Fix being unable to type uppercase characters on Windows

[kirillsemyonkin]

Fixes the problem with Windows being unable to type out uppercase characters when using Key::Unicode keys (e.g. Key::Unicode('T')).

Upon using the key function that types out Unicode keys for uppercase characters, the call returns the error below:

error when mapping keycode to keysym: (Could not translate the character to the corresponding virtual-key code and shift state for the current keyboard)

Apparently the call for VkKeyScanW in enigo's code does not handle the uppercase characters: the code is translated into two bytes (as huge comment above the call mentions) as a single u32 (upper u16 of which isn't used?), which later was not handled and passed over to MapVirtualKeyW, which fails because it cannot translate two-byte value (e.g. 340 for 'T') into a scancode. This PR handles the two bytes by emitting extra necessary scancodes (e.g. SCANCODE_LSHIFT) depending on the flags in the high-order byte.

P.S. Why does Windows implementation mention keysym in its errors?

[kirillsemyonkin]

Mind that for this PR I have not checked for Issues or other PRs or branches - I just did what happened to fix my problem, and would like to see it up on Cargo crates.io as soon as possible (it is a major feature that does not work on Windows without this fix), so that I can use it in a project I would like to publish soon-ish.

[kirillsemyonkin]

Something to probably consider for the future: these non-...ExW calls probably use the current (e.g. en-us) keyboard layout, which probably means it will be unable to type out keys from other languages, unless user explicitly switches layout (or an Alt+Shift combination is generated - which is configurable on many OSes and, thus, unusable in crossplatform way). The ...ExW allows to input a particular keyboard layout (which has to be preloaded with a LoadKeyboardLayoutW), which then I, honestly, have no idea how to figure out which layout the unicode char belongs to (maybe iterate over user's layouts and select non-failing mapping only from those?).

[pentamassiv]

Thank you very much for your PR. I hope to have time to test it in the beginning of next week.

Regarding the mapping of keys, I created a PR to check the current layout of the user and use that #280

[pentamassiv]

I didn't have time to test it with the borrowed machine but will have access to it at the end of may again. Hopefully I can test it then.

[kirillsemyonkin]

I can give you a remote Windows VM if you want