Daniel Berger wrote: > +%token tEQU /* ãé¡× */ I don't know...is the lex even capable of parsing unicode? if it is, you may have to change some more code, somewhere in the scanner...would be interesting. ¡â for !=, for example. [murphy]