From 579f4e92e6331d8b23a8ef46c58b1f3667b83f02 Mon Sep 17 00:00:00 2001
From: Disassembler <disassembler@dasm.cz>
Date: Sat, 7 Oct 2017 19:26:04 +0200
Subject: [PATCH] Add SeedDMS support for office document indexing

---
 10-seeddms.sh                               |  2 +-
 seeddms/etc/php/7.0/fpm/pool.d/seeddms.conf |  2 ++
 seeddms/srv/seeddms/www/conf/settings.xml   | 10 ++++++++--
 3 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/10-seeddms.sh b/10-seeddms.sh
index 19a0b24..5c6fedf 100755
--- a/10-seeddms.sh
+++ b/10-seeddms.sh
@@ -3,7 +3,7 @@
 SOURCE_DIR=$(realpath $(dirname "${0}"))
 
 # Install dependencies for SeedDMS
-apt-get -y --no-install-recommends install catdoc ghostscript gnumeric id3 imagemagick php7.0-gd php7.0-fpm php7.0-mbstring php7.0-pgsql php7.0-xml php-log php-mail php-pear poppler-utils
+apt-get -y --no-install-recommends install ghostscript id3 imagemagick libreoffice-calc libreoffice-writer php7.0-gd php7.0-fpm php7.0-mbstring php7.0-pgsql php7.0-xml php-log php-mail php-pear poppler-utils unoconv
 
 # Install SeedDMS
 wget https://sourceforge.net/projects/seeddms/files/seeddms-5.1.3/seeddms-quickstart-5.1.3.tar.gz/download -O /srv/seeddms.tgz
diff --git a/seeddms/etc/php/7.0/fpm/pool.d/seeddms.conf b/seeddms/etc/php/7.0/fpm/pool.d/seeddms.conf
index 45fcb40..24bb4e8 100644
--- a/seeddms/etc/php/7.0/fpm/pool.d/seeddms.conf
+++ b/seeddms/etc/php/7.0/fpm/pool.d/seeddms.conf
@@ -11,3 +11,5 @@ pm.max_children = 8
 
 php_admin_value[open_basedir] = /srv/seeddms:/tmp:/tmp:/usr/share/php
 php_admin_value[upload_max_filesize] = 100M
+
+env[LANG]=en_US.UTF-8
diff --git a/seeddms/srv/seeddms/www/conf/settings.xml b/seeddms/srv/seeddms/www/conf/settings.xml
index 66c4e15..3b38b19 100644
--- a/seeddms/srv/seeddms/www/conf/settings.xml
+++ b/seeddms/srv/seeddms/www/conf/settings.xml
@@ -24,10 +24,16 @@
         <server coreDir="" luceneClassDir="" contentOffsetDir="1048576" maxDirID="0" updateNotifyTime="86400" extraPath="/srv/seeddms/pear" cmdTimeout="5" maxExecutionTime="30"/>
         <converters target="fulltext">
             <converter mimeType="application/pdf">pdftotext -enc UTF-8 -nopgbrk %s - | sed -e 's/ [a-zA-Z0-9.]\{1\} / /g' -e 's/[0-9.]//g'</converter>
-            <converter mimeType="application/msword">catdoc %s</converter>
-            <converter mimeType="application/vnd.ms-excel">ssconvert -T Gnumeric_stf:stf_csv -S %s fd://1</converter>
+            <converter mimeType="text/rtf">unoconv -d document -f txt --stdout %s</converter>
+            <converter mimeType="application/msword">unoconv -d document -f txt --stdout %s</converter>
+            <converter mimeType="application/vnd.oasis.opendocument.text">unoconv -d document -f txt --stdout %s</converter>
+            <converter mimeType="application/vnd.openxmlformats-officedocument.wordprocessingml.document">unoconv -d document -f txt --stdout %s</converter>
+            <converter mimeType="application/vnd.ms-excel">unoconv -d spreadsheet -f csv --stdout %s</converter>
+            <converter mimeType="application/vnd.oasis.opendocument.spreadsheet">unoconv -d spreadsheet -f csv --stdout %s</converter>
+            <converter mimeType="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet">unoconv -d spreadsheet -f csv --stdout %s</converter>
             <converter mimeType="audio/mp3">id3 -l -R %s | egrep '(Title|Artist|Album)' | sed 's/^[^:]*: //g'</converter>
             <converter mimeType="audio/mpeg">id3 -l -R %s | egrep '(Title|Artist|Album)' | sed 's/^[^:]*: //g'</converter>
+            <converter mimeType="text/html">unoconv -d document -f txt --stdout %s</converter>
             <converter mimeType="text/plain">cat %s</converter>
         </converters>
     </advanced>