Path: utzoo!attcan!uunet!olivea!tymix!cirrusl!sunstorm!dhesi
From: dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi)
Newsgroups: comp.editors
Subject: Re: spelling from within vi
Message-ID: <2732@cirrusl.UUCP>
Date: 21 Nov 90 03:58:46 GMT
References: <1990Oct31.153710.7865@cec1.wustl.edu> <1990Nov7.005405.461@investor.pgh.pa.us> <59025@microsoft.UUCP> <21197@ttidca.TTI.COM>
Sender: news@cirrusl.UUCP
Distribution: na
Organization: Cirrus Logic Inc.
Lines: 31

Here's my spell filter for vi.  It's useful for LaTeX files, but works
for everything.  The stop list contains words one per line, enclosed in
backets.  E.g.:

[Emacs]
[Email]
[Feb]
[Fri]

#! /bin/csh -f
# filter text through this to see spelling errors summarized in
# a compact format;  make changes to avoid misinterpretation of
# Latex features; and use a stop list for final filtering.
# The stop list must be sorted in alphabetic order and is
# case-sensitive.   R.D. Wed Apr 18 21:17:43 PDT 1990

cat $* | 		# use stdin or list of files		\
 sed -e 's/\\/ /g' 	# replace backslash by blank		\
 	-e 's/^%//' 	# delete % at beginning of line		\
	-e 's/\$/ /g' 	# replace $ by blank			\
	-e 's/{/ /g' 	# replace { by blank			\
	-e 's/}/ /g' 	# replace } by blank			\
	| \
 spell | \
 sed -e 's/^/[/' -e 's/$/]/' |  # bracket so complete lines match \
 fgrep -v -f ~dhesi/.stoplist | # print only words not in stoplist \
 sed -e 's/^\[//' -e 's/\]$//'  |  # remove brackets		   \
 fmt				# ... and display in compact format
--
Rahul Dhesi <dhesi%cirrusl@oliveb.ATC.olivetti.com>
UUCP:  oliveb!cirrusl!dhesi