use MD5 password hasher under test

[chris48s]

This restores the optimisation I removed in #109 to get the tests working.
Now that we're using Factories rather than fixtures, we can use what we want under test.

A good production password hashing algorithm is slow by design and hard to optimise. This makes it difficult to brute-force. The reason cookiecutter suggests this is because MD5 is not very secure but it is fast to calculate. We don't care about security of data we generate while running tests, but we do want it to be fast to build, so using a known vulnerable hash algorithm makes sense here. Its a bit of a micro-optimisation, but its definitely not going to make the tests run any slower, so we might as well re-enable it now.

[codecov]

Codecov Report

Merging #130 into master will not change coverage by %.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master     #130   +/-   ##
=======================================
  Coverage   76.33%   76.33%           
=======================================
  Files          28       28           
  Lines         393      393           
=======================================
  Hits          300      300           
  Misses         93       93           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 412d7dd...008a04e. Read the comment docs.

[lpatmo]

Sounds good! Thanks for the context, and feel free to merge. :)