Select the search type
  • Site
  • Web
Search
You are here:  Support/Forums
Support

Bring2mind Forums

Search on contents only works with *.txt documents, not Word or PDF
Last Post 06/22/2012 4:48 PM by Peter Donker. 7 Replies.
Sort:
PrevPrev NextNext
You are not authorized to post a reply.
Author Messages
Kate
New Member
New Member
Posts:29


--
06/04/2012 9:03 PM
We've purchased DocX (version 6.0.3) and installed it on DNN community version 6.01.05. DNN is resident on a hosted server. We set up DocX to use Lucene Search provider.

Contents search works ONLY with .txt documents, NOT with Microsoft Word or PDF files. I had the hosting company install Adobe iFilter on the server and restart IIS. I've re-run all relevant scripts.

We are setting this up for a client who will mainly store upload PDF documents and needs the ability to search contents as well as title and key word. So we have to get content search to work for documents other than plain text, but I really do not know how to proceed from here.
Peter Donker
Veteran Member
Veteran Member
Posts:4536


--
06/05/2012 10:23 PM
Hi Kate,

This means the component can't leverage the server's iFilters. These are dlls installed on the server which allow other software to read proprietary format files. Like Word, Pdf, etc.

This can be due to two things:
1. The server doesn't have the necessary iFilters installed. In this case you'll typically find you can find something in a Word file (the MS Office iFilters are on nearly every Windows server AFAIK) but not in another format like Acrobat. Resolution: install the applicable iFilters.
2. The server doesn't allow the software to use the iFilters. The latter is a "trust issue" and must be discussed with the server admin.

Peter
Kate
New Member
New Member
Posts:29


--
06/11/2012 8:19 PM
[quote]


This can be due to two things:
1. The server doesn't have the necessary iFilters installed. In this case you'll typically find you can find something in a Word file (the MS Office iFilters are on nearly every Windows server AFAIK) but not in another format like Acrobat. Resolution: install the applicable iFilters.
2. The server doesn't allow the software to use the iFilters. The latter is a "trust issue" and must be discussed with the server admin.

Peter
[/quote]

As mentioned in my original post, I had the hosting company install Adobe iFilter. I checked out the trust level, and it was set to full. I re-ran all scripts that could be remotely applicable. I really don't know where else to go with this, do you have any other suggestions?
Rob Ralston
Basic Member
Basic Member
Posts:164


--
06/11/2012 11:46 PM
Hi Kate,

You may know all this, but just trying to help.

Since there is more than one place that trust level can be set: 1) at the top level web.config file for the server, and 2) at each individual web site, it would be good to look at a specific value within your DNN site, since the top level web.config can have the trust level set to "Medium" with "No Override" allowed. This would prevent your local web.config from having any affect (on Trust level)

To really see what trust level you have on the site, go to Host Settings, and look for the value of "Permissions". If it says "None", then the site is in a lower trust level, most likely Medium. If it is in Full Trust, the value of Permissions will be something else (I think it will say "Reflection").

Many hosting companies force Medium Trust as a security best practice for IIS.

I hope this is of some help.

Rob Ralston
SilverBullet Technologies LLC
Peter Donker
Veteran Member
Veteran Member
Posts:4536


--
06/18/2012 1:12 PM
Thanks Rob. And for completeness' sake:

1. The permission needed is the permission by the asp.net application to "run unmanaged code"
2. If the iFilter is outdated newer PDFs won't be read by the old iFilter.

Peter
Kate
New Member
New Member
Posts:29


--
06/18/2012 11:42 PM
Thanks to both of you, Rob and Peter, for answering me.

Meanwhile, however, I see that Permissions has the following value(s):

ReflectionPermission, WebPermission, AspNetHostingPermission

Does the above include "run unmanaged code" or does something else needed to be added to Permissions?

Kate
Rob Ralston
Basic Member
Basic Member
Posts:164


--
06/19/2012 8:20 PM
Kate,

From my experience, your site is indeed running under .NET Full Trust, which allows for "unmanaged code" execution, so theoretically, Lucene should work.

To further isolate, you may want to check the DNN Event log for any related errors, as well as the Windows Event logs on the server, assuming you have that kind of server access. I would re-run the indexing scripts, try a search, then look for errors/warnings.

If the application pool account for your site is a unique Windows account, it is possible the account was not properly registered with IIS.

I have not used Lucene (because of the .NET Trust issues), but I believe Peter has a recommended tool for testing Lucene indexing, which again I believe requires you to be able to log on to the server and install/run the query tool. On the surface, it would seem the Lucene index file is not being populated.

Also, not sure if your version of DNN is related to the problem, but Peter would probably have identified that already.

Ultimately, you may need to get your hosting provider's tech support involved to determine the problem. Hopefully they have experience with DNN hosting.

Rob Ralston
Peter Donker
Veteran Member
Veteran Member
Posts:4536


--
06/22/2012 4:48 PM
Check out this new thread:
http://www.bring2mind.net...v/topic/Default.aspx

Peter
You are not authorized to post a reply.