Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!mips!sgi!dragon!dblues.wpd.sgi.com!jackr
From: jackr@dblues.wpd.sgi.com (John Repenning)
Newsgroups: comp.emacs
Subject: gin-mode for GNU Emacs
Keywords: GNU, gin-mode, indentation, text-mode
Message-ID: <1990Dec3.190053.10160@relay.wpd.sgi.com>
Date: 3 Dec 90 19:00:53 GMT
Sender: news@relay.wpd.sgi.com ( CNews Account )
Reply-To: jackr@dblues.wpd.sgi.com
Organization: Silicon Graphics, Inc.
Lines: 53

I have a copy of a minor mode, called ``gin-mode'', which modifies
text-mode so that the indentation functions guess the appropriate
indentation and follow it - no more set-fill-prefix!  And, no more
"ah, shoot!  I forgot to reset set-fill-prefix" either!!!  The code
was written by Martin Neitzel, around a year ago, and I believe I got
it (indirectly) from this news group, so I wont repost it (but if
someone wants it, I'd be happy to provide you a copy - email me).

However, I've found what appears to be a simple, minor bug in it.  The
code that handles hanging indents doesn't recognize the ones that look
like this:

	1. 	Paragraph about something or other, long enough to
		require several lines
	...

This recognition is controled by a regular expression, which seems to
be trying to include this format, but misses it by just a hair:
instead of being defined like this:

(defvar gin-left-hang-indent-re
  "\\s *\\([-*]\\|([a-zA-Z0-9])\\|[a-zA-Z0-9]\\.?:]?\\)\\s +"
  "*Regexp that defines a hanging indent of a paragraph.
If it is seen by gin-guess-prefix, the next lines are indented with
white space beyond the hanging indent.  Setting this variable makes
it buffer-local.")

...it shouls be like this:

(defvar gin-left-hang-indent-re
  "\\s *\\([-*]\\|([a-zA-Z0-9])\\|[a-zA-Z0-9]\\.?:?]?\\)\\s +"
  "*Regexp that defines a hanging indent of a paragraph.
If it is seen by gin-guess-prefix, the next lines are indented with
white space beyond the hanging indent.  Setting this variable makes
it buffer-local.")

The change is in the sequence "\\.?:]?", near the end - I've made the
colon optional, like the period and the bracket.

Or, maybe the sequence should just be "[].:]?", to permit zero or one
of any of these characters - strictly, the present code allows a form
like 

	1.:]	stuff

...which doesn't look like anything anyone might want.

I thought of sending this change to Martin (with whom I've
corresponded before), but I've lost his new address.  (Do you read
this group, Martin?  Do you agree with my interpretation?)

Jack Repenning
jackr@dblues.wpd.sgi.com


Brought to you by Super Global Mega Corp .com