There was a great post a few months ago on a set of python tools for windows application GUI testing: https://community.alteryx.com/t5/Engine-Works-Blog/Total-Automation-Badassery/ba-p/153885
However, to my knowledge the pywinauto package used for these tools does not play nicely with web automation, so I wanted to start designing tools that did. Research pointed me to selenium as the best option for this kind of testing.
I started writing some simple python to get a feel for this, navigating to a web page, clicking on some links, scraping some information, etc.
However, I'm encountering issues maintaining a session between different scripts - I can maintain a process id and a host process, but seem to encounter errors trying to use these after an initial script has completed. Research suggests that most people don't use selenium this way! :)
I'm open to any advice, but I'm wondering if there is another way to architect this toolset to expose this automation functionality to non-coders. Maybe writing out python code and executing it, although that seems like it could have plenty of risks as well(?).
