Ah, sorry, I had not noticed that you were discarding the result of 9&u:
Thanks, -- Raul On Tue, Mar 22, 2022 at 11:23 PM Elijah Stone <[email protected]> wrote: > > Neither of those corresponds to the verbs I posted. > > latin2utf8_v1=. (9&u: ] ]) :: (8 u: 10 u: ]) > latin2utf8_v2=. 8 u: (9&u:) :: (10&u:) > valid_utf8=. 8 u: 10 u: ?1e6$16384 > invalid_utf8=. a.{~?1e6$256 > utf8s=. valid_utf8;invalid_utf8 > datatype@:latin2utf8_v1 &.> utf8s > ┌───────┬───────┐ > │literal│literal│ > └───────┴───────┘ > datatype@:latin2utf8_v1 &.> utf8s > ┌───────┬───────┐ > │literal│literal│ > └───────┴───────┘ > timex 'latin2utf8_v1 valid_utf8' > 0.0111741 > timex 'latin2utf8_v2 valid_utf8' > 0.0235571 > NB. performance is the same on invalid utf8 > > On Tue, 22 Mar 2022, Raul Miller wrote: > > > On Tue, Mar 22, 2022 at 7:47 PM Elijah Stone <[email protected]> wrote: > >> It's not; it should give exactly the same result. > > > > datatype 9 u: 8 u: 10 u: 166 97 98 99 { a. > > unicode4 > > datatype 8 u: 9 u: 8 u: 10 u: 166 97 98 99 { a. > > literal > > > > -- > > Raul > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
