As Eric said, it is in stdlib not addon, so that you have to wait until the next base library update.
It is the same as you patch except the name "unicode4" is used instead of "literal4" for consistency with "unicode" which is needed by backward compatibility. For j dictionary, correctness is more important than backward compatibility so that data type "unicode" has been renamed to "literal2" to be consistent with "literal4" On Aug 26, 2016 12:10 AM, "robert therriault" <[email protected]> wrote: > Hi Bill, > > Have you updated datatype to deal with literal4? > > datatype > 3 : 0 > n=. 1 2 4 8 16 32 64 128 1024 2048 4096 8192 16384 32768 65536 131072 > t=. '/boolean/literal/integer/floating/complex/boxed/extended/rational' > t=. t,'/sparse boolean/sparse literal/sparse integer/sparse floating' > t=. t,'/sparse complex/sparse boxed/symbol/unicode' > (n i. 3!:0 y) pick <;._1 t > ) > datatype 9 u: 16b1f600 > |index error: pick > | (n i.3!:0 y) pick<;._1 t > > datatype=: 3 : 0 NB. My definition to patch over the error > n=. 1 2 4 8 16 32 64 128 1024 2048 4096 8192 16384 32768 65536 131072 > 262144 > t=. '/boolean/literal/integer/floating/complex/boxed/extended/rational' > t=. t,'/sparse boolean/sparse literal/sparse integer/sparse floating' > t=. t,'/sparse complex/sparse boxed/symbol/unicode/literal4' > (n i. 3!:0 y) pick <;._1 t > ) > datatype 9 u: 16b1f600 > literal4 > > I copied libj.dylib across to a previous j805 beta to upgrade to version > 11. If you have updated datatype, then it could be that Package Manager is > showing everything up to date and has not caught a change to datatype > because of the way that I updated. Thoughts? > > Cheers, bob > > > On Aug 24, 2016, at 4:57 PM, bill lam <[email protected]> wrote: > > > > see updated documentation of u: in _j805_ addon docs/help. svn copy here > > > > http://www.jsoftware.com/svn/addons/trunk/docs/help/dictionary/duco.htm > > > > since j805 is still in beta, regular help pages are for j804 at the > moment. > > > > 3 u: 7 u: 16b1f600 > > 55357 56832 > > > > this is intended, 3&u: works as if it is rank-0, ie atom by atom. To find > > codepoint of any character or string, first convert it into utf-32, > > > > 3 u: 9 u: foo NB. see u: > > > > rendering of unicode characters depends on font engine and font being > > used. Try another session font or unifont. > > > > linux jqt can display smiley face emoji but in monochrome. > > > > Finally notice the differences between datatype and encoding. For all > > operations on literal/literal2/literal4 datatype, J engine ignores > unicode > > utf8, utf16 and utf32 encoding except, > > a. display to session/jconsole > > b. monad ": verb > > c. dyad 7/8/9 u: verb > > > > Even monad u: and 10&u: are unrelated to unicode encodings and they can > > give results with invalid unicode codepoints. > > ---------------------------------------------------------------------- > > 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
