Indexing XLSX Excel 2007 Documents

Install Prerequisite Perl packages

cpan Spreadsheet::XLSX

Create Filter script

We will create two scripts to convert an xlsx file to plain text.  The first is a perl script (xlsx2txt.pl) which takes a filename on the command line and converts it to plain text.  The second is a bash wrapper script that reads from STDIN and creates a temporary file then invokes xlsx2txt.pl.

#!/usr/bin/perl
use Text::Iconv;
my $converter = Text::Iconv -> new ("utf-8", "windows-1251");

# Text::Iconv is not really required.
# This can be any object with the convert method. Or nothing.

use Spreadsheet::XLSX;

my $file = $ARGV[0];
if(!$file) {
        print("Usage: xlsx2txt filename.xlsx\n");
        exit(1);
}
if(! -e $file) {
        printf("File: %s not found\n", $file);
        exit(1);
}

my $excel = Spreadsheet::XLSX -> new ($file, $converter);
foreach my $sheet (@{$excel -> {Worksheet}}) {
        $sheet -> {MaxRow} ||= $sheet -> {MinRow};
        foreach my $row ($sheet -> {MinRow} .. $sheet -> {MaxRow}) {
                $sheet -> {MaxCol} ||= $sheet -> {MinCol};
                foreach my $col ($sheet -> {MinCol} ..  $sheet -> {MaxCol}) {
                        my $cell = $sheet -> {Cells} [$row] [$col];
                        if ($cell) {
                            printf("%s\n", $cell -> {Val});
                        }

                }

        }
}

#!/bin/sh

# Setting PATH
PATH=$PATH:/bin:/usr/bin:/sbin:/usr/sbin:/usr/local/bin:/usr/local/sbin
export PATH

TEMP=`mktemp`
dd of=$TEMP > /dev/null 2>&1
xlsx2txt.pl $TEMP 2> /dev/null
rm $TEM

Configure MindTouch startup.xml

Copy the xlsx2txt.pl script to somewhere in your $PATH (ex: /usr/bin/xlsx2txt.pl)

Copy the xlsx2txt bash script to /var/www/dekiwiki/bin/filters

Make both scripts executable

Edit your /etc/dekiwiki/mindtouch.deki.startup.xml and add the following filter:

 <filter-path extension="xlsx">/var/www/dekiwiki/bin/filters/xlsx2txt</filter-path>