Add interface for setting project, crystal, and dataset names

[JBGreisman]

This PR addresses #103 by adding an interface for setting the project, crystal, and dataset names in DataSet.to_gemmi() and DataSet.write_mtz(). These names are then set in the MTZ file or gemmi.Mtz object.

This can be useful for record keeping, allowing custom titles in MTZ files that can be read by other programs.

These names default to "reciprocalspaceship" -- if these new arguments are left unset, a user will have the same behavior as before.

[codecov-commenter]

Codecov Report

Merging #157 (97124c4) into main (79a28fb) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #157   +/-   ##
=======================================
  Coverage   98.45%   98.46%           
=======================================
  Files          43       43           
  Lines        1688     1695    +7     
=======================================
+ Hits         1662     1669    +7     
  Misses         26       26           

Flag	Coverage Δ
unittests	98.46% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
reciprocalspaceship/dataset.py	98.03% <100.00%> (ø)
reciprocalspaceship/io/mtz.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 79a28fb...97124c4. Read the comment docs.

[kmdalton]

I think this code would be cleaner and your tests more straightforward if the signature of to_gemmi were changed to

def to_gemmi(
        self, 
        skip_problem_mtztypes=False, 
        project_name="reciprocalspaceship",
        crystal_name="reciprocalspaceship",
        dataset_name="reciprocalspaceship",
    ):

I think strings are immutable, and this is okay to do. Let me know if I am being obtuse, and I will have another look.

[JBGreisman]

I think that's a reasonable request. It certainly makes it more transparent what the default value will ultimately be.

[kmdalton]

I think that passing DataSet.to_gemmi anything other than a string for {project,crystal,dataset}_name should raise a ValueError. Passing None should probably not be silently corrected back to the default values in the signature. I think the tests should look for a ValueError when None is passed. I might be being overly picky.

[JBGreisman]

I'm fine with that -- I agree that silently coercing a value of None to "reciprocalspaceship" is a bit surprising from an interface perspective. The docstrings certainly specify that the values take str, so I think it's reasonable to raise a ValueError for anything else.

[kmdalton]

I think that the methods DataSet.to_gemmi and DataSet.write_mtz should be the only places with default arguments for the various name strings. The functions in io are not really meant to be called directly by users. I think it is good for them to have the names as required arguments. This way, if we ever decide to change the default names in the future, they only need to be updated at the top level.

[kmdalton]

i don't see any glaring issues here, and i like the solution.