Combine existing translations with foundry-magic-l10n

[ben]

This tool could help get us up and running with lots of l10n text.

• Run current text through the automated translation
• It looks like it comes back as key-value, and not structured JSON, so massage it back into shape
• Try to keep the manual translation work that's already been done

[rsek]

r20's ironsworn translations are apparently an MIT license, so there's probably a bunch we could grab from there too. r20 uses key-value pairs so it's a mainly a matter of mapping one key convention to another. the translations are complete for a few languages (French for instance), including all oracle strings, meaning there's over 2000 strings total, though some are repeated.

SF's translations theoretically exist, but last time i checked there wasn't a ton of movement on them.

where possible, i think it'd be useful to centralize this so that any project can benefit from it, and make it referenceable against dataforged.

one possibility for single-language output files is having a flat object with Parent/Object/ID#Key, e.g. Starforged/Assets/Module/Engine_Upgrade/Abilities/1#Text; alternatively, it could just be a matching "skeleton" of dataforged with only the user-facing string keys for a given language.

a multi-language "master" file might parallel either, except instead of string values it would have a hash object of values keyed by locale code.