157 lines
3.5 KiB
Bash
Executable File
157 lines
3.5 KiB
Bash
Executable File
#!/bin/sh
|
||
# @(#$Id: rclsoff,v 1.6.6.1 2007/01/21 16:41:49 dockes Exp $ (C) 2004 J.F.Dockes
|
||
# Parts taken from Estraier:
|
||
#================================================================
|
||
# Estraier: a personal full-text search system
|
||
# Copyright (C) 2003-2004 Mikio Hirabayashi
|
||
#================================================================
|
||
#================================================================
|
||
# rclsoff
|
||
# Extract text from an openoffice/soffice file
|
||
#
|
||
#================================================================
|
||
|
||
|
||
# set variables
|
||
LANG=C ; export LANG
|
||
LC_ALL=C ; export LC_ALL
|
||
progname="rclsoff"
|
||
|
||
|
||
# show help message
|
||
if test $# -ne 1 -o "$1" = "--help"
|
||
then
|
||
printf 'Convert an openoffice file to unformatted HTML text.\n'
|
||
printf 'Usage: %s [infile]\n' "$progname"
|
||
exit 1
|
||
fi
|
||
|
||
infile="$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
|
||
echo $cmd not found 1>&2
|
||
exit 1
|
||
fi
|
||
done
|
||
}
|
||
checkcmds awk iconv unzip
|
||
|
||
# check the input file existence
|
||
if test ! -f "$infile"
|
||
then
|
||
printf '%s: %s: no such file\n' "$progname" "$infile"
|
||
exit 1
|
||
fi
|
||
|
||
# We need a temporary directory
|
||
if test z"$RECOLL_TMPDIR" != z; then
|
||
ttdir=$RECOLL_TMPDIR
|
||
elif test z"$TMPDIR" != z ; then
|
||
ttdir=$TMPDIR
|
||
else
|
||
ttdir=/tmp
|
||
fi
|
||
tmpdir=$ttdir/rclsoff_tmp$$
|
||
mkdir $tmpdir || exit 1
|
||
mkdir $tmpdir/rclsofftmp || exit 1
|
||
|
||
cleanup()
|
||
{
|
||
# Note that we're using a constant part (rclsofftmp), that hopefully
|
||
# guarantees that we can't do big mistakes here.
|
||
rm -rf $tmpdir/rclsofftmp
|
||
rmdir $tmpdir
|
||
}
|
||
|
||
trap cleanup EXIT HUP QUIT INT TERM
|
||
|
||
# Unzip the input file and change to the unzipped directory
|
||
unzip -q -d $tmpdir/rclsofftmp "$infile"
|
||
cd $tmpdir/rclsofftmp
|
||
|
||
# Note: there can be newlines inside the description field, we don't want
|
||
# them...
|
||
descsedprog='/<dc:description>/,/<\/dc:description>/{
|
||
s!.*<dc:description>!!
|
||
s!</dc:description>.*!!
|
||
p
|
||
}
|
||
'
|
||
description=`sed -n -e "$descsedprog" < meta.xml | tr '\n' ' '`
|
||
|
||
subject=`sed -e "s/\"/'/" -e 's/.*<dc:subject>\([^<]*\).*/\1/p;d' < meta.xml`
|
||
|
||
title=`sed -e "s/\"/'/" -e 's/.*<dc:title>\([^<]*\).*/\1/p;d' < meta.xml`
|
||
|
||
keywords=`sed -e "s/\"/'/" -e 's/.*<meta:keyword>\([^<]*\).*/\1/p;d' \
|
||
< meta.xml`
|
||
|
||
# Note: next expr inserts a newline at each end of paragraph (for preview)
|
||
content="`sed -e 's!</text:p>!\\
|
||
!g' -e 's/<[^>]*>/ /g' < content.xml`"
|
||
|
||
#echo description "$description"
|
||
#echo subject "$subject"
|
||
#echo title "$title"
|
||
#echo keywords "$keywords"
|
||
#echo content "$content"
|
||
|
||
# output the result
|
||
echo '<html><head>'
|
||
echo '<title>' "$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 '</head><body><p>'
|
||
|
||
echo "$content" | sed -e "s/'/'/g" -e 's/"/"/g' |\
|
||
awk '
|
||
BEGIN {
|
||
cont = ""
|
||
}
|
||
{
|
||
$0 = cont $0
|
||
cont = ""
|
||
|
||
if ($0 ~ /[-]$/) {
|
||
# Note : soft-hyphen is iso8859 0xad
|
||
# Break at last whitespace
|
||
match($0, "[ \t][^ \t]+$")
|
||
line = substr($0, 0, RSTART)
|
||
cont = substr($0, RSTART, RLENGTH-1)
|
||
$0 = line
|
||
}
|
||
|
||
if($0 == "\f") {
|
||
print "</p>\n<hr>\n<p>"
|
||
next
|
||
}
|
||
|
||
print $0 "<br>"
|
||
}
|
||
END {
|
||
printf("</p></body></html>\n");
|
||
}' | iconv -f UTF-8 -t UTF-8 -c -s
|
||
|
||
cd /
|
||
# exit normally
|
||
exit 0
|