Add partial recongnize unit tests

.eslintrc

@@ -2,7 +2,8 @@
   "extends": "airbnb",
   "env": {
     "browser": true,
-    "node": true
+    "node": true,
+    "mocha": true
   },
   "rules": {
     "no-underscore-dangle": 0,

.gitignore

@@ -3,4 +3,6 @@ node_modules/*
 yarn.lock
 tesseract.dev.js
 worker.dev.js
-*.traineddata
+/*.traineddata
+/examples/**/*.traineddata
+.nyc_output

package.json

@@ -7,6 +7,7 @@
     "start": "node scripts/server.js",
     "build": "webpack --config scripts/webpack.config.prod.js",
     "prepublishOnly": "npm run build",
+    "test:node": "nyc mocha --exit --require ./scripts/test-helper.js ./tests/*.test.js",
     "lint": "eslint src"
   },
   "browser": {
@@ -23,7 +24,10 @@
     "eslint-plugin-import": "^2.14.0",
     "eslint-plugin-jsx-a11y": "^6.1.2",
     "eslint-plugin-react": "^7.11.1",
+    "expect.js": "^0.3.1",
     "express": "^4.16.4",
+    "mocha": "^5.2.0",
+    "nyc": "^13.1.0",
     "webpack": "^4.26.0",
     "webpack-cli": "^3.1.2",
     "webpack-dev-middleware": "^3.4.0"
@@ -31,7 +35,7 @@
   "dependencies": {
     "check-types": "^7.4.0",
     "is-url": "1.2.2",
-    "node-fetch": "^1.7.3",
+    "node-fetch": "^2.3.0",
     "resolve-url": "^0.2.1",
     "tesseract.js-core": "^2.0.0-beta.5",
     "tesseract.js-utils": "^1.0.0-beta.2"

scripts/server.js

@@ -7,9 +7,9 @@ const webpackConfig = require('./webpack.config.dev');
 const compiler = webpack(webpackConfig);
 const app = express();
 
-express.static.mime.types.wasm = 'application/wasm';
-
 app.use('/', express.static(path.resolve(__dirname, '..')));
 app.use(middleware(compiler, { publicPath: '/dist' }));
 
-app.listen(3000, () => console.log('Server is running on port 3000'));
+module.exports = app.listen(3000, () => {
+  console.log('Server is running on port 3000');
+});

scripts/test-helper.js

@@ -0,0 +1,17 @@
+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) => {
+  app.use('/', express.static(path.resolve(__dirname, '..')));
+  devServer = app.listen(3000, done);
+};
+
+global.stopServer = (done) => {
+  devServer.close(done);
+};

scripts/webpack.config.prod.js

@@ -7,7 +7,7 @@ const genConfig = ({
   devtool: 'source-map',
   entry,
   output: {
-    path: path.resolve(__dirname, 'dist'),
+    path: path.resolve(__dirname, '..', 'dist'),
     filename,
     library,
     libraryTarget,

src/common/worker.js

@@ -48,14 +48,21 @@ const handleInit = (req, res) => {
   return Promise.resolve();
 };
 
-const loadLanguage = (req) => {
-  const { options: { lang }, workerOptions: { langPath } } = req;
-  return loadLang({
+const loadLanguage = ({
+  options: { lang },
+  workerOptions: {
+    langPath, cachePath, cacheMethod, dataPath,
+  },
+}) => (
+  loadLang({
     langs: lang,
     tessModule: Module,
     langURI: langPath,
-  });
-};
+    cachePath,
+    cacheMethod,
+    dataPath,
+  })
+);
 
 const handleRecognize = (req, res) => (
   handleInit(req, res)

src/node/index.js

@@ -11,7 +11,7 @@ const readFile = util.promisify(fs.readFile);
 const loadImage = (imageURI) => {
   if (isURL(imageURI)) {
     return fetch(imageURI)
-      .then(resp => resp.buffer());
+      .then(resp => resp.arrayBuffer());
   }
   return readFile(imageURI);
 };

tests/.eslintrc

@@ -0,0 +1,5 @@
+{
+  "rules": {
+    "no-undef": 0
+  }
+}

tests/recognize.test.js

@@ -0,0 +1,100 @@
+const { TesseractWorker } = Tesseract;
+
+const IMAGE_PATH = 'http://localhost:3000/tests/assets/images';
+const cachePath = './tests/assets/traineddata';
+const cacheMethod = 'readOnly';
+const SIMPLE_TEXT = 'Tesseract.js\n';
+const COMSIC_TEXT = 'HellO World\nfrom beyond\nthe Cosmic Void\n';
+const TESTOCR_TEXT = 'This is a lot of 12 point text to test the\nocr code and see if it works on all types\nof file format.\n\nThe quick brown dog jumped over the\nlazy fox. The quick brown dog jumped\nover the lazy fox. The quick brown dog\njumped over the lazy fox. The quick\nbrown dog jumped over the lazy fox.\n';
+
+before((done) => {
+  if (typeof startServer !== 'undefined') {
+    startServer(done);
+  } else {
+    done();
+  }
+});
+
+after((done) => {
+  if (typeof stopServer !== 'undefined') {
+    stopServer(done);
+  }
+});
+
+describe('recognize()', () => {
+  describe('supports multiple formats', () => {
+    ['bmp', 'jpg', 'png', 'pbm'].forEach(format => (
+      it(`support ${format} format`, (done) => {
+        const worker = new TesseractWorker({ cachePath, cacheMethod });
+        worker
+          .recognize(`${IMAGE_PATH}/simple.${format}`)
+          .then((result) => {
+            expect(result.text).to.be(SIMPLE_TEXT);
+            worker.terminate();
+            done();
+          });
+      }).timeout(10000)
+    ));
+  });
+
+  describe('1 worker multiple recognition', () => {
+    [3, 10, 20].forEach(num => (
+      it(`recognize ${num} images with 1 worker`, (done) => {
+        const worker = new TesseractWorker({ cachePath, cacheMethod });
+        Promise.all(
+          Array(num).fill(0).map(() => worker.recognize(`${IMAGE_PATH}/simple.png`)),
+        ).then((results) => {
+          results.forEach(({ text }) => {
+            expect(text).to.be(SIMPLE_TEXT);
+          });
+          worker.terminate();
+          done();
+        });
+      }).timeout(30000)
+    ));
+  });
+
+  describe('should recognize in order', () => {
+    [1, 2].forEach((num) => {
+      it(`recognize ${num * 2} images with 1 worker in order`, (done) => {
+        const worker = new TesseractWorker({ cachePath, cacheMethod });
+        const cases = Array(num).fill(0)
+          .reduce(acc => (
+            acc.concat([
+              { name: 'simple.png', ans: SIMPLE_TEXT },
+              { name: 'cosmic.png', ans: COMSIC_TEXT },
+            ])
+          ),
+          []);
+        Promise.all(
+          cases.map(({ name }) => worker.recognize(`${IMAGE_PATH}/${name}`)),
+        ).then((results) => {
+          results.forEach(({ text }, idx) => {
+            expect(text).to.be(cases[idx].ans);
+          });
+          worker.terminate();
+          done();
+        });
+      }).timeout(30000);
+    });
+  });
+
+  describe('supports different complexity', () => {
+    [
+      { name: 'simple.png', desc: 'simple', ans: SIMPLE_TEXT },
+      { name: 'cosmic.png', desc: 'normal', ans: COMSIC_TEXT },
+      { name: 'testocr.png', desc: 'complex', ans: TESTOCR_TEXT },
+    ].forEach(({ name, desc, ans }) => (
+      it(`recongize ${desc} image`, (done) => {
+        const worker = new TesseractWorker({ cachePath, cacheMethod });
+        worker
+          .recognize(`${IMAGE_PATH}/${name}`)
+          .then(({ text }) => {
+            expect(text).to.be(ans);
+            worker.terminate();
+            done();
+          });
+      }).timeout(10000)
+    ));
+  });
+});