Run `npm install` to download the correct Chromium revision (1045629). Linux Relevant log output Error: Could not find expected browser (chrome) locally. Instead of specifying a download location using PUPPETEER_CACHE_DIR, is there any way the old behaviour can optionally be turned on, and let puppeteer figure out where it’s installed? This would avoid having to know the subtlety of the underlying build/file system.Ĩ.5.0 What operating system are you seeing the problem on? example: retrieve handler and call for funcaptcha token. Supported captcha tasks listed on capsolver-npm at capsolver-npmSupported API methods. Use the handler with: await page.capsolver() Perform any task that capsolver-npm brings. Handler its attached to any puppeteer page. I’ve tried playing around with PUPPETEER_CACHE_DIR, but again, we have very little guarantees over the build flow and it potentially changing. Based on capsolver-npm (nodejs api wrapper for api). While the Cloud Function environment is not a direct concern of the puppeteer project, the caching feature introduced in 19.0.0 makes it hard to reliably install Chromium in an environment where we don’t control the build/file system. Although the build (and probably the download of Chromium) succeeds, Chromium is not available at runtime.
My guess is that the new cache location is not retained in the final Docker image for the Cloud Function. Invoke the Cloud Function, the error below occurs.Deploy the Cloud Function, the build succeeds.
Depend on puppeteer 19.0.0 in a Cloud Function.
0 kommentar(er)
