I made a start yesterday at implementing it just to see what kinds of problems there are, and it is very wormy indeed.I agree this would be a good idea. I was going to have a go at implementing it and then I started to think it might turn into a can of worms.
At the moment, strings in source files are interpreted as being in the bbc character set unless they represent the names of files on the host, and then they are treated as being in the ambient encoding. Typically windows-1252 on Windows or utf-8 on Linux.
A simple fix would be to display a warning if any characters in the source have the top bit set. The warning could suggest replacing pound with backtick. This could be disabled with a command-line option "-source-encoding copy". Like now, this wouldn't do any translation.
This would have avoided lovebug's problem, doesn't break anything and leaves open the possibility of adding "-source-encoding utf8".
Statistics: Posted by ctr — Tue Apr 15, 2025 4:54 pm