Open patrickfrey opened 5 years ago
A possible solution was proposed by forcing the use of a virtual character table with the option %LEXER BYTECHAR in the pattern matcher program or the defineOption method in the pattern lexer interface. This is not a fix in general but may shift some character definitions and help in some cases (e.g. german umlauts).
For character sets in IsoLatin-1 the %LEXER BYTECHAR works.
Some word boundaries are not detected correctly in UTF-8 input.