Splink datasets - simplify + restructure

[ADBond]

Type of PR

• BUG
• FEAT
• MAINT
• DOC

Is your Pull Request linked to an existing Issue or Pull Request?

Closes #2369.

Much of this is long-overdue tidying and simplifying of the over-complex code that was in place. There are a few things repeated in a couple of places now (dataset names + descriptions), but I think this is okay, as these are unlikely to change / be added to particularly frequently.

Give a brief description for the solution you have provided

Functionality is largely the same. Details:

• splink_datasets should work now even if the user does not have write permissions for the caching folder. If not, they will download directly and use without on-disk caching
• datasets are cached on the object when accessed, so repeated calls will not cause multiple reads/downloads
• code is split into modules to make it more maintainable
• dataset attributes are defined statically, so can be accessed with autocomplete
• as well as descriptions, dataset docstrings include column information

PR Checklist

• Added documentation for changes
• Added feature to example notebooks or tutorial (if appropriate)
• Added tests (if appropriate)
• Updated CHANGELOG.md (if appropriate)
• Made changes based off the latest version of Splink
• Run the linter
• Run the spellchecker (if appropriate)