Changed

• Conceptually renamed "handlers" to "receivers"
• Made each module export only a list of receivers (Intent system that @9nine9nine9 described) so it is easily extensible and future-proof
  • Adding a new command in an existing module only requires a change in that module
  • Creating new modules requires you to register the module in EventHandler and the cabal file as before
• Abstracted away DiscordHandler (Either RestCallErrorCode Message) to a cleaner DiscordHandler () since the returned Message isn't used anyway
• Gave all receivers a consistent type signature (Message -> DiscordHandler (), or ReactionInfo -> DiscordHandler ())
• Use T.Text for all possible places apart from when interacting with functions that require String (e.g. disk IO), and Main
• Added new Util methods to create commands: newCommand and newDevCommand. This abstracts away RegEx matching and subgroup capturing as well as dev checking
  • This also puts all the RegEx into the receiver module itself, so all RE files (apart from TemplateRE) has been removed
• Formatted the entire codebase: aligned imports, comma-broken lists (diff becomes cleaner), and ~80 char wraps
  • Made code structure somewhat consistent throughout the codebase, so new additions can easily reference styles from wherever it wants
• Renamed several functions to be more representative of what they're responsible
• Removed any legacy unused code
• Updated README with more information, added a Dockerfile (untested) for better deployment
• Fixed several issues that were open
• Created a Parsec parser for CSV files, and delegated all CSV work into there (includes unit tests)
  • IMPORTANT: Changed location of csv storage from src/config/~~.conf to .owen/~~.csv.
  • IMPORTANT: All .csv files now require a new line at the end
  • IMPORTANT: fame.csv and devs.csv are now ONE ITEM PER LINE, instead of a comma separated single line

Unchanged

• Nothing...

Requesting some local testing or review before actually merging, because although I tested on my machine, most of the code is newly written and therefore prone to mistakes/errors. (...or we can just merge and test in production lmao)