Add names for bundles and test they are consistent

[stefpiatek]

Moved raw data to new structure with a separate file for the names, and added in some tests to ensure that we can detect inconsitencies. Bit of refactoring of raw data functions, expect that once we have a release serving off a package .rda file that we'll change the raw data to live outside of the package but seems fine for now as we still need to serve from the package.

• One question though, what domain does indices of deprivation go into?

[milanmlft]

expect that once we have a release serving off a package .rda file that we'll change the raw data to live outside of the package but seems fine for now as we still need to serve from the package.

Agreed, having it run off .rda would mean all the data is immediately available and we don't need to re-read the csv files every time we wan to access a bundle. On the other hand, it might make loading the package slower as all the data would immediately be loaded in as well. If the data gets big enough that might become a performance issue. Maybe worth doing some benchmarking on that once we get to that point.

[milanmlft]

Not sure why the ::: everywhere for get_raw_dir()?

[milanmlft]

Shouldn't have to call library() or omopbundles::* in tests, as the package code should work independently of that

[milanmlft]

LGTM! 🚀 Just some suggestions regarding library() and ::/::: calls
Just wondering how the bundle_names.csv files get generated? Might be worth adding a script to reproduce these files to the package. A common place for this could be data-raw/bundle_names.R

[stefpiatek]

Just wondering how the bundle_names.csv files get generated? Might be worth adding a script to reproduce these files to the package. A common place for this could be data-raw/bundle_names.R

I think the names should be manually curated, or exported from the source system