Non-ASCII Latin and Cyrillic chars count twice, and East-Asian chars count 3 times.
For example:
á, д and ß count twice: count "á" returns 2
And:
我, あ and 한 count 3 times.
This breaks commands such as select for selecting text.

This needs an awkward workaround atm:
count toArray "我"
and: toString (toArray "我" select [0,1])

Also related:
https://feedback.bistudio.com/T154207