How about: qualify name in the calling locale? In other words, if calling locale is FOO
abc becomes abc_FOO_ abc_xyz_ remains abc_xyz_ abc__Y with Y_FOO_ being X becomes abc_X_ This covers most of the important issues I can think of... The remaining issue (getting the calling locale an identifier) can be achieved by passing an empty string instead of a name. Thanks, -- Raul On Wed, Apr 5, 2017 at 3:49 PM, 'Pascal Jasmin' via Programming <programm...@jsoftware.com> wrote: > well, 18!:51 proposal, > > Returns name of the caller's locale (the locale that execution will return to > after execution of the current function). > > > But, using it would be awkward compared to passing qualified names to begin > with. This can already be done generically with: > > loc =: (,&'_'@[ ,&'_'@, ":@>@])"1 0 boxopen > locs_z_ =: 1 : 'm loc 18!:5 ''''' > > 'a' locs > a_base_ > > but I think the use case Raul is thinking of is that if the user/caller did > not provide a localized version of the name, the adverb/function that knows > its a name, could localize it for them. That would certainly be convenient. > For example, > > myadv_far_ =: locs (1 : 'u~ y') NB. cannot include locs (relevant/relative > to caller) as part of localized definition. > > > 'a' myadv_far_ > 'a_far_' (1 : 'u~ y') > > > ________________________________ > From: Henry Rich <henryhr...@gmail.com> > To: programm...@jsoftware.com > Sent: Wednesday, April 5, 2017 3:08 PM > Subject: Re: [Jprogramming] locales with adverbs and conjunctions? > > > > Oooh, let's talk about this. > > What would you like to see? > > I have long wanted to be able to evaluate a name in the context of the > caller of an adv/conj, so that I can pass a private name into a > conjunction and have it evaluated there. > > If we can come up with a sound proposal, it can be implemented. > > Henry Rich > > > On 4/5/2017 1:31 PM, Raul Miller wrote: >> The inability of determining the locale of the calling context (to >> qualify non-locative names with it) has bothered me for some time >> (many years). >> >> However, I did not have a convincing use case where it was significant >> (I was trying to work with trace, and that apparently was not >> important enough to warrant addressing this issue). >> >> I think something should be done about it, but I'll leave the details >> of that to the ISI folks. >> >> Thanks, >> > > ---------------------------------------------------------------------- > 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
