Parent:
[8e8ac4]
(diff)
Child:
[097bc5]
(diff)
Download this file
rclsiduxman
94 lines (79 with data), 2.5 kB
#!/bin/sh
# @(#$Id: rclsiduxman,v 1.1 2008-06-09 09:12:05 dockes Exp $ (C) 2004 J.F.Dockes
# Parts taken from Estraier:
#================================================================
# Estraier: a personal full-text search system
# Copyright (C) 2003-2004 Mikio Hirabayashi
#================================================================
#================================================================
# rclsiduxman
# Strip the menu part from sidux manual pages to improve search precision
#================================================================
# set variables
LANG=C ; export LANG
LC_ALL=C ; export LC_ALL
progname="rclsiduxman"
filetype="sidux manual htm"
#RECFILTCOMMONCODE
##############################################################################
# !! Leave the previous line unmodified!! Code imported from the
# recfiltcommon file
# Utility code common to all shell filters. This could be sourced at run
# time, but it's slightly more efficient to include the code in the
# filters at build time (with a sed script).
# Describe error in a way that can be interpreted by our caller
senderror()
{
echo RECFILTERROR $*
# Also alert on stderr just in case
echo ":2:$progname::: $*" 1>&2
exit 1
}
iscmd()
{
cmd=$1
case $cmd in
*/*)
if test -x $cmd -a ! -d $cmd ; then return 0; else return 1; fi ;;
*)
oldifs=$IFS; IFS=":"; set -- $PATH; IFS=$oldifs
for d in $*;do test -x $d/$cmd -a ! -d $d/$cmd && return 0;done
return 1 ;;
esac
}
checkcmds()
{
for cmd in $*;do
if iscmd $cmd
then
a=1
else
senderror HELPERNOTFOUND $cmd
fi
done
}
# show help message
if test $# -ne 1 -o "$1" = "--help"
then
echo "Convert a $filetype file to HTML text for Recoll indexing."
echo "Usage: $progname [infile]"
exit 1
fi
infile="$1"
# check the input file existence (may be '-' for stdin)
if test "X$infile" != X- -a ! -f "$infile"
then
senderror INPUTNOSUCHFILE "$infile"
fi
# protect access to our temp files and directories
umask 77
##############################################################################
# !! Leave the following line unmodified !
#ENDRECFILTCOMMONCODE
checkcmds sed
# Delete everything from <div id="menu"> to <div id="main-page">
# This prints an additional blank line at top which does not matter
sed -n -e '1,/<div id="menu">/{x;p' -e '}' \
-e '/<div id="main-page">/,$p' < $infile
# exit normally
exit 0