Switching to processes and whisper.cpp to avoid transformers.js memory leaks

src/vision.ts

@@ -2,43 +2,22 @@ import {createWorker} from 'tesseract.js';
 import {AbortablePromise, Ai} from './ai.ts';
 
 export class Vision {
-	private worker: any = null;
-	private queue: Array<{ path: string, resolve: any, reject: any }> = [];
-	private busy = false;
 
 	constructor(private ai: Ai) {}
 
-	private async processQueue() {
-		if(this.busy || !this.queue.length) return;
-		this.busy = true;
-		const job = this.queue.shift()!;
-		if(!this.worker) this.worker = await createWorker(this.ai.options.ocr || 'eng', 2, {cachePath: this.ai.options.path});
-		try {
-			const {data} = await this.worker.recognize(job.path);
-			job.resolve(data.text.trim() || null);
-		} catch(err) {
-			job.reject(err);
-		}
-		this.busy = false;
-		this.processQueue();
-	}
-
 	/**
 	 * Convert image to text using Optical Character Recognition
 	 * @param {string} path Path to image
 	 * @returns {AbortablePromise<string | null>} Promise of extracted text with abort method
 	 */
 	ocr(path: string): AbortablePromise<string | null> {
-		let aborted = false;
-		const abort = () => { aborted = true; };
-		const p = new Promise<string | null>((resolve, reject) => {
-			this.queue.push({
-				path,
-				resolve: (text: string | null) => !aborted && resolve(text),
-				reject: (err: Error) => !aborted && reject(err)
-			});
-			this.processQueue();
+		let worker: any;
+		const p = new Promise<string | null>(async res => {
+			worker = await createWorker(this.ai.options.ocr || 'eng', 2, {cachePath: this.ai.options.path});
+			const {data} = await worker.recognize(path);
+			await worker.terminate();
+			res(data.text.trim() || null);
 		});
-		return Object.assign(p, {abort});
+		return Object.assign(p, {abort: () => worker?.terminate()});
 	}
 }