drop regex-posix in favour of regex-tdfa?
|Reported by:||simonmar||Owned by:|
|Type of failure:||None/Unknown||Test Case:|
|Related Tickets:||Differential Rev(s):|
regex-tdfa is written in Haskell (regex-posix uses a C backend), and it uses the standard DFA algorithm compared to regex-posix's naive backtracking, so it is asymptotically better (though it would probably be a good idea to check performance on some non-pathalogical examples to make sure the constant factor isn't too bad).
For GHC 6.8 we should drop regex-posix in favour of this.