Closed rainoftime closed 3 years ago
That's a good observation! The differences in behaviour are due to the supported escape sequences. CVC4 follows the standard literally, and only accepts the escape sequences "" and \u..., whereas Ostrich and Z3 also support \x... (as common in JavaScript, etc.), and therefore interpret \x4a and \x4A as the same character.
This is something I'm trying to clean up in Ostrich at the moment.
Hi, for the following formula
ostrich and z3 yield
sat
, but cvc4 givesunsat
.Intuitively, I may respect the result would be
unsat
. Is this different because of the changes in the SMT-LIB2 standard?