Text() cannot print combined Unicode characters with hardware keyboard method on Windows

[comodoro]

If I invoke the Text action with combined characters (like ď, ć) and I use a keyboard layout that supports them, I get an error and no output.

For example, on the Czech keyboard, common letters ď, ť, ň can only be created by pressing the combining caret character (shift+= on the english keyboard) followed by the particular base letter.
Text('ď').executre() gives the error

Execution failed: Keyboard interface cannot type this character: 'ď'

while Text('ěščřžýáíé').executre() executes without problems.

This can be avoided by setting unicode_keyboard = True in settings.cfg in .dragonfly-speech In user directory, but I wouldn't be comfortable modifying user settings just for my program and have not found a mechanism for setting this when importing dragonfly; does it exist? Still it would be nice to have the option to use scan codes as some applications seem not to be able to handle unicode.

[comodoro]

For the current master (not the pypi version) I'm able to do

        if os.name == 'nt':
            dragonfly.actions.keyboard._win32.Win32Keyboard.unicode_keyboard = True

But this is a hack.

I'm biased and this is a matter for discussion, but I think respecting modifier keys in Text does not outweigh these problems and therefore that the direct text method (unicode_keyboard = True) should be the default. Even for English; after all, even the English type foreign names.