From 5fc5a9f9f18872cb0f1b54359338b2e189dd2bb1 Mon Sep 17 00:00:00 2001
From: ThibG <thib@sitedethib.com>
Date: Tue, 1 Sep 2020 00:26:10 +0200
Subject: Update Tesseract.js (#14708)

* [WiP] Update Tesseract.js

- Update Tesseract.js to 2.2.1
- Use versioned file names
- differentiate two progression states: preparing OCR and detecting picture

* Get rid of copy-webpack-plugin
---
 .../features/ui/components/focal_point_modal.js    | 49 ++++++++++++++++------
 1 file changed, 36 insertions(+), 13 deletions(-)

(limited to 'app')

diff --git a/app/javascript/mastodon/features/ui/components/focal_point_modal.js b/app/javascript/mastodon/features/ui/components/focal_point_modal.js
index 7348d9599..926a495d1 100644
--- a/app/javascript/mastodon/features/ui/components/focal_point_modal.js
+++ b/app/javascript/mastodon/features/ui/components/focal_point_modal.js
@@ -18,6 +18,8 @@ import { length } from 'stringz';
 import { Tesseract as fetchTesseract } from 'mastodon/features/ui/util/async-components';
 import GIFV from 'mastodon/components/gifv';
 import { me } from 'mastodon/initial_state';
+import tesseractCorePath from 'tesseract.js-core/tesseract-core.wasm.js';
+import tesseractWorkerPath from 'tesseract.js/dist/worker.min.js';
 
 const messages = defineMessages({
   close: { id: 'lightbox.close', defaultMessage: 'Close' },
@@ -104,6 +106,7 @@ class FocalPointModal extends ImmutablePureComponent {
     dirty: false,
     progress: 0,
     loading: true,
+    ocrStatus: '',
   };
 
   componentWillMount () {
@@ -219,11 +222,18 @@ class FocalPointModal extends ImmutablePureComponent {
 
     this.setState({ detecting: true });
 
-    fetchTesseract().then(({ TesseractWorker }) => {
-      const worker = new TesseractWorker({
-        workerPath: `${assetHost}/packs/ocr/worker.min.js`,
-        corePath: `${assetHost}/packs/ocr/tesseract-core.wasm.js`,
-        langPath: `${assetHost}/ocr/lang-data`,
+    fetchTesseract().then(({ createWorker }) => {
+      const worker = createWorker({
+        workerPath: tesseractWorkerPath,
+        corePath: tesseractCorePath,
+        langPath: assetHost,
+        logger: ({ status, progress }) => {
+          if (status === 'recognizing text') {
+            this.setState({ ocrStatus: 'detecting', progress });
+          } else {
+            this.setState({ ocrStatus: 'preparing', progress });
+          }
+        },
       });
 
       let media_url = media.get('url');
@@ -236,12 +246,18 @@ class FocalPointModal extends ImmutablePureComponent {
         }
       }
 
-      worker.recognize(media_url)
-        .progress(({ progress }) => this.setState({ progress }))
-        .finally(() => worker.terminate())
-        .then(({ text }) => this.setState({ description: removeExtraLineBreaks(text), dirty: true, detecting: false }))
-        .catch(() => this.setState({ detecting: false }));
-    }).catch(() => this.setState({ detecting: false }));
+      (async () => {
+        await worker.load();
+        await worker.loadLanguage('eng');
+        await worker.initialize('eng');
+        const { data: { text } } = await worker.recognize(media_url);
+        this.setState({ description: removeExtraLineBreaks(text), dirty: true, detecting: false });
+        await worker.terminate();
+      })();
+    }).catch((e) => {
+      console.error(e);
+      this.setState({ detecting: false });
+    });
   }
 
   handleThumbnailChange = e => {
@@ -261,7 +277,7 @@ class FocalPointModal extends ImmutablePureComponent {
 
   render () {
     const { media, intl, account, onClose, isUploadingThumbnail } = this.props;
-    const { x, y, dragging, description, dirty, detecting, progress } = this.state;
+    const { x, y, dragging, description, dirty, detecting, progress, ocrStatus } = this.state;
 
     const width  = media.getIn(['meta', 'original', 'width']) || null;
     const height = media.getIn(['meta', 'original', 'height']) || null;
@@ -282,6 +298,13 @@ class FocalPointModal extends ImmutablePureComponent {
       descriptionLabel = <FormattedMessage id='upload_form.description' defaultMessage='Describe for the visually impaired' />;
     }
 
+    let ocrMessage = '';
+    if (ocrStatus === 'detecting') {
+      ocrMessage = <FormattedMessage id='upload_modal.analyzing_picture' defaultMessage='Analyzing picture…' />;
+    } else {
+      ocrMessage = <FormattedMessage id='upload_modal.preparing_ocr' defaultMessage='Preparing OCR…' />;
+    }
+
     return (
       <div className='modal-root__modal report-modal' style={{ maxWidth: 960 }}>
         <div className='report-modal__target'>
@@ -333,7 +356,7 @@ class FocalPointModal extends ImmutablePureComponent {
               />
 
               <div className='setting-text__modifiers'>
-                <UploadProgress progress={progress * 100} active={detecting} icon='file-text-o' message={<FormattedMessage id='upload_modal.analyzing_picture' defaultMessage='Analyzing picture…' />} />
+                <UploadProgress progress={progress * 100} active={detecting} icon='file-text-o' message={ocrMessage} />
               </div>
             </div>
 
-- 
cgit