emscripten: Make an attempt at correct keyboard scancode/keycodes. #6364
Conversation
This uses a newer browser API to get physical scancodes, but still uses the (deprecated) event field that we were already using for scancodes, but for keycodes instead now, which appears to be more accurate. Since keyboard layout isn't (generally) available to web apps, this adds an internal interface to send key events with both scancode and keycode to SDL's internals, instead of sending just scancodes and expecting SDL to use its own keymap to generate keycodes. Future work in this area would be to use the keyboard layout APIs on browsers that support them, which would allow us to use SDL's usual keymap code and not rely on a deprecated browser API, but until we get there, this patch gives significantly more correct results than we would have before. Fixes libsdl-org#2098.
|
If only there was a (standard) property for "with layout, but not modifiers" on key events... Anyway, this does fix the scancode for the '#' key from Edit: Hmm, the |
I assume the solution to this will eventually be the Keyboard Layout Map API, which works in Chrome-based browsers but not Firefox or Safari yet. I definitely don't have time to mess with this right now, but it might be worth making the current code into a fallback for when that API isn't available (and eventually remove it entirely). But for now, I'm merging this PR, as it seems to improve the current state of things. |
|
I'm currently trying to roll this into emscripten and running into failure in the The 2.24.2 the test is generating: With 2.26.0 (I assume due to this change) the test is generating: I'm hoping this is perhaps the test harness that needs updating in the way it sends fake key events? Hopefully this isn't a breaking change for users? |
|
Indeed, it looks like it was out fake keyboard events that were causing the issue. Fixed in emscripten-core/emscripten#21337 |
emscripten: Make an attempt at correct keyboard scancode/keycodes.
This uses a newer browser API to get physical scancodes, but still uses the (deprecated) event field that we were already using scancodes, but for keycodes instead now, which appears to be more accurate.
Since keyboard layout isn't (generally) available to web apps, this adds an internal interface to send key events with both scancode and keycode to SDL's internals, instead of sending just scancodes and expecting SDL to use its own keymap to generate keycodes.
Future work in this area would be to use the keyboard layout APIs on browsers that support them, which would allow us to use SDL's usual keymap code and not rely on a deprecated browser API, but until we get there, this patch gives significantly more correct results than we would have before.
Fixes #2098.