优化后端

2026-09-11 17:56:56 +08:00
parent 7eed11b774
commit 812e67422d
7299 changed files with 1744959 additions and 112 deletions
@@ -0,0 +1,12 @@
+'use strict';
+
+const version = require('../../../package.json').version;
+const defaultOptions = require('../../constants/defaultOptions');
+
+/*
+ * Default options for browser worker
+ */
+module.exports = {
+  ...defaultOptions,
+  workerPath: `https://cdn.jsdelivr.net/npm/tesseract.js@v${version}/dist/worker.min.js`,
+};
@@ -0,0 +1,26 @@
+'use strict';
+
+/**
+ *
+ * Tesseract Worker adapter for browser
+ *
+ * @fileoverview Tesseract Worker adapter for browser
+ * @author Kevin Kwok <antimatter15@gmail.com>
+ * @author Guillermo Webster <gui@mit.edu>
+ * @author Jerome Wu <jeromewus@gmail.com>
+ */
+const defaultOptions = require('./defaultOptions');
+const spawnWorker = require('./spawnWorker');
+const terminateWorker = require('./terminateWorker');
+const onMessage = require('./onMessage');
+const send = require('./send');
+const loadImage = require('./loadImage');
+
+module.exports = {
+  defaultOptions,
+  spawnWorker,
+  terminateWorker,
+  onMessage,
+  send,
+  loadImage,
+};
@@ -0,0 +1,71 @@
+'use strict';
+
+/**
+ * readFromBlobOrFile
+ *
+ * @name readFromBlobOrFile
+ * @function
+ * @access private
+ */
+const readFromBlobOrFile = (blob) => (
+  new Promise((resolve, reject) => {
+    const fileReader = new FileReader();
+    fileReader.onload = () => {
+      resolve(fileReader.result);
+    };
+    fileReader.onerror = ({ target: { error: { code } } }) => {
+      reject(Error(`File could not be read! Code=${code}`));
+    };
+    fileReader.readAsArrayBuffer(blob);
+  })
+);
+
+/**
+ * loadImage
+ *
+ * @name loadImage
+ * @function load image from different source
+ * @access private
+ */
+const loadImage = async (image) => {
+  let data = image;
+  if (typeof image === 'undefined') {
+    return 'undefined';
+  }
+
+  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 {
+      const resp = await fetch(image);
+      data = await resp.arrayBuffer();
+    }
+  } else if (typeof HTMLElement !== 'undefined' && image instanceof HTMLElement) {
+    if (image.tagName === 'IMG') {
+      data = await loadImage(image.src);
+    }
+    if (image.tagName === 'VIDEO') {
+      data = await loadImage(image.poster);
+    }
+    if (image.tagName === 'CANVAS') {
+      await new Promise((resolve) => {
+        image.toBlob(async (blob) => {
+          data = await readFromBlobOrFile(blob);
+          resolve();
+        });
+      });
+    }
+  } else if (typeof OffscreenCanvas !== 'undefined' && image instanceof OffscreenCanvas) {
+    const blob = await image.convertToBlob();
+    data = await readFromBlobOrFile(blob);
+  } else if (image instanceof File || image instanceof Blob) {
+    data = await readFromBlobOrFile(image);
+  }
+
+  return new Uint8Array(data);
+};
+
+module.exports = loadImage;
@@ -0,0 +1,7 @@
+'use strict';
+
+module.exports = (worker, handler) => {
+  worker.onmessage = ({ data }) => { // eslint-disable-line
+    handler(data);
+  };
+};
@@ -0,0 +1,12 @@
+'use strict';
+
+/**
+ * send
+ *
+ * @name send
+ * @function send packet to worker and create a job
+ * @access public
+ */
+module.exports = async (worker, packet) => {
+  worker.postMessage(packet);
+};
@@ -0,0 +1,22 @@
+'use strict';
+
+/**
+ * spawnWorker
+ *
+ * @name spawnWorker
+ * @function create a new Worker in browser
+ * @access public
+ */
+module.exports = ({ workerPath, workerBlobURL }) => {
+  let worker;
+  if (Blob && URL && workerBlobURL) {
+    const blob = new Blob([`importScripts("${workerPath}");`], {
+      type: 'application/javascript',
+    });
+    worker = new Worker(URL.createObjectURL(blob));
+  } else {
+    worker = new Worker(workerPath);
+  }
+
+  return worker;
+};
@@ -0,0 +1,12 @@
+'use strict';
+
+/**
+ * terminateWorker
+ *
+ * @name terminateWorker
+ * @function terminate worker
+ * @access public
+ */
+module.exports = (worker) => {
+  worker.terminate();
+};
@@ -0,0 +1,12 @@
+'use strict';
+
+const path = require('path');
+const defaultOptions = require('../../constants/defaultOptions');
+
+/*
+ * Default options for node worker
+ */
+module.exports = {
+  ...defaultOptions,
+  workerPath: path.join(__dirname, '..', '..', 'worker-script', 'node', 'index.js'),
+};
@@ -0,0 +1,26 @@
+'use strict';
+
+/**
+ *
+ * Tesseract Worker impl. for node (using child_process)
+ *
+ * @fileoverview Tesseract Worker impl. for node
+ * @author Kevin Kwok <antimatter15@gmail.com>
+ * @author Guillermo Webster <gui@mit.edu>
+ * @author Jerome Wu <jeromewus@gmail.com>
+ */
+const defaultOptions = require('./defaultOptions');
+const spawnWorker = require('./spawnWorker');
+const terminateWorker = require('./terminateWorker');
+const onMessage = require('./onMessage');
+const send = require('./send');
+const loadImage = require('./loadImage');
+
+module.exports = {
+  defaultOptions,
+  spawnWorker,
+  terminateWorker,
+  onMessage,
+  send,
+  loadImage,
+};
@@ -0,0 +1,38 @@
+'use strict';
+
+const util = require('util');
+const fs = require('fs');
+// Use built-in fetch if available, otherwise fallback to node-fetch
+const fetch = global.fetch || require('node-fetch');
+const isURL = require('is-url');
+
+const readFile = util.promisify(fs.readFile);
+
+/**
+ * loadImage
+ *
+ * @name loadImage
+ * @function load image from different source
+ * @access public
+ */
+module.exports = async (image) => {
+  let data = image;
+  if (typeof image === 'undefined') {
+    return image;
+  }
+
+  if (typeof image === 'string') {
+    if (isURL(image) || image.startsWith('moz-extension://') || image.startsWith('chrome-extension://') || image.startsWith('file://')) {
+      const resp = await fetch(image);
+      data = await resp.arrayBuffer();
+    } else if (/data:image\/([a-zA-Z]*);base64,([^"]*)/.test(image)) {
+      data = Buffer.from(image.split(',')[1], 'base64');
+    } else {
+      data = await readFile(image);
+    }
+  } else if (Buffer.isBuffer(image)) {
+    data = image;
+  }
+
+  return new Uint8Array(data);
+};
@@ -0,0 +1,5 @@
+'use strict';
+
+module.exports = (worker, handler) => {
+  worker.on('message', handler);
+};
@@ -0,0 +1,12 @@
+'use strict';
+
+/**
+ * send
+ *
+ * @name send
+ * @function send packet to worker and create a job
+ * @access public
+ */
+module.exports = async (worker, packet) => {
+  worker.postMessage(packet);
+};
@@ -0,0 +1,12 @@
+'use strict';
+
+const { Worker } = require('worker_threads');
+
+/**
+ * spawnWorker
+ *
+ * @name spawnWorker
+ * @function fork a new process in node
+ * @access public
+ */
+module.exports = ({ workerPath }) => new Worker(workerPath);
@@ -0,0 +1,12 @@
+'use strict';
+
+/**
+ * terminateWorker
+ *
+ * @name terminateWorker
+ * @function kill worker
+ * @access public
+ */
+module.exports = (worker) => {
+  worker.terminate();
+};