:
##########################################################################
# Shellscript:	striphtml - remove HTML tags from input
# Version    :	$Revision: 1.1.1.1 $
# Author     :	Heiner Steven (stv)
# Category   :  HTML, File Conversion
# Date       :	09.03.1999
# RCS-Id.    :	$Id: striphtml.txt,v 1.1.1.1 1999/06/15 19:29:16 heiner Exp $
##########################################################################
# Description
#
##########################################################################

PN=`basename "$0"`			# Program name
VER=`echo '$Revision: 1.1.1.1 $' | cut -d\  -f2`

Usage () {
    echo >&2 "$PN - strip HTML tags from input, $VER (stv '99)
usage: $PN [file ...]"
    exit 1
}

Msg () {
    for MsgLine
    do echo "$PN: $MsgLine" >&2
    done
}

Fatal () { Msg "$@"; exit 1; }

set -- `getopt h "$@"`
[ $# -lt 1 ] && Usage			# "getopt" detected an error

while [ $# -gt 0 ]
do
    case "$1" in
					# your flags here
	--)	shift; break;;
	-h)	Usage;;
	-*)	Usage;;
	*)	break;;			# First file name
    esac
    shift
done

# Transform the input the following way:
#    1. Replace "<BR>" tags with a newline (special handling)
#    2. Remove all Tags delimited by "<..>"
#    3. Remove character entities (i.e. "&copy;" or "&nbsp;")
sed -e 's:<[bB][rR]>:\
:g' \
    -e 's:<[^>]*>::g' \
    -e 's:&..;::g; s:&...;::g; s:&....;::g'	\
    "$@"