Re: Using readline with Emacs shell or eshell or ???

[Robert L. Knighten]

Ilya Zakharevich <nospam-abuse@ilyaz.org> writes:

> [A complimentary Cc of this posting was sent to
> Stefan Monnier
> <monnier+gnu.emacs.help/news/@flint.cs.yale.edu>], who wrote in article 
> <5lhe8fion1.fsf@rum.cs.yale.edu>:
> > > For an example that works (but might not be the best implementation)
> > > take a look at the inferior-octave-complete function in the Emacs mode
> > > for running Octave inside Emacs (octave-inf.el).  The Emacs function
> > > for completion sends a command to Octave to generate completions and
> > > then displays them.  It uses comint for some of the ugly stuff, so the
> > 
> > I haven't looked at how octave-inf.el does it, but gud.el does it
> > as well when interacting with gdb, so you can also look at that
> > one.  If you look at both and get some insight, I'd like to
> > hear about it.
> 
> IIRC, gdb has "complete" command.  GUD erases the current line, sends
> the complete command, processes the output, then restores the current
> line and acts on the completion list.
> 
> Do not think one can do anything with a random non-cooperating program.
> 
> Hope this helps,
> Ilya
> 
gud was the first thing I looked at, but as noted it is based on a really
tight interaction with the debugger and is very complex as a result.

I'v now played with octave-inf.el and even it is overkill for the rather
simple program that I am building.  But it is very clean and it appears that
it will serve as an excellent base for what I do need.

Thank you for the comments.

-- Bob

-- 
Robert L. Knighten
Robert@Knighten.org