My source database mainly uses UTF8, but some old library encoded Windows Latin1 encoded strings encoded inside it.
I was hoping Boost would have a clear conversion function, but I did not find one. Do I need to manually code this common solution?
Looking for a portable solution that runs on Linux.
( This Q is similar, but not quite the same)
Edit: ICU seems to be the correct answer, but it will outsmart a little for my needs. I ended up with a string replacement for the well-known several extended characters that were used.
source share