More pervasive measure word audio generation

[batterseapower]

MW audio is only being generated in expression onfocuslost. We could add either:

1. A feature like we have for the audio field currently where you can type stuff into the MW Audio field and get it reformatted automatically.

2. An onfocuslost hook for the MW field which generates the audio. A bit trickier, because we have to parse the MW field to make that work.

[Nick3C]

The second option is a much better solution in my view. It will be a lot easier to add things (paste or type in the characters and have the audio field update automatically).

I am not really that keen on feature 1 even for the audio field because it causes redundant typing to fill both reading and audio field. While I'm not in a rush to change the audio field (because it does work) I don't think it's a good idea to add the behaviour somewhere else.