now available and under development in the master branch, read a story about v2: <a href="https://jeromewu.github.io/why-i-refactor-tesseract.js-v2/">Why I refactor tesseract.js v2?</a><br>
+  Check the <a href="https://github.com/naptha/tesseract.js/tree/support/1.x">support/1.x</a> branch for version 1
+</h3>
+
+<br>
+
 Tesseract.js is a javascript library that gets words in [almost any language](./docs/tesseract_lang_list.md) out of images. ([Demo](http://tesseract.projectnaptha.com/))
 
 Image Recognition
@@ -62,16 +69,6 @@ const worker = createWorker({
 
 [Check out the docs](#documentation) for a full explanation of the API.
 
-## Major changes in v3
-- Significantly faster performance
-   - Runtime reduction of 84% for Browser and 96% for Node.js when recognizing the [example images](./examples/data)
-- Upgrade to Tesseract v5.1.0 (using emscripten 3.1.18)
-- Added SIMD-enabled build for supported devices
-- Added support:
-   - Node.js version 18
-- Removed support:
-   - ASM.js version, any other old versions of Tesseract.js-core (<3.0.0) 
-   - Node.js versions 10 and 12
 
 ## Major changes in v2
 - Upgrade to tesseract v4.1.1 (using emscripten 1.39.10 upstream)
@@ -80,8 +77,7 @@ const worker = createWorker({
 - Support WebAssembly (fallback to ASM.js when browser doesn't support)
 - Support Typescript
 
-Read a story about v2: <a href="https://jeromewu.github.io/why-i-refactor-tesseract.js-v2/">Why I refactor tesseract.js v2?</a><br>
-  Check the <a href="https://github.com/naptha/tesseract.js/tree/support/1.x">support/1.x</a> branch for version 1
+
 ## Installation
 Tesseract.js works with a `<script>` tag via local copy or CDN, with webpack via `npm` and on Node.js with `npm/yarn`.
 
@@ -99,16 +95,16 @@ After including the script the `Tesseract` variable will be globally available.
 
 ### Node.js
 
-**Tesseract.js v3 requires Node.js v14 or higher**
+**Tesseract.js currently requires Node.js v6.8.0 or higher**
 
 ```shell
-# For v3
+# For v2
 npm install tesseract.js
 yarn add tesseract.js
 
-# For v2
-npm install tesseract.js@2
-yarn add tesseract.js@2
+# For v1
+npm install tesseract.js@1
+yarn add tesseract.js@1
 ```
 
 

docs/image-format.md

@@ -1,18 +1,17 @@
 # Image Format
 
-The main Tesseract.js functions (ex. recognize, detect) take an `image` parameter.  The image formats and data types supported are listed below. 
+Support Format: **bmp, jpg, png, pbm**
 
-Support Image Formats: **bmp, jpg, png, pbm, webp**
+The main Tesseract.js functions (ex. recognize, detect) take an `image` parameter, which should be something that is like an image. What's considered "image-like" differs depending on whether it is being run from the browser or through NodeJS.
 
-For browser and Node, supported data types are:
- - string with base64 encoded image (fits `data:image\/([a-zA-Z]*);base64,([^"]*)` regexp)
- - buffer
+On a browser, an image can be:
+- an `img`, `video`, or `canvas` element
+- a `File` object (from a file `<input>`)
+- a `Blob` object
+- a path or URL to an accessible image
+- a base64 encoded image fits `data:image\/([a-zA-Z]*);base64,([^"]*)` regexp
 
-For browser only, supported data types are:
- - `File` or `Blob` object
- - `img` or `canvas` element
-
-For Node only, supported data types are:
- - string containing a path to local image
-
-Note: images must be a supported image format **and** a supported data type.  For example, a buffer containing a png image is supported.  A buffer containing raw pixel data is not supported. 
+In Node.js, an image can be
+- a path to a local image
+- a Buffer storing binary image 
+- a base64 encoded image fits `data:image\/([a-zA-Z]*);base64,([^"]*)` regexp

examples/browser/demo.html

@@ -71,6 +71,7 @@ async function recognizeFile(file) {
 <option value='meme'     > Internet Meme                </option>
 <option value='epo'     > Esperanto             </option>
 <option value='epo_alt' > Esperanto alternative </option>
+<option value='equ'     > Math                  </option>
 <option value='est'     > Estonian              </option>
 <option value='eus'     > Basque                </option>
 <option value='fin'     > Finnish               </option>

package.json

@@ -1,6 +1,6 @@
 {
   "name": "tesseract.js",
-  "version": "3.0.3",
+  "version": "2.1.5",
   "description": "Pure Javascript Multilingual OCR",
   "main": "src/index.js",
   "types": "src/index.d.ts",
@@ -8,7 +8,7 @@
   "jsdelivr": "dist/tesseract.min.js",
   "scripts": {
     "start": "node scripts/server.js",
-    "build": "rimraf dist && webpack --config scripts/webpack.config.prod.js && rollup -c scripts/rollup.esm.js",
+    "build": "rimraf dist && webpack --config scripts/webpack.config.prod.js",
     "profile:tesseract": "webpack-bundle-analyzer dist/tesseract-stats.json",
     "profile:worker": "webpack-bundle-analyzer dist/worker-stats.json",
     "prepublishOnly": "npm run build",
@@ -35,12 +35,10 @@
   ],
   "license": "Apache-2.0",
   "devDependencies": {
-    "@babel/core": "^7.18.7",
-    "@babel/preset-env": "^7.18.7",
-    "@rollup/plugin-commonjs": "^22.0.2",
+    "@babel/core": "^7.7.7",
+    "@babel/preset-env": "^7.7.7",
     "acorn": "^6.4.0",
-    "babel-loader": "^8.2.0",
-    "buffer": "^6.0.3",
+    "babel-loader": "^8.1.0",
     "cors": "^2.8.5",
     "eslint": "^7.2.0",
     "eslint-config-airbnb-base": "^14.2.0",
@@ -52,26 +50,25 @@
     "npm-run-all": "^4.1.5",
     "nyc": "^15.1.0",
     "rimraf": "^2.7.1",
-    "rollup": "^2.79.0",
     "wait-on": "^3.3.0",
-    "webpack": "^5.74.0",
-    "webpack-bundle-analyzer": "^4.6.0",
-    "webpack-cli": "^4.10.0",
-    "webpack-dev-middleware": "^5.3.3"
+    "webpack": "^4.44.2",
+    "webpack-bundle-analyzer": "^3.6.0",
+    "webpack-cli": "^3.3.12",
+    "webpack-dev-middleware": "^3.7.2"
   },
   "dependencies": {
-    "babel-eslint": "^10.1.0",
+    "blueimp-load-image": "^3.0.0",
     "bmp-js": "^0.1.0",
     "file-type": "^12.4.1",
     "idb-keyval": "^3.2.0",
     "is-electron": "^2.2.0",
     "is-url": "^1.2.4",
+    "jpeg-autorotate": "^7.1.1",
     "node-fetch": "^2.6.0",
     "opencollective-postinstall": "^2.0.2",
     "regenerator-runtime": "^0.13.3",
     "resolve-url": "^0.2.1",
-    "tesseract.js-core": "^3.0.2",
-    "wasm-feature-detect": "^1.2.11",
+    "tesseract.js-core": "^2.2.0",
     "zlibjs": "^0.3.1"
   },
   "repository": {

scripts/rollup.esm.js

@@ -1,13 +0,0 @@
-import commonjs from "@rollup/plugin-commonjs";
-
-export default [
-  {
-    input: "dist/tesseract.min.js",
-    output: {
-      file: "dist/tesseract.esm.min.js",
-      format: "esm",
-      banner: "/* eslint-disable */",
-    },
-    plugins: [commonjs()],
-  },
-];

scripts/webpack.config.common.js

@@ -1,9 +1,4 @@
 module.exports = {
-  resolve: {
-    fallback: {
-      buffer: require.resolve('buffer/'),
-    },
-  },
   module: {
     rules: [
       {

scripts/webpack.config.dev.js

@@ -15,9 +15,6 @@ const genConfig = ({
     libraryTarget,
   },
   plugins: [
-    new webpack.ProvidePlugin({
-      Buffer: ['buffer', 'Buffer'],
-    }),
     new webpack.DefinePlugin({
       'process.env': {
         TESS_ENV: JSON.stringify('development'),

scripts/webpack.config.prod.js

@@ -1,6 +1,5 @@
 const path = require('path');
 const common = require('./webpack.config.common');
-const webpack = require('webpack');
 
 const genConfig = ({
   entry, filename, library, libraryTarget,
@@ -15,11 +14,6 @@ const genConfig = ({
     library,
     libraryTarget,
   },
-  plugins: [
-    new webpack.ProvidePlugin({
-      Buffer: ['buffer', 'Buffer'],
-    }),
-  ]
 });
 
 module.exports = [

src/constants/PSM.js

@@ -15,5 +15,4 @@ module.exports = {
   SINGLE_CHAR: '10',
   SPARSE_TEXT: '11',
   SPARSE_TEXT_OSD: '12',
-  RAW_LINE: '13',
 };

src/index.d.ts

@@ -19,18 +19,12 @@ declare namespace Tesseract {
     readText(path: string, jobId?: string): Promise<ConfigResult>
     removeText(path: string, jobId?: string): Promise<ConfigResult>
     FS(method: string, args: any[], jobId?: string): Promise<ConfigResult>
-    loadLanguage(langs?: string | Lang[], jobId?: string): Promise<ConfigResult>
-    initialize(langs?: string | Lang[], oem?: OEM, jobId?: string): Promise<ConfigResult>
+    loadLanguage(langs?: string, jobId?: string): Promise<ConfigResult>
+    initialize(langs?: string, oem?: OEM, jobId?: string): Promise<ConfigResult>
     setParameters(params: Partial<WorkerParams>, jobId?: string): Promise<ConfigResult>
     recognize(image: ImageLike, options?: Partial<RecognizeOptions>, jobId?: string): Promise<RecognizeResult>
     detect(image: ImageLike, jobId?: string): Promise<DetectResult>
     terminate(jobId?: string): Promise<ConfigResult>
-    getPDF(title?: string, textonly?: boolean, jobId?: string):Promise<GetPDFResult>
-  }
-
-  interface Lang {
-    code: string;
-    data: unknown;
   }
 
   interface WorkerOptions {
@@ -68,10 +62,6 @@ declare namespace Tesseract {
     jobId: string
     data: Page
   }
-  interface GetPDFResult {
-    jobId: string
-    data: number[]
-  }
   interface DetectResult {
     jobId: string
     data: DetectData
@@ -89,13 +79,13 @@ declare namespace Tesseract {
     width: number
     height: number
   }
-  enum OEM {
+  const enum OEM {
     TESSERACT_ONLY,
     LSTM_ONLY,
     TESSERACT_LSTM_COMBINED,
     DEFAULT,
   }
-  enum PSM {
+  const enum PSM {
     OSD_ONLY = '0',
     AUTO_OSD = '1',
     AUTO_ONLY = '2',
@@ -109,7 +99,6 @@ declare namespace Tesseract {
     SINGLE_CHAR = '10',
     SPARSE_TEXT = '11',
     SPARSE_TEXT_OSD = '12',
-    RAW_LINE = '13'
   }
   type ImageLike = string | HTMLImageElement | HTMLCanvasElement | HTMLVideoElement
     | CanvasRenderingContext2D | File | Blob | ImageData | Buffer;

src/utils/resolvePaths.js

@@ -4,7 +4,7 @@ const resolveURL = isBrowser ? require('resolve-url') : s => s; // eslint-disabl
 module.exports = (options) => {
   const opts = { ...options };
   ['corePath', 'workerPath', 'langPath'].forEach((key) => {
-    if (options[key]) {
+    if (typeof options[key] !== 'undefined') {
       opts[key] = resolveURL(opts[key]);
     }
   });

src/worker-script/browser/getCore.js

@@ -1,26 +1,15 @@
-const { simd } = require('wasm-feature-detect');
-const { dependencies } = require('../../../package.json');
-
-module.exports = async (corePath, res) => {
+module.exports = (corePath, res) => {
   if (typeof global.TesseractCore === 'undefined') {
     res.progress({ status: 'loading tesseract core', progress: 0 });
-
-    // If the user specifies a core path, we use that
-    // Otherwise, we detect the correct core based on SIMD support
-    let corePathImport = corePath;
-    if (!corePathImport) {
-      const simdSupport = await simd();
-      if (simdSupport) {
-        corePathImport = `https://unpkg.com/tesseract.js-core@v${dependencies['tesseract.js-core'].substring(1)}/tesseract-core-simd.wasm.js`;
-      } else {
-        corePathImport = `https://unpkg.com/tesseract.js-core@v${dependencies['tesseract.js-core'].substring(1)}/tesseract-core.wasm.js`;
-      }
-    }
-
-    global.importScripts(corePathImport);
-
+    global.importScripts(corePath);
+    /*
+     * Depending on whether the browser supports WebAssembly,
+     * the version of the TesseractCore will be different.
+     */
     if (typeof global.TesseractCoreWASM !== 'undefined' && typeof WebAssembly === 'object') {
       global.TesseractCore = global.TesseractCoreWASM;
+    } else if (typeof global.TesseractCoreASM !== 'undefined') {
+      global.TesseractCore = global.TesseractCoreASM;
     } else {
       throw Error('Failed to load TesseractCore');
     }

src/worker-script/index.js

@@ -28,10 +28,10 @@ let latestJob;
 let adapter = {};
 let params = defaultParams;
 
-const load = async ({ workerId, jobId, payload: { options: { corePath, logging } } }, res) => {
+const load = ({ workerId, jobId, payload: { options: { corePath, logging } } }, res) => {
   setLogging(logging);
   if (!TessModule) {
-    const Core = await adapter.getCore(corePath, res);
+    const Core = adapter.getCore(corePath, res);
 
     res.progress({ workerId, status: 'initializing tesseract', progress: 0 });
 
@@ -72,7 +72,7 @@ const loadLanguage = async ({
     },
   },
 },
-res) => {
+  res) => {
   const loadAndGunzipFile = async (_lang) => {
     const lang = typeof _lang === 'string' ? _lang : _lang.code;
     const readCache = ['refresh', 'none'].includes(cacheMethod)
@@ -144,7 +144,15 @@ res) => {
     res.progress({ workerId, status: 'loaded language traineddata', progress: 1 });
     res.resolve(langs);
   } catch (err) {
-    res.reject(err.toString());
+    if (isWebWorker && err instanceof DOMException) {
+      /*
+       * For some reason google chrome throw DOMException in loadLang,
+       * while other browser is OK, for now we ignore this exception
+       * and hopefully to find the root cause one day.
+       */
+    } else {
+      res.reject(err.toString());
+    }
   }
 };
 
@@ -177,10 +185,7 @@ const initialize = ({
       api.End();
     }
     api = new TessModule.TessBaseAPI();
-    const status = api.Init(null, langs, oem);
-    if (status === -1) {
-      res.reject('initialization failed');
-    }
+    api.Init(null, langs, oem);
     params = defaultParams;
     setParameters({ payload: { params } });
     res.progress({

src/worker-script/node/getCore.js

@@ -1,19 +1,12 @@
-const { simd } = require('wasm-feature-detect');
-
 let TesseractCore = null;
 /*
  * getCore is a sync function to load and return
  * TesseractCore.
  */
-module.exports = async (_, res) => {
+module.exports = (_, res) => {
   if (TesseractCore === null) {
-    const simdSupport = await simd();
     res.progress({ status: 'loading tesseract core', progress: 0 });
-    if (simdSupport) {
-      TesseractCore = require('tesseract.js-core/tesseract-core-simd');
-    } else {
-      TesseractCore = require('tesseract.js-core/tesseract-core');
-    }
+    TesseractCore = require('tesseract.js-core');
     res.progress({ status: 'loaded tesseract core', progress: 1 });
   }
   return TesseractCore;

src/worker-script/utils/setImage.js

@@ -17,12 +17,10 @@ module.exports = (TessModule, api, image) => {
   let w = 0;
   let h = 0;
 
-  const exif = buf.slice(0, 500).toString().match(/\x01\x12\x00\x03\x00\x00\x00\x01\x00(.)/)?.[1]?.charCodeAt(0) || 1;
-
   /*
-   * Leptonica supports uncompressed but not compressed bmp files
-   * @see https://github.com/DanBloomberg/leptonica/issues/607#issuecomment-1068802516
-   * We therefore use bmp-js to process all bmp files
+   * Although leptonica should support reading bmp, there is a bug of "compressed BMP files".
+   * As there is no solution, we need to use bmp-js for now.
+   * @see https://groups.google.com/forum/#!topic/tesseract-ocr/4mPD9zTxdxE
    */
   if (type && type.mime === 'image/bmp') {
     const bmpBuf = bmp.decode(buf);
@@ -55,9 +53,9 @@ module.exports = (TessModule, api, image) => {
    *
    */
   if (data === null) {
-    api.SetImage(pix, undefined, undefined, undefined, undefined, exif);
+    api.SetImage(pix);
   } else {
-    api.SetImage(data, w, h, bytesPerPixel, w * bytesPerPixel, exif);
+    api.SetImage(data, w, h, bytesPerPixel, w * bytesPerPixel);
   }
   return data === null ? pix : data;
 };

src/worker/browser/defaultOptions.js

@@ -1,5 +1,5 @@
 const resolveURL = require('resolve-url');
-const { version } = require('../../../package.json');
+const { version, dependencies } = require('../../../package.json');
 const defaultOptions = require('../../constants/defaultOptions');
 
 /*
@@ -14,5 +14,5 @@ module.exports = {
    * If browser doesn't support WebAssembly,
    * load ASM version instead
    */
-  corePath: null,
+  corePath: `https://unpkg.com/tesseract.js-core@v${dependencies['tesseract.js-core'].substring(1)}/tesseract-core.${typeof WebAssembly === 'object' ? 'wasm' : 'asm'}.js`,
 };

src/worker/browser/loadImage.js

@@ -1,4 +1,5 @@
 const resolveURL = require('resolve-url');
+const blueimpLoadImage = require('blueimp-load-image');
 
 /**
  * readFromBlobOrFile
@@ -20,6 +21,19 @@ const readFromBlobOrFile = (blob) => (
   })
 );
 
+const fixOrientationFromUrlOrBlobOrFile = (blob) => (
+  new Promise((resolve) => {
+    blueimpLoadImage(
+      blob,
+      (img) => img.toBlob(resolve),
+      {
+        orientation: true,
+        canvas: true,
+      },
+    );
+  })
+);
+
 /**
  * loadImage
  *
@@ -34,14 +48,18 @@ const loadImage = async (image) => {
   }
 
   if (typeof image === 'string') {
-    // Base64 Image
-    if (/data:image\/([a-zA-Z]*);base64,([^"]*)/.test(image)) {
-      data = atob(image.split(',')[1])
-        .split('')
-        .map((c) => c.charCodeAt(0));
-    } else {
+    if (image.endsWith('.pbm')) {
       const resp = await fetch(resolveURL(image));
       data = await resp.arrayBuffer();
+    } else {
+      let img = image;
+      // If not Base64 Image
+      if (!/data:image\/([a-zA-Z]*);base64,([^"]*)/.test(image)) {
+        img = resolveURL(image);
+      }
+      data = await readFromBlobOrFile(
+        await fixOrientationFromUrlOrBlobOrFile(img),
+      );
     }
   } else if (image instanceof HTMLElement) {
     if (image.tagName === 'IMG') {
@@ -59,7 +77,11 @@ const loadImage = async (image) => {
       });
     }
   } else if (image instanceof File || image instanceof Blob) {
-    data = await readFromBlobOrFile(image);
+    let img = image;
+    if (!image.name.endsWith('.pbm')) {
+      img = await fixOrientationFromUrlOrBlobOrFile(img);
+    }
+    data = await readFromBlobOrFile(img);
   }
 
   return new Uint8Array(data);

src/worker/node/loadImage.js

@@ -2,6 +2,7 @@ const util = require('util');
 const fs = require('fs');
 const fetch = require('node-fetch');
 const isURL = require('is-url');
+const jo = require('jpeg-autorotate');
 
 const readFile = util.promisify(fs.readFile);
 
@@ -31,5 +32,9 @@ module.exports = async (image) => {
     data = image;
   }
 
+  try {
+    data = (await jo.rotate(data, { quality: 100 })).buffer;
+  } catch (_) {} /* eslint-disable-line */
+
   return new Uint8Array(data);
 };