Deployed on GitHub pages here.

:warn: IMPORTANT NOTE: The file transcript.html is manually created:

• login into AWS
• navigate to "Transcript"
• select "Completed"
• use Chrome tools (inspect)
• select top html tag
• right-click and 'copy-outer-html'
• paste into transcript.html
• this became neccessary when AWS started blocking the automated login from Puppeteer.

pup-learning-aws

A tool to gather Learning classes completed along with details.

When run as a Node.js tool, it will parse all the Completed Courses of the configured user, It will save that data as ./artifact/sample/json, and it will generate

• public/index.html
• artifacts/learning-summary-aws.mdx

The associated GH Action will deploy the index.html file.

The learning-summary.mdx can be manually copied to the my-blog2 directory and committed to that repo as post in the blog.

Mac Install

Install npm from the website or homebrew. You will also need yarn.

NOTE: Normally, you can use either npm or yarn to install modules. The debug of Chrome with DNS Manager (mising a valid certificate), discovered that we need older versions of some packages. This can be done with either npm or yarn. We currently only have this setup with the yarn.lock file.

Linux Install

For a Unbuntu system, use:

sudo apt-get install -y npm yarn

For other distributions, please search on Google.

Install then run

First, there is a one-time setup of enviroment variables like:

export PUP_USERNAME=<AWS Learning User Name>
export PUP_PASSWORD=<AWS Learning User Password>
yarn install

From a command line:

yarn start

You can also runs this 'headless' or without a browser window. Look for 'headless' in site.js. To verify, you should see the screenshots generated.

Internal Settings

There are number of things can be changed quickly in the source code. Eventually they will be added as program options.

Look for "INTERNAL OPTION".

    browserType:     "firefox", // "chrome, firefox"
    headless:        false,     // run without windows
    forceFullGather:  true,     // skip test for number of course
    scrollToBottom:   true,     // scroll page to bottom (WARNING: non-visible thumbnails are not loaded until page is scrolled)
    gatherDetails:    true,     // parse the details
    useSampleData:   false,     // skip browser and use sample data file
    saveSampleData:   true,     // save to sample data file
    screenshot:      false,     // take snapshots
    screenshotDir:    "/tmp/pup_learning_aws_screenshots"

Chromium vs Firefox for Puppeteer

Puppeteer can run automated tests with both Chrome (technically the Chromium build) and Firefox.

Headless

Headless is the ability to run a web page without showing on the screen. This mode can be used for automated testing. This works with both Chrome and Firefox.

Screenshots

This feature allows capturing an image of the web page while the test is running.

Local Test of index.html

• cd public
• python -m SimpleHTTPServer
• open http://localhost:8000/

Know Issues

• options must be set in code

TODO List:

• set up with GH Actions to run "update" with docker/ubuntu18/node12...
  • get
  • yarn install
  • yarn start
  • verify index
  • commit /public /artifacts (.md)
• go back to HACK (limit to 10 detail pages per run of tool) DEBUG firefox

Future changes:

TBD