Puppeteer improvements (#186)

* improving puppeteer handling

* upgrade dependencies

* reduce logging

* upgrade nanoid

lib/provider/neubauKompass.js

@@ -23,7 +23,7 @@ const config = {
   url: null,
   crawlContainer: '.col-12.mb-4',
   sortByDateParam: 'Sortierung=Id&Richtung=DESC',
-  waitForSelector: '.nbk-section',
+  waitForSelector: 'div[data-live-name-value="SearchList"]',
   crawlFields: {
     id: 'a@href',
     title: 'a@title | removeNewline | trim',

lib/services/extractor/parser/parser.js

@@ -9,12 +9,12 @@ export function loadParser(text) {
 
 export function parse(crawlContainer, crawlFields, text, url) {
   if (!text) {
-    logger.warn('No content found for ', url);
+    logger.debug('No content found for ', url);
     return null;
   }
 
   if (!crawlContainer || !crawlFields) {
-    logger.warn('Cannot parse, selector was empty for url ', url);
+    logger.debug('Cannot parse, selector was empty for url ', url);
     return null;
   }
 

lib/services/extractor/puppeteerExtractor.js

@@ -2,30 +2,56 @@ import puppeteer from 'puppeteer-extra';
 import StealthPlugin from 'puppeteer-extra-plugin-stealth';
 import { debug, DEFAULT_HEADER, botDetected } from './utils.js';
 import logger from '../logger.js';
+import fs from 'fs';
+import os from 'os';
+import path from 'path';
 
 puppeteer.use(StealthPlugin());
 
 export default async function execute(url, waitForSelector, options) {
   let browser;
+  let page;
+  let result = null;
+  let userDataDir;
+  let removeUserDataDir = false;
   try {
     debug(`Sending request to ${url} using Puppeteer.`);
 
+    // Prepare a dedicated temporary userDataDir to avoid leaking /tmp/.org.chromium.* dirs
+    if (options && options.userDataDir) {
+      userDataDir = options.userDataDir;
+      removeUserDataDir = !!options.cleanupUserDataDir;
+    } else {
+      const prefix = path.join(os.tmpdir(), 'puppeteer-fredy-');
+      userDataDir = fs.mkdtempSync(prefix);
+      removeUserDataDir = true;
+    }
+
     browser = await puppeteer.launch({
       headless: options.puppeteerHeadless ?? true,
-      args: ['--no-sandbox', '--disable-gpu', '--disable-setuid-sandbox'],
+      args: [
+        '--no-sandbox',
+        '--disable-gpu',
+        '--disable-setuid-sandbox',
+        '--disable-dev-shm-usage',
+        '--disable-crash-reporter',
+      ],
       timeout: options.puppeteerTimeout || 30_000,
+      userDataDir,
     });
-    let page = await browser.newPage();
+    page = await browser.newPage();
     await page.setExtraHTTPHeaders(DEFAULT_HEADER);
     const response = await page.goto(url, {
       waitUntil: 'domcontentloaded',
     });
     let pageSource;
-    //if we're extracting data from a spa, we must wait for the selector
+    // if we're extracting data from a SPA, we must wait for the selector
     if (waitForSelector != null) {
-      await page.waitForSelector(waitForSelector);
+      const selectorTimeout = options?.puppeteerSelectorTimeout ?? options?.puppeteerTimeout ?? 30_000;
+      await page.waitForSelector(waitForSelector, { timeout: selectorTimeout });
       pageSource = await page.evaluate((selector) => {
-        return document.querySelector(selector).innerHTML;
+        const el = document.querySelector(selector);
+        return el ? el.innerHTML : '';
       }, waitForSelector);
     } else {
       pageSource = await page.content();
@@ -35,16 +61,35 @@ export default async function execute(url, waitForSelector, options) {
 
     if (botDetected(pageSource, statusCode)) {
       logger.warn('We have been detected as a bot :-/ Tried url: => ', url);
-      return null;
+      result = null;
+    } else {
+      result = pageSource || (await page.content());
     }
-
-    return await page.content();
   } catch (error) {
     logger.error('Error executing with puppeteer executor', error);
-    return null;
+    result = null;
   } finally {
-    if (browser != null) {
-      await browser.close();
+    try {
+      if (page) {
+        await page.close();
+      }
+    } catch {
+      // ignore
+    }
+    try {
+      if (browser != null) {
+        await browser.close();
+      }
+    } catch {
+      // ignore
+    }
+    try {
+      if (removeUserDataDir && userDataDir) {
+        await fs.promises.rm(userDataDir, { recursive: true, force: true });
+      }
+    } catch {
+      // ignore
     }
   }
+  return result;
 }

package.json

@@ -1,6 +1,6 @@
 {
   "name": "fredy",
-  "version": "12.1.5",
+  "version": "12.1.6",
   "description": "[F]ind [R]eal [E]states [d]amn eas[y].",
   "scripts": {
     "prepare": "husky",
@@ -70,7 +70,7 @@
     "handlebars": "4.7.8",
     "lodash": "4.17.21",
     "markdown": "^0.5.0",
-    "nanoid": "5.1.5",
+    "nanoid": "5.1.6",
     "node-cron": "^4.2.1",
     "node-fetch": "3.3.2",
     "node-mailjet": "6.0.9",

yarn.lock

@@ -5406,10 +5406,10 @@ nano-spawn@1.0.3:
   resolved "https://registry.yarnpkg.com/nano-spawn/-/nano-spawn-1.0.3.tgz#ef8d89a275eebc8657e67b95fc312a6527a05b8d"
   integrity sha512-jtpsQDetTnvS2Ts1fiRdci5rx0VYws5jGyC+4IYOTnIQ/wwdf6JdomlHBwqC3bJYOvaKu0C2GSZ1A60anrYpaA==
 
-nanoid@5.1.5:
-  version "5.1.5"
-  resolved "https://registry.yarnpkg.com/nanoid/-/nanoid-5.1.5.tgz#f7597f9d9054eb4da9548cdd53ca70f1790e87de"
-  integrity sha512-Ir/+ZpE9fDsNH0hQ3C68uyThDXzYcim2EqcZ8zn8Chtt1iylPT9xXJB0kPCnqzgcEGikO9RxSrh63MsmVCU7Fw==
+nanoid@5.1.6:
+  version "5.1.6"
+  resolved "https://registry.yarnpkg.com/nanoid/-/nanoid-5.1.6.tgz#30363f664797e7d40429f6c16946d6bd7a3f26c9"
+  integrity sha512-c7+7RQ+dMB5dPwwCp4ee1/iV/q2P6aK1mTZcfr1BTuVlyW9hJYiMPybJCcnBlQtuSmTIWNeazm/zqNoZSSElBg==
 
 nanoid@^3.3.11:
   version "3.3.11"