recoll/src/filters/rcldoc
2005-10-20 15:42:29 +00:00

101 lines
2.2 KiB
Bash
Executable File

#!/bin/sh
# @(#$Id: rcldoc,v 1.4 2005-10-20 15:42:29 dockes Exp $ (C) 2004 J.F.Dockes
# Parts taken from Estraier:
#================================================================
# Estraier: a personal full-text search system
# Copyright (C) 2003-2004 Mikio Hirabayashi
#================================================================
#================================================================
# rcldoc
# Extract text from an msword file by executing either antiword
# (or wvware maybe if we need it one day)
#
# The default is to use antiword, the code would need modifications to
# work with wvWare
#
#================================================================
# set variables
LANG=C ; export LANG
LC_ALL=C ; export LC_ALL
progname="rcldoc"
decoder="antiword -i -1 -m UTF-8"
# Not ready to use this for now (it outputs html, so the code below has to
# be simplified.)
#decoder="wvWare -1 -c UTF-8"
# show help message
if test $# -ne 1 -o "$1" = "--help"
then
printf 'Convert a word file to unformatted HTML text.\n'
printf 'Usage: %s [infile]\n' "$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 $cmd not found 1>&2
exit 1
fi
done
}
checkcmds awk antiword iconv
# check the input file existence
if test ! -f "$infile"
then
printf '%s: %s: no such file\n' "$progname" "$infile"
exit 1
fi
# output the result
$decoder "$infile" |
awk '
BEGIN {
printf("<html><head><title></title>\n")
printf("<meta http-equiv=\"Content-Type\" content=\"text/html;charset=UTF-8\">\n")
printf("</head>\n<body><p>");
esc = 1
}
{
if ($0 ~ /-$/) {
sub(/-$/, "", $0)
printf("%s", $0);
} else if($0 == "\f") {
printf("</p>\n<hr>\n<p>")
} else {
if(esc > 0) {
gsub(/&/, "\\&amp;", $0)
gsub(/</, "\\&lt;", $0)
gsub(/>/, "\\&gt;", $0)
}
print $0
}
}
END {
printf("</p></body></html>\n");
}' | iconv -f UTF-8 -t UTF-8 -c -s
# exit normally
exit 0