`main.py` is currently untested and not easily testable

Although we run using Coverage and configure it to fail at less than 100% we don't actually tell it what directories to cover, so it ends up not enforcing anything.

The fix is to change this line:
https://github.com/opensafely-core/tpp-database-utils/blob/067273b699a5d8d8946ea337dd8466af7205390f/justfile#L103

To this:

 uv run coverage run --source tpp_database_utils,tests --module pytest {{ args }}

This then fails because main.py is untested:

uv run coverage report || uv run coverage html
Name                         Stmts   Miss Branch BrPart  Cover   Missing
------------------------------------------------------------------------
tpp_database_utils/main.py      45     45     16      0     0%   1-74
------------------------------------------------------------------------
TOTAL                          253     45     40      0    79%

However, it's not (straightforwardly) possible to write tests for it because it assumes it's being invoked directly as a script and therefore that its sibling modules are all on sys.path.

If you try creating tests/test_main.py:

from tpp_database_utils import main

def test_main():
   assert main

It will fail with an import error:

$ pytest tests/test_main.py 
========================== test session starts ==========================
platform linux -- Python 3.11.14, pytest-9.0.2, pluggy-1.6.0
rootdir: /home/dave/projects/bennett/tpp-database-utils
configfile: pyproject.toml
collected 0 items / 1 error                                             

================================ ERRORS =================================
__________________ ERROR collecting tests/test_main.py __________________
ImportError while importing test module '/home/dave/projects/bennett/tpp-database-utils/tests/test_main.py'.
Hint: make sure your test modules/packages have valid Python names.
Traceback:
../../../.local/share/uv/python/cpython-3.11.14-linux-x86_64-gnu/lib/python3.11/importlib/__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
tests/test_main.py:1: in <module>
    from tpp_database_utils import main
tpp_database_utils/main.py:4: in <module>
    import maintenance_mode
E   ModuleNotFoundError: No module named 'maintenance_mode'
======================== short test summary info ========================
ERROR tests/test_main.py
!!!!!!!!!!!!!!!! Interrupted: 1 error during collection !!!!!!!!!!!!!!!!!
=========================== 1 error in 0.07s ============================

The simplest fix here is probably to move to main.py to __main__.py, rewrite the imports to use the package name and then change the Docker entrypoint to use python -m tpp_database_utils:
https://github.com/opensafely-core/tpp-database-utils/blob/067273b699a5d8d8946ea337dd8466af7205390f/docker/Dockerfile#L155

Start with tpp_database_utils/main.py, the coverage command at justfile line 103, and the Docker entrypoint at docker/Dockerfile line 155. Try the proposed main.py and package-import approach, then add tests/test_main.py and run the coverage command. Done means main.py is testable and coverage reports the intended package and tests without the import error.