[Python-ideas] Accessible tools

Stephen J. Turnbull stephen at xemacs.org
Sat Feb 21 07:37:40 CET 2015


We should probably take this off-list, since the OP hasn't actually
shown interest yet.  Assuming there are any replies ;-), Reply-To set
to me, will summarize.

Nikolaus Rath writes:
 > On Feb 19 2015, "Stephen J. Turnbull" <stephen-Sn97VrDLz2sdnm+yROfE0A at public.gmane.org> wrote:

 > > Emacs has most of the features you're talking about in some form
 > > already.  I admit it is questionable whether the quality is up to the
 > > standards of Eclipse or Xcode.  For example, Emacs's completion
 > > feature is currently not based on the full AST and so doesn't come up
 > > to "intellisense" standards.
 > 
 > Are you sure? I believe that jedi effective uses the AST. It actually
 > parses the code in a subordinate Python interpreter and uses its
 > interspection capabilities.

I'm sure you're right about jedi.  I was speaking in general.  As
perhaps you are aware, emacs-devel recently had an intense flamewar
over using the ASTs from LLVM and GCC, and that proposal is currently
on hold (RMS is "studying the matter").

 > > but it is based on a generic parser called "semantic", so it wouldn't
 > > be hard to teach it.
 > 
 > Well, not so sure about that. There's the slight hurdle of learning
 > ELisp first.

I know Elisp (and have already volunteered some assistance), as does
Dr. Raman (whose availability I can't speak to, but I'm sure he'd
offer at least moral support).  The Emacs developers will surely be
helpful.  I believe the effort would be small enough to be justified.

 > Make sure to look at Emacs 24. All of the above can also be set up
 > for Emacs 23, but using 24 will save you a lot of time.

I would say go straight to trunk (which will be Emacs 25), which will
be in pretest in the near future.  By the time the new IDE is release
ready, Emacs 24 will be out of support (Emacs only supports the
current release, so that's less than a year away).



More information about the Python-ideas mailing list