The file contents could be of str type. Should use byte length instead
of str length, otherwise utf-8 encoded files may not get properly parsed
Source issue: https://github.com/ncm2/ncm2-pyclang#2
Nitpicking: the description only mentions changes related to file contents but this modification (fspath(name) -> b(fspath(name))) likely fixes a different issue, it may be worth mentioning this in the commit message. Unless this extra b() call is not strictly necessary here but added for consistency?