176 lines
4.1 KiB
Bash
Executable File
176 lines
4.1 KiB
Bash
Executable File
#!/bin/sh
|
|
# @(#$Id: rclabw,v 1.2 2007-06-15 11:41:50 dockes Exp $ (C) 2004 J.F.Dockes
|
|
# Parts taken from Estraier:
|
|
#================================================================
|
|
# Estraier: a personal full-text search system
|
|
# Copyright (C) 2003-2004 Mikio Hirabayashi
|
|
#================================================================
|
|
#================================================================
|
|
# Extract text from an abiword file
|
|
#================================================================
|
|
|
|
# set variables
|
|
LANG=C ; export LANG
|
|
LC_ALL=C ; export LC_ALL
|
|
progname="rclabw"
|
|
filetype=abiword
|
|
|
|
#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 $cmd ; then return 0; else return 1; fi ;;
|
|
*)
|
|
oldifs=$IFS; IFS=":"; set -- $PATH; IFS=$oldifs
|
|
for d in $*;do test -x $d/$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 iconv sed
|
|
|
|
# check the input file existence
|
|
if test ! -f "$infile"
|
|
then
|
|
printf '%s: %s: no such file\n' "$progname" "$infile"
|
|
exit 1
|
|
fi
|
|
|
|
encoding=`sed -e '/<?xml version=/s/"?>$//' \
|
|
-e '/^<?xml version=/s/.*encoding="//p;D;q' \
|
|
-e D \
|
|
< $infile`
|
|
if test X$encoding = X ; then encoding=UTF-8;fi
|
|
|
|
# Note: there can be newlines inside the description field, we don't want
|
|
# them... Have 2 use 2 different selectors for the single-line and
|
|
# multiple-line cases because of the generic tag end (</m> for all meta
|
|
# tags)
|
|
descsedprog='
|
|
/<m key="dc.description">\([^<]*\)<\/m>/ {
|
|
s//\1/
|
|
p
|
|
q
|
|
}
|
|
/<m key="dc.description">/,/<\/m>/ {
|
|
s!.*<m key="dc.description">!!
|
|
s!</m>.*!!
|
|
H
|
|
}
|
|
${
|
|
g
|
|
s/\n/ /g
|
|
p
|
|
}
|
|
'
|
|
|
|
description=`sed -n -e "$descsedprog" < "$infile"`
|
|
#echo description: "$description"
|
|
|
|
# Set program for the single line meta elements. Takes element name as
|
|
# parameter
|
|
setmetasedprog() {
|
|
metasedprog='/<m key="'$1'">/{
|
|
s/.*<m key="'$1'">\([^<]*\).*/\1/
|
|
'"s/\"/'/g"'
|
|
p
|
|
}'
|
|
}
|
|
|
|
setmetasedprog dc.subject
|
|
subject=`sed -n -e "$metasedprog" "$infile"`
|
|
#echo subject: "$subject"
|
|
|
|
setmetasedprog dc.title
|
|
title=`sed -n -e "$metasedprog" "$infile"`
|
|
#echo titre: "$title"
|
|
|
|
setmetasedprog abiword.keywords
|
|
keywords=`sed -n -e "$metasedprog" "$infile"`
|
|
#echo keywords: "$keywords"
|
|
|
|
setmetasedprog dc.creator
|
|
creator=`sed -n -e "$metasedprog" "$infile"`
|
|
#echo creator: "$creator"
|
|
|
|
# Note: next expr supposes that paragraphs are always all by themselves on
|
|
# a single line in the xml (no multiple <p> per line, no embedded newlines
|
|
# in text).
|
|
contentsedprog='
|
|
/<p[ >]/{
|
|
s/<[^>]*>/ /g
|
|
p
|
|
}
|
|
'
|
|
content=`sed -n -e "$contentsedprog" "$infile"`
|
|
#echo content: "$content"
|
|
|
|
# output the result
|
|
(echo '<html><head><title>' "$title" '</title>'
|
|
echo '<meta http-equiv="Content-Type" content="text/html;charset=UTF-8">'
|
|
echo '<meta name="description" content="' "$description $subject" '">'
|
|
echo '<meta name="keywords" content="' "$keywords" '">'
|
|
echo '<meta name="author" content="' "$creator" '">'
|
|
echo '</head><body><pre>'
|
|
echo "$content"
|
|
echo '</pre></body></html>') \
|
|
| iconv -f $encoding -t UTF-8 -c -s
|
|
|
|
|
|
# exit normally
|
|
exit 0
|