Improve steno "keymap" display

[lexidevs]

Currently, the keymap display for steno is largely ineffective. In static mode, it works fine, but in react or next mode, it shows letters as if typed individually, without chords. For example, when the next word is "thing," it highlights the "t" keys, suggesting a stroke of T-T, while the correct steno stroke is "THEUPBG." Additionally, if a word starts with a letter not in the steno layout, nothing is displayed.

This limits the keymap display's usefulness for stenography. From experience, having a visual representation of the keys needed for a stroke is very helpful when learning steno, as it requires understanding which buttons correspond to specific sounds, rather than just single letters.

I would improve this feature by translating words into their corresponding strokes, and displaying those strokes, rather than letters. This is somewhat similar to DiDoesDigital's Typey Type (github), though Typey Type hard-codes the corresponding stroke into the dictionaries for lessons. By implementing a simple stroke lookup within plover-compatible dictionaries (perhaps supporting custom dictionaries as well?) the correct stroke could be displayed to the user.

Perhaps the best place to start would be, similar to Typey Type, allowing custom lessons to have a list of words and their corresponding strokes. This would be much easier to implement than full translation.

Is this something that others are interested in being part of monkeytype? I'm open to implementing it, if so. I have prior experience with similar word-to-steno translation. I know that previously, the idea of emulating steno has been shut down, which makes sense -- that's a much harder problem and there already exists purpose-built software (namely, plover) to do the same thing.

[Miodec]

Doesnt that require a list of which words can be typed with what chords? Isnt that going to be massive? What if a word is not included?
I kinda feel like this is out of the scope of this project as its built primairly for 'normal' typing. Steno keymap was added as just a novelty, nothing serious.

[lexidevs]

The list of words -> chords would probably just be the plover dictionary, or perhaps user-uploaded dictionary. The main plover dictionary is ~4MB, though I think it'd be reasonable to only fetch it when the user enables the feature.

If a word isn't included, you can fall back to fingerspelling (letter by letter, as is done for rare proper nouns and such).

I sort of felt like you might say that it's out of scope, and that's totally fair -- that's why I wanted to gauge interest among others. If you figure it's unlikely to go upstream, I may implement it myself anyway and just use it for my own practice.

[Miodec]

I mean, as long as its not a massive rewrite of the keymap system, maybe its not that big of an issue.

[lexidevs]

Also to note:

Lots of steno people who use monkeytype have basically tried to do this themselves, either by themeing plover keyboard layout (itself a plugin), e.g. https://www.youtube.com/watch?v=yChj7aD1vHY & https://www.youtube.com/watch?v=8Ixtfo3I4VE, or by writing their own extensions (e.g. https://chromewebstore.google.com/detail/monkeytype-steno-assist/cjpibklglegjgalionmdnhajmpeepphf) to display the stroke/key layout of upcoming words on monkeytype.