Document Management Systems On Linux...

[Lee Harris_5]

Hi,

Not sure if this is the correct forum to post this thread in, but here goes...

As I'm sure you all know...working in IT you tend to collect millions of documents of various kinds. We currently have a network drive with thousands of docs on relating to support of different systems and so on.

The issue we are facing is probably something some of you must have faced at some time. We have that much documentation, being added and updated by so many different people and teams that it becomes practically impossible to find what you're after.

What I want to do, is get all the documentation into one place, preferably a central web server. However this is my challenge, and I wondered if any of you had faced a similar task and would like to share the general gist of your solution...

I want all the documentation to be indexed on the webserver (regardless of format) so that it is fully searchable through the web interface. So basically this allows people to still write their docs in Microsoft Word, or whatever application they are used to, but when they uploaded to the webserver, something indexs all the content of the files and allows you to do free text queries against it.

The other requirement of this solution is that it must be free (GPL / opensource) and run on Linux (RedHat Enterprise Linux 3 AS to be precise).

I was looking at a product called knowledgeTree (kt-dms.sourceforge.net) but due to an issue with MySQL / PHP and RHEL3 this looks like it is out of the window.

Any suggestions?

[Chakravarthi]

check twiki

http://twiki.org/

[Wim Van den Wyngaert]

Managing documents requires structure and discipline. May be a better solution is to use google for indexing the mess you have.

http://www.google.com/appliance/faq.html

Wim

[Lee Harris_5]

Thanks for the replies so far...

As far as I can see though, TWiki does not provide the ability to index existing documents which are in Word format etc.

Also, the Google Appliance, as nice as it would be, is a black box solution offered by Google, but it fails to fulfil one of my major requirements....it must be a free solution.

[Wim Van den Wyngaert]

http://www.htdig.org/

is free. I tried it on OpenVMS but it was only suitable for small indexing jobs. May be try it on Linux ?

Wim

[Wim Van den Wyngaert]

To be more complete : I wanted to index several Gigabyte of very small HTML files. It ran for a week and delivered something corrupt. Too many data ? VMS bug ?

I retried it with only a a few hundred MB and then it worked fine.

Wim

[Dave Falloon]

I home built a quick perl script that used Win32:Ole ( had to run on a windows machine that had office installed ) to read in the first line ( all our documents had descriptive titles ) and to drop that line into a text file, that had the title without common words like and or the, and the path to the file.

Here's a link with some helpful code:

http://www.wellho.net/solutions/perl-using-perl-to-read-microsoft-word-documents.html

You may be able to do something similar and then make a web frontend to search your index file. The perl script ran for a REALLY long time ( all night and some of the morning ) on about 5 gigs of documents.

--Dave

[Davorin Bengez]

Hi,

a little late, but how about mnogosearch?

http://mnogosearch.org

it can index local files as well, and there is support for external parsers - if there is program that can read source document from file|stdin and convert it to file|stdout, it should work. Also, many backend databases, international character support, etc...

hope it helps

regards,

davor

[Lee Harris_5]

Thanks for the replies, but a different solution has been found (one that cost enormous amounts of money I imagine!)