Gavin's Sharepoint Blog » ifilter

Adobe labs acrobat v8 ifilter for 64bit sharepoint – prerelease

gavinadams — Fri, 25 Jan 2008 03:05:34 +0000

Just stumbled across this on a Google search.

Adobe Labs have a pre-release of how to support Adobe Reader v8.1 ifilter on 64bit sharepoint.

I’ll admit that I haven’t tested this yet and it involves a 64bit ‘thunking’ layer. In the past I have seen some instructions on to use this technique to ‘hack’ together a custom 32bit ifilter handler onto a 64bit system.

The main significance of this is that currently the only cleanest way of supporting pdf indexing on sharepoint 64bit platforms is to use foxit pdf ifilter. However this is commercial software that is priced based on the number of CPU cores. (As at 25JAN2008 this was “License cost for each server with up to two CPU cores is $329.99. Each additional CPU core costs $129.99.”) Now with most production servers these days being dual quad core processors (ie 8 processing cores) This would be US$1109.93 for each production server. Congrats does go to foxit for seeing a large hole left in the market by Adobe being so slow to support the 64bit platform and providing a solution.

The other comment worth making is that the foxit 64bit ifilter may be coded natively in 64bit but I’m not sure of this. Whereas the Adobe solution is taking the existing 32bit ifilter in wrapping it into 64bit which does involve a processing cost overhead. If you had a large index with lots pdf documents then going for a native pdf ifilter could bring a significant performance difference.

Enabling the inbuilt msg ifilter on sharepoint (even 64bit!)

gavinadams — Tue, 09 Oct 2007 04:58:20 +0000

One of our big decisions on the proposed physical architecture of our sharepoint farms has been can we go 64bit on the index server. For us the ability to index .msg files is crucial and until now I thought I would have to purchase a third party ifilter. The problem being that the third party does not have a 64bit version of their ifilter and initial discussion with the vendor indicated that building a 64bit ifilter would not be easy.

Some recent updates to the technet sharepoint doco were pointed out to me yesterday that listed msg indexing support as being available out of the box.
This is the relevant page:-

File types and IFilter reference (Office SharePoint Server)
URL: http://technet2.microsoft.com/Office/en-us/library/09357d8e-37b9-4e96-b8fd-f17b990d010a1033.mspx

So this doc basically states that msg filters are inbuilt. So I do a search of some content that is in the body of a .msg file and no results are returned.

Now I’ve been down this path before with the foxit pdf ifilters so I’m pretty sure that it is a problem with missing values in the registry. I also note at the bottom of the above reference page is a link to an article on search onenote 2007 files. This document provides hints as to the missing registry keys.

Install and register the OneNote IFilter (Office SharePoint Server 2007)
URL: http://technet2.microsoft.com/Office/en-us/library/2e715e42-c09b-4b4f-a082-b19e1cad96031033.mspx

Crank up regedit check that the :-

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Office Server\12.0\Search\Setup\ContentIndexCommon\Filters\Extension\.msg keys exist.
(They should do, all my installs have this by default)

The following key is the one that is missing.

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Office Server\12.0\Search\Setup\Filters\.msg]
“Extension”=”msg”
“FileTypeBucket”=dword:00000001
“MimeTypes”=”application/msoutlook”

[Edit 08JUL2010 - changed MimeTypes key above as per testing from others. Thanks to devparts comment below]

Now, I have taken a guess on the mimetype for outlook. I can find no mimetypes for .msg files in the registries on my servers and my workstations. A web search yielded 3 possibilities, “application/vnd.ms-outlook”, “application/x-msg” & “application/msoutlook”

So the next thing was to restart the office search service, and iisreset. I think reset all crawled content (purge the index). Then I ran a full crawl.

Fortunately I got lucky on the mimetype and the index successfully crawled the content of the email! Now for the caveats. I’ve only done basic testing. I haven’t checked if the msg ifilter will pull an attachment (eg word doc) and index that as well. Also I haven’t checked this with Microsoft, though I think it is worthy of a bug report (at least in the doco).

Also thanks to Mutaz from MCS for his initial guidance on pointing me in the right direction this.

Configuring Foxit 64bit PDF ifilter

gavinadams — Fri, 20 Jul 2007 00:00:11 +0000

Adir Ron has posted on his blog the instructions for making the Foxit 64bit PDF ifilter work with Sharepoint 2007.

Taking from his article is the following reg hack instructions on registering the ifilter with the sharepoint search.

After you install the Foxit, add a pdf extension in MOSS search settings
Open regedit, locate [HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Office Server\12.0\Search\Setup\ContentIndexCommon\Filters\Extension\.pdf]
Change the default value to {987f8d1a-26e6-4554-b007-6b20e2680632} . (You can get this handler addin value using Citeknet ifilter explorer)
Cmdline: net stop osearch
Cmdline: net start osearch
Do a full crawl.

(PS Adir, love the MOSS club tag line on your blog.)

(PPS, To everyone else Adir Ron has some great articles on sharepoint search on his blog.)