GradeScraper
Find out if your exam at Fernuni Hagen has finally been graded.
This is my first Elixir program, so don't expect anything special. ;)
Prerequisites
Need to have selenium-webserver and elixir installed. On a mac:
brew install selenium-webserver-standalone
brew install elixir
On Linux, use your favourite package manager. On Windows, you're on your own.
You'll also need to run
mix local.hex
mix local.rebar
once initially.
You'll also need an older version of Firefox (<= 46).
Unfortunately, newer versions of Firefox break selenium. Installing FF46 and disabling upgrades worked for me. The sad state of browser automation seems to be that nothing works out of the box anymore without tedious configuration. Chromedriver didn't work for me. Phantomjs (1 or 2) didn't work for me (I suspect the page it's trying to open is too broken for phantomjs to work). If anybody knows anything that actually works, please let me know.
Select exam
Open up lib/grade_scraper.ex and exchange the @exam_search_string constant (currently, it's checking for the probability module). Basically, the scraper will just check whether that string appears somewhere on the grade listing page.
Run
Issuing ./start should be enough. This starts a selenium process in the background and runs the scraper against it. It will spin up a browser (haven't been able to make it headless yet), so be prepared for that.
You will also be asked for your credentials - these aren't stored anywhere though.
Store credentials
If you're tired of inputting your credentials all the time, you can (at your own risk!) store them in the FERNUNI_POS_USERNAME and FERNUNI_POS_PASSWORD environment variables (e.g. in your shell startup script).
TODO
- actually output the grade instead of just alerting you that there is a grade
- make exam title configurable more easily (parameter maybe?)
- run this in headless mode somehow?