Gnu Emacs EDT Emulation - Introduction - 1/3
Mike Clarkson
mike at yetti.UUCP
Thu Aug 28 09:30:33 AEST 1986
Here's my EDT emulation for GNU Emacs that is based on the EDT emulation
for Gosling's Emacs sent out on the net a couple of years ago by Lynn Olson
at Tektronics. This emulation was widely distributed as the file edt.ml
in the maclib directory of most Emacs distributions.
My emulation consists of two files: edt.el and edtdoc.el. The edtdoc.el file
is the documentation, that you can add to the beginning of edt.el if you
want. I have split them because I have been loading the edt.el file a lot
during debugging.
I will gladly take all criticisms and complaints to heart, and will fix
what bugs I can find. As this is my first elisp hack, you may have to
root out a few nasties hidden in the code. Please let me know if you
find any (sorry,
no rewards :-). I would also be interested if there are better,
cleaner, faster ways of doing some of the things that I have done.
You must understand some design considerations that I had in mind.
The intention was not really to "emulate" EDT, but rather to take advantage
of the years of EDT experience that had accumulated in my right hand,
while at the same time taking advantage of EMACS.
Some major differences are:
HELP is describe-key;
GOLD/HELP is describe-function;
FIND is isearch-forward/backward;
GOLD/HELP is occur-menu, which finds all occurrences of a search string;
ENTER is other-window;
SUBS is subprocess-command. Note that you will have to change this
yourself to shell if you are running Un*x;
PAGE is next-paragraph, because that's more useful than page.
SPECINS is copy-to-killring;
GOLD/GOLD is mark-section-wisely, which is my command to mark the
section in a manner consistent with the major-mode. It
uses mark-defun for emacs-lisp, lisp, mark-c-function for C,
and mark-paragraph for other modes.
Some subtle differences are:
APPEND is append-to-buffer. One doesn't append to the kill ring much
and SPECINS is now copy-to-killring;
REPLACE is replace-regexp;
FILL is fill-region-wisely, which uses indent-region for C, lisp
emacs-lisp, and fill-region for others. It asks if you really
want to fill-region in TeX-mode, because I find this to be
very dangerous.
CHNGCASE is case-flip for the character under the cursor only.
I felt that case-flip region is unlikely, as usually you
upcase-region or downcase region. Also, unlike EDT it
is independent of the direction you are going, as that
drives me nuts.
I use Emacs definition of what a word is. This is considerably different from
what EDT thinks a word is. This is not good for dyed-in-the-wool EDT fans,
but is probably preferable for experienced Emacs users. My assumption is that
the former are a dying breed now that GNU Emacs has made it to VMS, but let me
know how you feel. Also, when you undelete a word it leave the point at the
end of the undeleted text, rather than the beginning. I might change this
as I'm not sure if I like this or not. I'm also not sure if I want it to
set the mark each time you delete a character or word.
Backspace does not invoke beginning-of-line, because ^H is the help prefix,
and I felt it should be left as such. You can change this if you like.
The ADVANCE and BACKUP keys do not work as terminators for forward or
backward searches. In Emacs, all search strings are terminated by return.
The searches will however go forward or backward depending on your current
direction. Also, when you change directions, the mode line will not be
updated immediately, but only when you next execute an emacs function.
Personally, I consider this to be a bug, not a feature.
This should also work with VT-2xx's, though I haven't tested it extensively
on those terminals. It assumes that the CSI-map of vt_200.el has been defined.
There are also a whole bunch of GOLD letter, and GOLD character bindings:
look at edtdoc.el for them, or better still, look at the edt.el lisp code,
because after all, in the true Lisp tradition, the source code is *assumed*
to be self-documenting :-)
Mike Clarkson, ...!allegra \ BITNET: mike at YUYETTI or
CRESS, York University, ...!decvax \ SYMALG at YUSOL
4700 Keele Street, ...!ihnp4 > !utzoo!yetti!mike
North York, Ontario, ...!linus /
CANADA M3J 1P3. ...!watmath / Phone: +1 (416) 736-2100 x 7767
Note that I am not on ARPA, and must gateway any ARPA mail through BITNET or
UUCP. If you have a UUCP or BITNET address please use it for communication
so that I can reach you directly. If you have both, the BITNET address
is preferred.
--
Mike Clarkson, ...!allegra \ BITNET: mike at YUYETTI or
CRESS, York University, ...!decvax \ SYMALG at YUSOL
4700 Keele Street, ...!ihnp4 > !utzoo!yetti!mike
North York, Ontario, ...!linus /
CANADA M3J 1P3. ...!watmath / Phone: +1 (416) 737-2100 x 7767
"...the most inevitable business communications system on the planet."
- ROLM magazine advertisement
which planet?
More information about the Comp.sources.unix
mailing list