Line length calculated in bytes

[yurikhan]

Characters represented by multiple bytes are counted incorrectly. For example, consider a comment written in Russian language using Cyrillic script which, when encoded in UTF-8, takes 2 bytes per letter:

-- ..5...10....5...20....5...30....5...40
-- Я могу есть стекло, оно мне не вредит.

$ locale|grep CTYPE
LC_CTYPE="en_US.UTF-8"

$ luacheck --max-line-length 50 test1.lua
Checking test1.lua                                1 warning

    test1.lua:2:51: line is too long (70 > 50)

Total: 1 warning / 0 errors in 1 file

[yurikhan]

Sorry, dupe of #136.