Create a `dora.py` script to calculate Release Frequency and Lead Time for Changes

[grahamalama]

This script will use only git commit data to infer these metrics:

• Deployment frequency will be inferred by <average # of tagged commits> / <some period of time>
• Lead time for changes will be inferred by the average of the difference of commit timestamp and the timestamp of the release it became a part of. That is, if we had these commits:

  commit days_before_release release
  a      -                   v1.9.1
  b      1 day
  c      3 days
  d      -                   v1.9.0
  e      2 days
  f      5 days

  We'd call Lead time for changes "about 2.75 days".

This script will assume that a tagged commit in git represents a successful release to production.

[grahamalama]

Closing this is issue since dora metrics ultimately won't live as a script. This logic of deriving lead time to change may live on elsewhere, though.