When Kitty key disambiguation protocol is in use, shifted symbols show up as non-shifted.

[Spritetm]

What I would like is to get a hold of the keysym that is pressed when taking into account modifier keys, specifically 'shift'. When not using the Kitty key disambiguation protocol, ncurses-input returns this (from what I can see in the id and/or utf8 fields). When using the protocol, ncurses-input returns the expected output for alphabetical keys (shift-a returns 'A') but not for other keys (shift-1 returns '1' rather than '!').

(An easy way to test this is to run notcurses-input both on Kitty and some other terminal, e.g. screen, and compare what happens when you press shift-1.)

Because it's impossible to calculate the shifted symbol from the non-shifted one (as I don't think there's a way to know the keyboard layout in use), I'd like some way to have the shifted symbol returned in the ncinput event struct. From what I can tell, kitty does return this so the info is available.

Note that this also affects widgets like ncreader, to the point of making it impossible to enter symbols like @#$%^&*() in this widget when using kitty. As such, I'm filing this as a bug rather than a feature request.

Please include the following data:

• export | egrep 'LANG|LC_CTYPE|TERM'
  declare -x COLORTERM="truecolor"
  declare -x LANG="en_US.UTF-8"
  declare -x LANGUAGE="en_US:en"
  declare -x TERM="xterm-kitty"

• notcurses version (available from notcurses-demo i)
  Latest master (v3.0.9-4-g90135203e)

• terminal name + version
  kitty 0.21.2 created by Kovid Goyal

[dankamongmen]

thanks for the great writeup! i haven't been as responsive to notcurses issues of late as i was when i was developing it fulltime, but i hope to address this soon.

[dankamongmen]

so we're done here, yes? outstanding work, thank you so much! i hope you're enjoying the library otherwise.

[Spritetm]

Yes :) Library is great otherwise, it's really nice to have a non-antique abstraction over all the things a terminal can do.