Renaming module part 2.

[ocefpaf]

Should Fix #12

[rhattersley]

Looks like this needs rebasing.

[ocefpaf]

@rhattersley not anymore 😉

[rhattersley]

I did a quick grep for "iris" and "unit>" and it threw up a couple of things:

1. IRIS_EPOCH
2. Lots of import unit and subsequent unit.<method> calls in the docstrings.

[ocefpaf]

Is IRIS_EPOCH -> EPOCH OK?

[rhattersley]

Is IRIS_EPOCH -> EPOCH OK?

Yeah, seems OK.

[ocefpaf]

@rhattersley should I those to __all__ in cf_units.py or is this OK?

[ocefpaf]

@rhattersley this time I tried to be very careful with the renaming and I believe we should be OK now. I also ran nosetests --verbose --with-doctest cf_units to check the doctests and all but one passed:

File "cf_units.py", line 1961, in cf_units.cf_units.Unit.date2num
Failed example:
    u.date2num(datetime.datetime(1970, 1, 1, 5))
Expected:
    5.00000000372529
Got:
    5.0000000037252903

But I don't think that this precision is an issue.

[ocefpaf]

@rhattersley I want to start adding the tests (#2), but I want to avoid adding them and then re-naming everything. So I am waiting for this one to start the test migration.

Do you think this PR is OK to go?

[rhattersley]

Do you think this PR is OK to go?

It's close enough. 😁