A way to read (from a console) a mix of “bindable keypresses” and Unicode?

[ilyaza]

[Given a dearth of documentation on robust handling of user character input on Windows, I do not know whether this is a feature request, or not. So I post it here.]

As far as I understand, there is no way for a console application to process a mix of input of “accelerators” keypresses (such as Ctrl-Shift-= etc) and of Unicode characters. The major stumbling block is the recognition of dead keys: the application cannot tell a dead key from an “accelerator”.

(Using the undocumented flag of ToUnicodeEx(), one could do it if one knew the current “input locale” — but I see no way to find the input locale due to the console wrapper running with very high elevation…)

Question 1: is the analysis above valid? Or maybe there is some clever trick to circumvent this?

Question 2: is there going to be changes in this regard, so that applications can recognize dead keys

Question 3: A way to see changes of the “input locale” would be also beneficial!

[ilyaza]

This defect makes it impossible to write a robust console application in Windows with customizable key-bindings. And a backwards-compatible solution seems to be quite simple:

• for events generating a deadkey, set an unused bit in the “extended scancode” field.

The bit DONTCARE_BIT is guaranteed to be not set on the key events delivered to/by the console. So just use this bit — so that the applications know that they do not need to interpret this keypress as a bindable event.

(I discuss this defect in detail in my list of defects of keyboard input on Windows.)