The update is awesome. Thanks!
How on earth did you get the furigana working correctly for vocabulary? That seemed hard to me. I thought the API only returned the complete readings (including any embedded kana). Very cool.
I notice that it doesn’t show the reading furigana after meaning quizzes. Not a problem, but I wonder if that’s related to how you generate the <ruby> markup?
Also, just FYI: It looks great with a userstylesheet to put the furigana below and dark theme. This is what I see during reviews (with a locally modified version of WKElementaryDarkTheme and furigana below user stylesheets):
You could optionally inject style="ruby-position: under;" on the <ruby> tag to force the furigana underneath instead of on top it you wanted.
Now to train my fingers to stop reflexively typing fe！
Final edit.
This is really great!
It doesn’t seem like this would be that big of a time saver, but I’m flying through my reviews without having to manually expand everything. I notice that I ignore the script output the majority of the time (correct answers). I just hit return immediately. But not having to expand and scroll for the occasional incorrect answer really seems to remove some mental friction and speeds things up considerably.
Seriously: this is a significant UX improvement. Nice work!