tesseract.js - npm Package Compare versions

Comparing version 2.0.0-alpha.3 to 2.0.0-alpha.4

docs/api.md

@@ -16,3 +16,3 @@ # API
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 worker
@@ -27,3 +27,3 @@ .recognize(myImage)
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 // if we know our image is of spanish words without the letter 'e':
@@ -48,3 +48,3 @@ worker
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 worker
@@ -63,3 +63,3 @@ .detect(myImage)
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 worker.recognize(myImage)
@@ -74,3 +74,3 @@ .progress(message => console.log(message))
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 const job1 = worker.recognize(myImage);
@@ -95,3 +95,3 @@
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 worker.recognize(myImage)
@@ -122,3 +122,3 @@ .progress(function(message){console.log('progress is: ', message)});
 ```javascript
-const worker = new Tessearct.TesseractWorker();
+const worker = new Tesseract.TesseractWorker();
 worker.recognize(myImage)
@@ -125,0 +125,0 @@ .then(function(result){console.log('result is: ', result)});

docs/examples.md

@@ -11,3 +11,3 @@ # Tesseract.js Examples
 const { TesseractWorker } = Tesseract;
-const worker = new TessearctWorker();
+const worker = new TesseractWorker();
 
@@ -27,3 +27,3 @@ worker
 const { TesseractWorker } = Tesseract;
-const worker = new TessearctWorker();
+const worker = new TesseractWorker();
 
@@ -40,3 +40,3 @@ worker
 
-### with multiple languages (separate by '+'')
+### with multiple languages, separate by '+'
 
@@ -47,3 +47,3 @@ ```javascript
 const { TesseractWorker } = Tesseract;
-const worker = new TessearctWorker();
+const worker = new TesseractWorker();
 
@@ -62,1 +62,28 @@ worker
 ```
+
+### with whitelist chars (^2.0.0-alpha.4)
+
+Sadly, whitelist chars is not supported in tesseract.js v4, so in tesseract.js we need to switch to tesseract v3 mode to make it work.
+
+```javascript
+import Tesseract from 'tesseract.js';
+
+const { TesseractWorker, OEM } = Tesseract;
+const worker = new TesseractWorker();
+
+worker
+  .recognize(
+    'http://jeroen.github.io/images/testocr.png',
+    'eng',
+    {
+      'init_oem': OEM.TESSERACT_ONLY,
+      'tessedit_char_whitelist': '0123456789-.',
+    }
+  )
+  .progress((p) => {
+    console.log('progress', p);
+  })
+  .then((result) => {
+    console.log(result);
+  });
+```

docs/local-installation.md

 ## Local Installation
 
+Check here for an example: https://github.com/jeromewu/tesseract.js-offline
+
 In browser environment, `tesseract.js` simply provides the API layer. Internally, it opens a WebWorker to handle requests. That worker itself loads code from the Emscripten-built `tesseract.js-core` which itself is hosted on a CDN. Then it dynamically loads language files hosted on another CDN.
 
-Because of this we recommend loading `tesseract.js` from a CDN. But if you really need to have all your files local, you can pass extra arguments to `TessearctWorker` to specify custom paths for workers, languages, and core.
+Because of this we recommend loading `tesseract.js` from a CDN. But if you really need to have all your files local, you can pass extra arguments to `TesseractWorker` to specify custom paths for workers, languages, and core.
 
@@ -11,5 +13,5 @@ In Node.js environment, the only path you may want to customize is languages/langPath.
 const worker = Tesseract.TesseractWorker({
-  workerPath: 'https://unpkg.com/tesseract.js@v2.0.0-alpha.3/dist/worker.min.js',
+  workerPath: 'https://unpkg.com/tesseract.js@v2.0.0-alpha.4/dist/worker.min.js',
   langPath: 'https://tessdata.projectnaptha.com/4.0.0',
-  corePath: 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.5/tesseract-core.js',
+  corePath: 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.wasm.js',
 });
@@ -25,2 +27,4 @@ ```
 ### corePath
-A string specifying the location of the [tesseract.js-core library](https://github.com/naptha/tesseract.js-core), with default value 'https://cdn.jsdelivr.net/gh/naptha/tesseract.js-core@v2.0.0-beta.5/tesseract-core.js'.
+A string specifying the location of the [tesseract.js-core library](https://github.com/naptha/tesseract.js-core), with default value 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.wasm.js' (fallback to tesseract-core.asm.js when WebAssembly is not available).
+
+Another WASM option is 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.js' which is a script that loads 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.wasm'. But it fails to fetch at this moment.

package.json

 {
   "name": "tesseract.js",
-  "version": "2.0.0-alpha.3",
+  "version": "2.0.0-alpha.4",
   "description": "Pure Javascript Multilingual OCR",
@@ -10,4 +10,9 @@ "main": "src/index.js",
     "prepublishOnly": "npm run build",
-    "test": "npm run test:node",
+    "wait": "wait-on http://localhost:3000/package.json",
+    "test": "npm-run-all -p -r start test:all",
+    "test:all": "npm-run-all wait test:browser:* test:node",
     "test:node": "nyc mocha --exit --bail --require ./scripts/test-helper.js ./tests/*.test.js",
+    "test:browser-tpl": "mocha-headless-chrome -a incognito -a no-sandbox -a disable-setuid-sandbox -t 300000",
+    "test:browser:detect": "npm run test:browser-tpl -- -f ./tests/browser/detect.test.html",
+    "test:browser:recognize": "npm run test:browser-tpl -- -f ./tests/browser/recognize.test.html",
     "lint": "eslint src"
@@ -24,2 +29,3 @@ },
   "devDependencies": {
+    "cors": "^2.8.5",
     "eslint": "^5.9.0",
@@ -33,3 +39,6 @@ "eslint-config-airbnb": "^17.1.0",
     "mocha": "^5.2.0",
+    "mocha-headless-chrome": "^2.0.2",
+    "npm-run-all": "^4.1.5",
     "nyc": "^13.1.0",
+    "wait-on": "^3.2.0",
     "webpack": "^4.26.0",
@@ -44,3 +53,3 @@ "webpack-cli": "^3.1.2",
     "resolve-url": "^0.2.1",
-    "tesseract.js-core": "^2.0.0-beta.5",
+    "tesseract.js-core": "^2.0.0-beta.8",
     "tesseract.js-utils": "^1.0.0-beta.5"
@@ -47,0 +56,0 @@ },

README.md

@@ -12,7 +12,7 @@ # [Tesseract.js](http://tesseract.projectnaptha.com/)
 
-**Tessearct.js v2 is now available and under development in master branch, check [support/1.x](https://github.com/naptha/tesseract.js/tree/support/1.x) branch for v1.**
+**Tesseract.js v2 is now available and under development in master branch, check [support/1.x](https://github.com/naptha/tesseract.js/tree/support/1.x) branch for v1.**
 
 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/))
 
-[![fancy demo gif](./docs/demo.gif)](http://tesseract.projectnaptha.com)
+[![fancy demo gif](./docs/images/demo.gif)](http://tesseract.projectnaptha.com)
 
@@ -43,3 +43,3 @@ Tesseract.js works with script tags, [webpack](https://webpack.js.org/), and [Node.js](https://nodejs.org/en/). [After you install it](#installation), using it is as simple as
 ```html
-<script src='https://unpkg.com/tesseract.js@v2.0.0-alpha.3/dist/tesseract.min.js'></script>
+<script src='https://unpkg.com/tesseract.js@v2.0.0-alpha.4/dist/tesseract.min.js'></script>
 ```
@@ -85,2 +85,3 @@
 * [Local Installation](./docs/local-installation.md)
+* [FAQ](./docs/faq.md)
 
@@ -87,0 +88,0 @@ # Contributing

scripts/server.js

@@ -5,2 +5,3 @@ const webpack = require('webpack');
 const path = require('path');
+const cors = require('cors');
 const webpackConfig = require('./webpack.config.dev');
@@ -11,2 +12,3 @@
 
+app.use(cors());
 app.use('/', express.static(path.resolve(__dirname, '..')));
@@ -13,0 +15,0 @@ app.use(middleware(compiler, { publicPath: '/dist' }));

scripts/test-helper.js

@@ -1,26 +0,3 @@
-const express = require('express');
-const path = require('path');
 global.expect = require('expect.js');
 global.fetch = require('node-fetch');
 global.Tesseract = require('../src');
-
-const app = express();
-let devServer = null;
-
-global.startServer = (done) => {
-  if (devServer === null) {
-    app.use('/', express.static(path.resolve(__dirname, '..')));
-    devServer = app.listen(3000, done);
-  } else {
-    done();
-  }
-};
-
-global.stopServer = (done) => {
-  if (devServer !== null) {
-    devServer.close(done);
-    devServer = null;
-  } else {
-    done();
-  }
-};

src/browser/index.js

@@ -86,3 +86,3 @@ /**
    */
-  corePath: `https://unpkg.com/tesseract.js-core@v2.0.0-beta.5/tesseract-core${typeof WebAssembly === 'object' ? '' : '.asm'}.js`,
+  corePath: `https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.${typeof WebAssembly === 'object' ? 'wasm' : 'asm'}.js`,
 };
@@ -102,5 +102,7 @@
   let worker;
-  if (window.Blob && window.URL) {
-    const blob = new Blob([`importScripts("${workerPath}");`]);
-    worker = new Worker(window.URL.createObjectURL(blob));
+  if (Blob && URL) {
+    const blob = new Blob([`importScripts("${workerPath}");`], {
+      type: 'application/javascript',
+    });
+    worker = new Worker(URL.createObjectURL(blob));
   } else {
@@ -107,0 +109,0 @@ worker = new Worker(workerPath);

src/browser/worker.js

@@ -34,3 +34,9 @@ /**
        */
-      global.TesseractCore = typeof WebAssembly === 'object' ? global.TesseractCoreWASM : global.TesseractCoreASM;
+      if (check.not.undefined(global.TesseractCoreWASM) && typeof WebAssembly === 'object') {
+        global.TesseractCore = global.TesseractCoreWASM;
+      } else if (check.not.undefined(global.TesseractCoreASM)){
+        global.TesseractCore = global.TesseractCoreASM;
+      } else {
+        throw Error('Failed to load TesseractCore');
+      }
       res.progress({ status: 'loading tesseract core', progress: 1 });
@@ -37,0 +43,0 @@ }

src/common/workerUtils.js

@@ -11,2 +11,3 @@ /**
 const { readImage, loadLang } = require('tesseract.js-utils');
+const check = require('check-types');
 const dump = require('./dump');
@@ -126,2 +127,5 @@
         .then(() => {
+          const OEM = check.undefined(params['init_oem'])
+            ? TessModule.OEM_DEFAULT
+            : params['init_oem'];
           const progressUpdate = (progress) => {
@@ -131,5 +135,5 @@ res.progress({ status: 'initializing api', progress });
           progressUpdate(0);
-          api.Init(null, lang);
+          api.Init(null, lang, OEM);
           progressUpdate(0.3);
-          Object.keys(params).forEach((key) => {
+          Object.keys(params).filter(key => !key.startsWith('init_')).forEach((key) => {
             api.SetVariable(key, params[key]);
@@ -136,0 +140,0 @@ });

src/index.js

@@ -12,2 +12,3 @@ /**
 const TesseractWorker = require('./common/TesseractWorker');
+const { OEM } = require('./common/types');
 
@@ -19,2 +20,4 @@ module.exports = {
   utils,
+  /** Check ./common/types for more details */
+  OEM,
 };

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

1346961

increased by17.64%

Number of package files

51

increased by37.84%

Lines of code

1933

increased by0.89%

Number of lines in readme file

115

increased by0.88%

Worsened metrics

Dev dependency count

16

increased by33.33%

Dependency changes

• Updatedtesseract.js-core@^2.0.0-beta.8