Crawled Properties not accessible from content application in SharePoint 2010

As we are in the process of porting our FC.ImageSearch product to SharePoint 2010 we came across a couple of Properties that are inaccessible if the code is running in the context of a Content Application, e.g. the AllCategories property of the schema object (more details on that here).
 
To make a long story short: Microsoft has opted to block access to certain data in the object model in order to prevent administration work to be done from a content application (instead of utilizing the central administration web application).
 
While I agree that security is always a concern I think that the current implementation is too tight.
 
One Example: In our scenario we want to give the site administrator the information which columns of the site's lists are mapped by which Managed Property's, so that he/she then knows how to configure the Advanced Search page properties accordingly. However, that's not possible as the CrawledProperty's - the "link" between columns and ManagedProperty's - are inaccessible from the content application.
 
Again, I consider it right to prevent any modifications to be made to the search' objects from a content application, but why cut off read access?
 
To solve this we implemented a custom WCF service on the central administration web application which the content application's code can talk to. While that's not bad or difficult it still doesn't feel right.
 
What do you think?

Published: Sep-20-10 | 0 Comments | Link to this post

Leverage the FREE Windows TIFF iFilter for OCR in SharePoint

I knew about the TIFF iFilter available in Windows 2008R2 and its OCR capabilities, but I had not had the time to try it out until very recently when it became a priority as a customer requested it.
 
The setup for it is very SIMPLE and well described in John Liu's blog post. It consists of no more than two essential steps: activating the TIFF iFilter Windows Feature and to configure the OCR Group Policy properties.
 
I might want to add a couple of things to John's description though:
 
1) You should also configure the option "Select OCR languages from a code page".
 
2) You do NOT need to restart your server: simply restart the SharePoint Search services and then execute an iisreset.
 
3) You only need to start a "Full Crawl" if you want to reindex TIFF files that are already in SharePoint. All TIFF files uploaded AFTER the iisreset will be indexed automatically.
 
Of course you need to consider the amount of text being added to the general search index file. However, the upside of searching on all the scanned documents' text is certainly quite obvious.

Published: Sep-20-10 | 0 Comments | Link to this post