add service number

[caleb15]

I think CIRCLE_WORKFLOW_ID maps to the number variable (aka service_number) but I'm not 100% sure.
Note that node coveralls uses CIRCLE_WORKFLOW_ID as the service number. See https://github.com/nickmerwin/node-coveralls/blob/705c3b5963e3cc76f8e70381dbae20979008c9b8/lib/getOptions.js#L66

[caleb15]

Circleci pre-commit failed because of pylint but that is a unrelated to my change.

pylint...................................................................Failed
- hook id: pylint
- exit code: 16

[caleb15]

I got the parallel option working with using workflow id as the service num so I'm a bit more confident this is the right approach now:

CI Config:

# js tests:
      - run: yarn test --collectCoverage --coverageDirectory=ff/static/coverage --ci --maxWorkers=3 --reporters=default --reporters=jest-junit

      - run: export COVERALLS_PARALLEL=true;yarn coveralls -v < ff/static/coverage/lcov.info

# python tests (runs after js tests)
      - run:
          name: run tests
          command: venv/bin/coverage run --parallel-mode manage.py test --noinput --parallel=4 --verbosity=2
          no_output_timeout: 20m
      - run: venv/bin/coverage combine
      - run: venv/bin/coverage report
      - run: export COVERALLS_SERVICE_JOB_NUMBER=$CIRCLE_WORKFLOW_ID;export COVERALLS_PARALLEL=true;venv/bin/coveralls
      # coveralls --finish is broken so we need to specify job num
      - run: export COVERALLS_SERVICE_JOB_NUMBER=$CIRCLE_WORKFLOW_ID;venv/bin/coveralls --finish

[TheKevJames]

@caleb15 thanks for the contribution here -- just catching up on #234 and #235 now. Why did this PR get closed?

[caleb15]

@TheKevJames sorry about that, I accidentally closed it. In my PR https://github.com/15five/fifteen5/pull/30230 I wrote the following:

I experienced a bug in the coveralls python library so I raised a PR with a fix #233 and did export COVERALLS_SERVICE_JOB_NUMBER=$CIRCLE_WORKFLOW_ID; as a temporary workaround.

Github thought that me saying "fix #233" meant that the PR had been fixed and could be closed 🙄 Too smart for its own good!

[TheKevJames]

Oh, wow, very weird that Github will close things cross-repos like that -- especially since you made the comment in a private repo and it closed a PR on an OSS repo owned by a completely different Github org. "Too smart for its own good!" no kidding!

Anyway, this changeset looks good to me -- I'm just going to wait for an answer on #235 to make sure it works for that case as well, then will merge and released this. Thanks for the contribution!