Files renamed to .py, tests added

[federicaspoto]

@stevechesley I just learned this new feature about "Draft pull request". It will allow me to write what I'm doing while I'm doing it and you should be able to see everything.

• Step 1: we want to rename the scripts in Python/bin and call them adding the *.py at the end. This is how we call them in our system.
• Step 2: we want to make the main script a routine, so that you can easily call them inside the code. But you can still call them from command line in exactly the same way as it was before.
• Step 3: I'm writing new tests for every script I'm changing. For the moment, they are in a separate directory called "new_tests".
• I will need to change a few threshold in the validation, because a few observations don't pass the validation right now. I'll add more details when I am at that point.

[federicaspoto]

@stevechesley @matthewjohnpayne I believe that this is complete.

• I have changed almost all the names in Python/bin (e.g. mpc80coltoxml is now mpc80coltoxml.py).
• I have removed the /usr/bin/python from the beginning of the files
• I have changed the structure so that you can import the single module and use the functions, but the call is exactly as it was before (side note: I could not change xmltopsv.py - too complex)
• I have created new tests and added new files (the tests are in new_tests and run with pytest)
• I have merged the changes from the master in my branch @stevechesley we can wait from when you are back from vacations to check the code and do the pull review.

[matthewjohnpayne]

Everything looks good, with the following exception:

I recommend changing test_mpc80coltoxml.py as described below, so that when pytest is executed, everything behaves in a standard manner:

• Import the pytest package inside test_mpc80coltoxml.py
  • import pytest
• Add decorator before expected fail ...
  • @pytest.mark.xfail
  • def test_mpc80coltoxml_submitted_split():

Then when you run "pytest test_mpc80coltoxml.py" you should see ...

• test_mpc80coltoxml.py ..x [100%]

  2 passed, 1 xfailed in 0.25s

[federicaspoto]

@matthewjohnpayne thank you! I have modified the test.

[matthewjohnpayne]

I'm happy to merge But I'll leave that to you/Steve once Steve has had chance to review M

[stevechesley]

@federicaspoto I think this could be ready to merge, but I propose that we just keep working on this branch before merging. I see three things that could/should be done before the merge:

1. Incorporate argparse into the scripts so that arguments handling would be more flexible and standardized, with automatic documentation.
2. Adapt a traditional functional approach, rather than just passing the command line array as a sole argument. So you would call function(arg1, arg2) rather than function(["", "arg1", "arg2"]).
3. implement tests for invocation as a function, in addition to the current tests for command line use.

I think items 1 & 2 are intertwined and would best be done together. For item 2, nobody outside of the MPC should have implemented these functions yet since this is still a development branch. So it seems a good moment to standardize/modernize the functional form before merging.

We can take care of all of these in fairly short order if you concur.

[stevenstetzler]

@federicaspoto I've implemented suggestion 1 + 2 from Steve's list and pushed the code in PR #23 so you can review it. If you give the go-ahead, I can merge that into this branch. And I can implement 3 from Steve's list if you approve. Thanks!

[federicaspoto]

@stevechesley @stevenstetzler I added the new tests, as we talked about.

[stevechesley]

@stevenstetzler Is the argparse ready to merge? Looks like it needs to first have the master merged into the branch. But not sure if it's ready yet. Feel free to go ahead with a pull request when it's ready.