w

Dockerfile

@@ -0,0 +1,11 @@
+FROM mcr.microsoft.com/playwright:v1.40.0-jammy
+
+#WORKDIR /code
+
+#COPY package.json ./package.json
+
+RUN npm i [email protected]
+
+COPY . .
+
+CMD [ "node", "index.mjs" ]

index.mjs

@@ -0,0 +1,81 @@
+import * as http from 'http';
+import * as url from 'url';
+import * as path from 'path';
+import { chromium } from 'playwright';
+import { locks } from 'web-locks';
+
+// npm i playwright
+// npm i web-locks
+// nohup cloudflared tunnel --url http://localhost:8080 --no-autoupdate &  (or setsid)
+// setsid node playwright.mjs  (nohup doesnt work)
+
+globalThis.state = Object.assign(globalThis.state||{}, {
+    browser: null,
+    context: null,
+});
+
+
+async function text_from_url(url, cookie) {
+    let { browser, context } = globalThis.state;
+    if (!browser) {
+        await locks.request('playwright_browser', async lock => {
+            browser = globalThis.state.browser = await chromium.launch();
+            context = globalThis.state.context = await browser.newContext({        
+                    javaScriptEnabled: false
+            }/*devices['iPhone 11']*/);
+        });
+    }
+
+    const page = await context.newPage();
+
+    if (cookie) {
+        if (cookie.endsWith(';')) cookie = cookie.slice(0, -1);
+        let cookies = cookie.split(';');
+        cookies = cookies.map(it=>{
+            let [name, value] = it.split('=');
+            return {name: name.trim(), value: value.trim(), url};
+        });
+        context.addCookies(cookies);
+    }
+
+    await context.route("**/*.{png,jpg,jpeg,css,js}", route => route.abort());
+    await page.goto(url);
+
+    let text;
+    let i = 0;
+    while (true) {
+        let new_text = await page.evaluate(() => document.body.innerText);
+        if (i > 5 || new_text?.length > 200) {
+            text = new_text;
+            break;
+        }
+        i++;
+        await new Promise(resolve=>setTimeout(resolve, 1000));
+    }
+
+    await page.close();
+    //await context.close();
+    // await browser.close();
+
+    return text;
+}
+
+
+const server = http.createServer(async function(req, res) {
+    const {query, pathname} = url.parse(req.url, true);
+
+    res.setHeader('Access-Control-Allow-Origin', '*')
+
+    let _url = query.url;
+    let cookie = query.cookie;
+
+    try {
+    	let text = await text_from_url(_url, cookie);
+
+    	res.end(text);
+    } catch (e) {
+    	res.end(e.toString());
+    }
+});
+
+server.listen(7860);