Remove use of python ua-parser from the backend.

[rtibbles]

Summary

• The Python ua-parser module is exhaustive, and as a result requires at least a thousand regexes.
• Currently it initializes them all at module import, which causes anywhere up to a 5 second delay in initializing Kolibri on a highly specced dev machine
• We could defer import until after initialization, but that would still mean a very lengthy initialization process at some point
• To avoid this, instead just do all user agent parsing in the frontend using the ua-parser-js library (which is far more performant than other user agent libraries), and which we were already partially using in vendored form
• Removes vendoring, and imports this library
• Uses this for all browser and os info inspection from the user agent
• Explicitly sends this info with session requests to be recorded in the user session logs
• Changes all session requests to use POST or PUT
• Sets 'active' on login POSTs so that user session logs are created
• Removes unneeded SessionResource

References

Removes major initialization bottleneck, reducing startup time from ~5s to < 0.1s.

Follow up from #7658

Reviewer guidance

Does login and subsequent pings to the session endpoint work?
Does appropriate data get recorded in the UserSessionLog?
Is startup performance on Android improved?

---

Testing checklist

• Contributor has fully tested the PR manually
• If there are any front-end changes, before/after screenshots are included
• Critical user journeys are covered by Gherkin stories
• Critical and brittle code paths are covered by unit tests

PR process

• PR has the correct target branch and milestone
• PR has 'needs review' or 'work-in-progress' label
• If PR is ready for review, a reviewer has been added. (Don't use 'Assignees')
• If this is an important user-facing change, PR or related issue has a 'changelog' label
• If this includes an internal dependency change, a link to the diff is provided

Reviewer checklist

• Automated test coverage is satisfactory
• PR is fully functional
• PR has been tested for accessibility regressions
• External dependency files were updated if necessary (yarn and pip)
• Documentation is updated
• Contributor is in AUTHORS.md

[indirectlylit]

~5s to < 0.1s.

Wow, that's huge!

Will review and test later this week

Follow up from #7658

Since that was merged into develop after 0.14, presumably the slow startup time didn't hit most users (except I suppose S2S?)

[rtibbles]

Since that was merged into develop after 0.14, presumably the slow startup time didn't hit most users (except I suppose S2S?)

Yes, and that's why I targeted it to 0.15, so we could prevent this issue even there.

[jredrejo]

Code looks good, nothing to comment.
All the functionalities work well without regressions.
Tested it in linux (dev), android and windows 10. I haven't measured the improvement when starting but data logs are now better, at least detecting the Linux version.

Before:

After: