Update test & examples

5 years ago · 0e968d1829
13 changed files with 117 additions and 45 deletions
--- a/examples/browser/basic.html
+++ b/examples/browser/basic.html
@ -6,7 +6,7 @@
				@@ -6,7 +6,7 @@
    <input type="file" id="uploader">
    <script>
      const recognize = async ({ target: { files }  }) => {
-        const { text } = await Tesseract.recognize(files[0], 'eng', {
+        const { data: { text } } = await Tesseract.recognize(files[0], 'eng', {
          corePath: '../../node_modules/tesseract.js-core/tesseract-core.wasm.js',
          logger: m => console.log(m),
        });
--- a/examples/browser/download-pdf.html
+++ b/examples/browser/download-pdf.html
@ -0,0 +1,52 @@
				@@ -0,0 +1,52 @@
+<html>
+  <head>
+    <script src="/dist/tesseract.dev.js"></script>
+  </head>
+  <body>
+    <div>
+      <input type="file" id="uploader">
+      <button id="download-pdf" disabled="true">Download PDF</button>
+    </div>
+    <textarea id="board" readonly rows="8" cols="80">Upload an image file</textarea>
+    <script>
+      const { createWorker } = Tesseract;
+      const worker = createWorker({
+        corePath: '/node_modules/tesseract.js-core/tesseract-core.wasm.js',
+        logger: m => console.log(m),
+      });
+      const uploader = document.getElementById('uploader');
+      const dlBtn = document.getElementById('download-pdf');
+      const recognize = async ({ target: { files }  }) => {
+        await worker.load();
+        await worker.loadLanguage('eng');
+        await worker.initialize('eng');
+        const { data: { text } } = await worker.recognize(files[0]);
+        const board = document.getElementById('board');
+        board.value = text;
+        dlBtn.disabled = false;
+      };
+      const downloadPDF = async () => {
+        const filename = 'tesseract-ocr-result.pdf';
+        const { data } = await worker.getPDF('Tesseract OCR Result');
+        const blob = new Blob([new Uint8Array(data)], { type: 'application/pdf' });
+        if (navigator.msSaveBlob) {
+          // IE 10+
+          navigator.msSaveBlob(blob, filename);
+        } else {
+          const link = document.createElement('a');
+          if (link.download !== undefined) {
+            const url = URL.createObjectURL(blob);
+            link.setAttribute('href', url);
+            link.setAttribute('download', filename);
+            link.style.visibility = 'hidden';
+            document.body.appendChild(link);
+            link.click();
+            document.body.removeChild(link);
+          }
+        }
+      };
+      uploader.addEventListener('change', recognize);
+      dlBtn.addEventListener('click', downloadPDF);
+    </script>
+  </body>
+</html>
--- a/examples/node/detect.js
+++ b/examples/node/detect.js
@ -8,6 +8,6 @@ const image = path.resolve(__dirname, (imagePath || '../../tests/assets/images/c
				@@ -8,6 +8,6 @@ const image = path.resolve(__dirname, (imagePath || '../../tests/assets/images/c
 console.log(`Recognizing ${image}`);

 Tesseract.detect(image, { logger: m => console.log(m) })
-  .then((data) => {
+  .then(({ data }) => {
    console.log(data);
  });
--- a/examples/node/download-pdf.js
+++ b/examples/node/download-pdf.js
@ -0,0 +1,22 @@
				@@ -0,0 +1,22 @@
+#!/usr/bin/env node
+const path = require('path');
+const fs = require('fs');
+const { createWorker } = require('../../');
+
+const [,, imagePath] = process.argv;
+const image = path.resolve(__dirname, (imagePath || '../../tests/assets/images/cosmic.png'));
+
+console.log(`Recognizing ${image}`);
+
+(async () => {
+  const worker = createWorker();
+  await worker.load();
+  await worker.loadLanguage('eng');
+  await worker.initialize('eng');
+  const { data: { text } } = await worker.recognize(image);
+  console.log(text);
+  const { data } = await worker.getPDF('Tesseract OCR Result');
+  fs.writeFileSync('tesseract-ocr-result.pdf', Buffer.from(data));
+  console.log('Generate PDF: tesseract-ocr-result.pdf');
+  await worker.terminate();
+})();
--- a/examples/node/recognize.js
+++ b/examples/node/recognize.js
@ -8,11 +8,6 @@ const image = path.resolve(__dirname, (imagePath || '../../tests/assets/images/c
				@@ -8,11 +8,6 @@ const image = path.resolve(__dirname, (imagePath || '../../tests/assets/images/c
 console.log(`Recognizing ${image}`);

 Tesseract.recognize(image, 'eng', { logger: m => console.log(m) })
-  .then(({ text }) => {
+  .then(({ data: { text } }) => {
    console.log(text);
  });
-
-//Tesseract.detect(image, { logger: m => console.log(m) })
-//  .then((data) => {
-//    console.log(data);
-//  });
--- a/src/Tesseract.js
+++ b/src/Tesseract.js
@ -1,29 +1,24 @@
				@@ -1,29 +1,24 @@
-const createScheduler = require('./createScheduler');
 const createWorker = require('./createWorker');

 const recognize = async (image, langs, options) => {
-  const scheduler = createScheduler();
  const worker = createWorker(options);
  await worker.load();
  await worker.loadLanguage(langs);
  await worker.initialize(langs);
-  scheduler.addWorker(worker);
-  return scheduler.addJob('recognize', image)
-    .finally(() => {
-      scheduler.terminate();
+  return worker.recognize(image)
+    .finally(async () => {
+      await worker.terminate();
    });
 };

 const detect = async (image, options) => {
-  const scheduler = createScheduler();
  const worker = createWorker(options);
  await worker.load();
  await worker.loadLanguage('osd');
  await worker.initialize('osd');
-  scheduler.addWorker(worker);
-  return scheduler.addJob('detect', image)
-    .finally(() => {
-      scheduler.terminate();
+  return worker.detect(image)
+    .finally(async () => {
+      await worker.terminate();
    });
 };

--- a/src/createScheduler.js
+++ b/src/createScheduler.js
@ -1,5 +1,5 @@
				@@ -1,5 +1,5 @@
 const createJob = require('./createJob');
-const log = require('./utils/log');
+const { log } = require('./utils/log');
 const getId = require('./utils/getId');

 let schedulerCounter = 0;
--- a/src/createWorker.js
+++ b/src/createWorker.js
@ -1,7 +1,7 @@
				@@ -1,7 +1,7 @@
 const resolvePaths = require('./utils/resolvePaths');
 const circularize = require('./utils/circularize');
 const createJob = require('./createJob');
-const log = require('./utils/log');
+const { log } = require('./utils/log');
 const getId = require('./utils/getId');
 const { defaultOEM } = require('./constants/config');
 const {
@ -106,9 +106,12 @@ module.exports = (_options = {}) => {
				@@ -106,9 +106,12 @@ module.exports = (_options = {}) => {
    }))
  );

-  const terminate = async () => {
+  const terminate = async (jobId) => {
    if (worker !== null) {
-      await startJob('terminate');
+      await startJob(createJob({
+        id: jobId,
+        action: 'terminate',
+      }));
      terminateWorker(worker);
      worker = null;
    }
--- a/src/index.js
+++ b/src/index.js
@ -14,6 +14,7 @@ const createJob = require('./createJob');
				@@ -14,6 +14,7 @@ const createJob = require('./createJob');
 const Tesseract = require('./Tesseract');
 const OEM = require('./constants/OEM');
 const PSM = require('./constants/PSM');
+const { setLogging } = require('./utils/log');

 module.exports = {
  OEM,
@ -21,5 +22,6 @@ module.exports = {
				@@ -21,5 +22,6 @@ module.exports = {
  createScheduler,
  createWorker,
  createJob,
+  setLogging,
  ...Tesseract,
 };
--- a/src/utils/log.js
+++ b/src/utils/log.js
@ -1,2 +1,9 @@
				@@ -1,2 +1,9 @@
-module.exports = (typeof process.env !== 'undefined' && process.env.NODE_ENV === 'development')
-  ? console.log : () => {};
+let logging = false;
+
+exports.logging = logging;
+
+exports.setLogging = (_logging) => {
+  logging = _logging;
+};
+
+exports.log = (...args) => (logging ? console.log.apply(this, args) : null);
--- a/src/worker-script/index.js
+++ b/src/worker-script/index.js
@ -15,7 +15,7 @@ const dump = require('./utils/dump');
				@@ -15,7 +15,7 @@ const dump = require('./utils/dump');
 const isBrowser = require('../utils/getEnvironment')('type') === 'browser';
 const setImage = require('./utils/setImage');
 const defaultParams = require('./constants/defaultParams');
-const log = require('../utils/log');
+const { log, setLogging } = require('../utils/log');

 /*
 * Tesseract Module returned by TesseractCore.
@ -29,7 +29,8 @@ let latestJob;
				@@ -29,7 +29,8 @@ let latestJob;
 let adapter = {};
 let params = defaultParams;

-const load = ({ workerId, jobId, payload: { options: { corePath } } }, res) => {
+const load = ({ workerId, jobId, payload: { options: { corePath, logging } } }, res) => {
+  setLogging(logging);
  if (!TessModule) {
    const Core = adapter.getCore(corePath, res);

--- a/tests/detect.test.js
+++ b/tests/detect.test.js
@ -1,7 +1,5 @@
				@@ -1,7 +1,5 @@
-const { createScheduler, createWorker } = Tesseract;
-const scheduler = createScheduler();
+const { createWorker } = Tesseract;
 const worker = createWorker(OPTIONS);
-scheduler.addWorker(worker);
 before(function cb() {
  this.timeout(0);
  return worker.load();
@ -14,7 +12,7 @@ describe('detect()', async () => {
				@@ -14,7 +12,7 @@ describe('detect()', async () => {
    ].forEach(async ({ name, ans: { script } }) => {
      await worker.loadLanguage('osd');
      await worker.initialize('osd');
-      const { data: { script: s } } = await scheduler.addJob('detect', `${IMAGE_PATH}/${name}`);
+      const { data: { script: s } } = await worker.detect(`${IMAGE_PATH}/${name}`);
      expect(s).to.be(script);
    });
  }).timeout(TIMEOUT);
--- a/tests/recognize.test.js
+++ b/tests/recognize.test.js
@ -1,7 +1,5 @@
				@@ -1,7 +1,5 @@
-const { createScheduler, createWorker, PSM } = Tesseract;
-const scheduler = createScheduler();
+const { createWorker, PSM } = Tesseract;
 const worker = createWorker(OPTIONS);
-scheduler.addWorker(worker);
 before(async function cb() {
  this.timeout(0);
  await worker.load();
@ -13,7 +11,7 @@ describe('recognize()', () => {
				@@ -13,7 +11,7 @@ describe('recognize()', () => {
    FORMATS.forEach(format => (
      it(`support ${format} format`, async () => {
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', `${IMAGE_PATH}/simple.${format}`);
+        const { data: { text } } = await worker.recognize(`${IMAGE_PATH}/simple.${format}`);
        expect(text).to.be(SIMPLE_TEXT);
      }).timeout(TIMEOUT)
    ));
@ -26,7 +24,7 @@ describe('recognize()', () => {
				@@ -26,7 +24,7 @@ describe('recognize()', () => {
    ].forEach(({ format, image, ans }) => (
      it(`recongize ${format} in base64`, async () => {
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', image);
+        const { data: { text } } = await worker.recognize(image);
        expect(text).to.be(ans);
      }).timeout(TIMEOUT)
    ));
@ -38,7 +36,7 @@ describe('recognize()', () => {
				@@ -38,7 +36,7 @@ describe('recognize()', () => {
    ].forEach(({ name, lang, ans }) => (
      it(`recongize ${lang}`, async () => {
        await worker.initialize(lang);
-        const { data: { text } } = await scheduler.addJob('recognize', `${IMAGE_PATH}/${name}`);
+        const { data: { text } } = await worker.recognize(`${IMAGE_PATH}/${name}`);
        expect(text).to.be(ans);
      }).timeout(TIMEOUT)
    ));
@ -52,7 +50,7 @@ describe('recognize()', () => {
				@@ -52,7 +50,7 @@ describe('recognize()', () => {
    ].forEach(({ name, desc, ans }) => (
      it(`recongize ${desc} image`, async () => {
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', `${IMAGE_PATH}/${name}`);
+        const { data: { text } } = await worker.recognize(`${IMAGE_PATH}/${name}`);
        expect(text).to.be(ans);
      }).timeout(TIMEOUT)
    ));
@ -68,8 +66,7 @@ describe('recognize()', () => {
				@@ -68,8 +66,7 @@ describe('recognize()', () => {
    }) => (
      it(`recongize half ${name}`, async () => {
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob(
-          'recognize',
+        const { data: { text } } = await worker.recognize(
          `${IMAGE_PATH}/${name}`,
          {
            rectangles: [
@ -94,7 +91,7 @@ describe('recognize()', () => {
				@@ -94,7 +91,7 @@ describe('recognize()', () => {
          await worker.setParameters({
            tessedit_pageseg_mode: mode,
          });
-          const { data } = await scheduler.addJob('recognize', `${IMAGE_PATH}/simple.png`);
+          const { data } = await worker.recognize(`${IMAGE_PATH}/simple.png`);
          expect(Object.keys(data).length).not.to.be(0);
        }).timeout(TIMEOUT)
      ));
@ -105,7 +102,7 @@ describe('recognize()', () => {
				@@ -105,7 +102,7 @@ describe('recognize()', () => {
      it(`support ${format} format`, async () => {
        const buf = fs.readFileSync(path.join(__dirname, 'assets', 'images', `simple.${format}`));
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', buf);
+        const { data: { text } } = await worker.recognize(buf);
        expect(text).to.be(SIMPLE_TEXT);
      }).timeout(TIMEOUT)
    ));
@ -117,7 +114,7 @@ describe('recognize()', () => {
				@@ -117,7 +114,7 @@ describe('recognize()', () => {
        const imageDOM = document.createElement('img');
        imageDOM.setAttribute('src', `${IMAGE_PATH}/simple.${format}`);
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', imageDOM);
+        const { data: { text } } = await worker.recognize(imageDOM);
        expect(text).to.be(SIMPLE_TEXT);
      }).timeout(TIMEOUT)
    ));
@ -129,7 +126,7 @@ describe('recognize()', () => {
				@@ -129,7 +126,7 @@ describe('recognize()', () => {
        const videoDOM = document.createElement('video');
        videoDOM.setAttribute('poster', `${IMAGE_PATH}/simple.${format}`);
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', videoDOM);
+        const { data: { text } } = await worker.recognize(videoDOM);
        expect(text).to.be(SIMPLE_TEXT);
      }).timeout(TIMEOUT)
    ));
@ -161,7 +158,7 @@ describe('recognize()', () => {
				@@ -161,7 +158,7 @@ describe('recognize()', () => {
    formats.forEach(format => (
      it(`support ${format} format`, async () => {
        await worker.initialize('eng');
-        const { data: { text } } = await scheduler.addJob('recognize', canvasDOM);
+        const { data: { text } } = await worker.recognize(canvasDOM);
        expect(text).to.be(SIMPLE_TEXT);
      }).timeout(TIMEOUT)
    ));