Issue 856: refactoring of how filenames are chosen for Detailed Summary

[artoonie]

Streamlines how filenames are constructed, and rename:

• summary.json -> detailed_report.json
• extended_summary.csv -> detailed_report.csv
• summary.csv -> summary_report.json

[yezr]

Something that I am noticing as I get acquainted with this update is the need for some vocabulary for function names that won't change if the file names of the RCTab output changes. What I want to differentiate is

• Round by Round results, what previously were the "summary files". These can include entire data for an entire contest or for just a slice, and can further be either detailed or summary format. I suggest we refer to these as "Results" since "Round by Round Results" is already in use generally.
• ALL files that are generated from a run. Above plus rctab_cvr and the CDF formatted CVR. I suggest we refer to these as "OutputFiles" or simply "Output".

If we move forward with those names there are some refactors now that I think can prevent refactors in the future if explicit names change (like the new ResultFile being OutputFile)

Other stuff

• A couple comments related to ResultFile knowing it's own path. I feel I can't explain concretely what needs to happen, but something seems off there especially related to sequential
• Awesome that all of the tests just need a rename! Proves that all these code updates are just refactors!

[artoonie]

Agree with the rename! Thanks for the commit.

OutputFileIdentifiers is better than what it would have been (OutputTypeAndSlice), and I can't think of a better name, but did want to flag that it is only some of the identifiers -- namely, it doesn't include the Sequence ID. If anyone has a suggestion for a better name, I'd be open to hearing it! Otherwise, 👍

[yezr]

notes from in person code review today

[alyssahursh]

Hey, folks! My apologies for taking so long on this review; I found the size of the diff a little overwhelming. Overall, I think I agree with Mat that the existing tests give me a certain amount of confidence that these changes have the intended result, but I struggled hard with the newOutputFileIdentifiers class.

It's possible that by reviewing commits sequentially I got caught up and confused in the same naming weeds you did, but I would love to have more time to consider the full shape of the OutputWriter class alongside the new OutputFileIdentifiers class to see if there are other threads we might pull on for refactoring.

I left a few comments. The one about thread safety feels the most important to me; the others are mostly just suggestions to make the code slightly more functional and less procedural. Do with them what you will! Let me know if you have any questions.