On Apr 7, 1:51 pm, "Edward K. Ream" <edream...@gmail.com> wrote:
> However, the real problem is inferring (deducing) the type of x.y.z > given only x as the starting point. In general, this is a hard > problem. This morning I thought that some slight of hand might allow > the use to specify types "the first time". This can be done, but > notice that given x, we still have to specify the type of x.y, and > similar for longer chains. I doubt this scheme would be pleasant to > use. Imo, we really do want to do type inference. > I'm missing something here, the only thing I see is having to determine is if we are invoking a callable, (hence requiring parens with some sort of argument "entry assistance"), or attribute, (for retrieving or setting). Its hard to see what is being gone for ahead of time as you, usually, haven't written that part yet. Even with an "inspect to build dictionaries" scheme, there is a place for data structure populating via python_cmpltns.py, after all, we have all of the standard modules which can be "pre-inspected", (note, we probably have to have 2.x and 3.x and a way to specify the target python version, something else to take into consideration. Tom -- You received this message because you are subscribed to the Google Groups "leo-editor" group. To post to this group, send email to leo-edi...@googlegroups.com. To unsubscribe from this group, send email to leo-editor+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/leo-editor?hl=en.
