Hi all, I've converted all the string code in the scratchpad application to unicode.
Text files and text from mzscheme is converted from utf-8 into utf-32 internal representation, and all output back into utf-8. This means that multibyte characters can now be edited normally (rather than the cursor disappearing in the middle of them), and should be displayed assuming the font contains a glyph for them. The problem is keyboard input, which glut seems to squash into a single byte for the one multibyte key I can find on my keyboard (£) could someone test this a bit more? Portuguese characters seem to now work in the console help properly. This is quite a big change, for those want to play with the scratchpad, be aware that you have to use wstring everywhere and declare string constants L"like this". Although the wchar_t is 4 bytes on linux (and probably/I hope OSX too) its 2 bytes long on windows - so don't de-reference these pointers directly. cheers, dave
