FCI and the Content Classifier

Classifying files by the data they contain

One of the ways you can create a rule to classify a file’s properties using the FCI (File Classification Infrastructure) in Server 2008 is to examine its contents, using the “Content Classifier.” You can perform a string search or look for data in a particular pattern. You can use a “regular expression” (or more than one) to identify files for classification.

That’s all well and good, but is there any limit on the file types that Server 2008 can parse in order to examine their content and execute the rules that you define in the File Server Resource Manager? Turns out that there is, and it’s a restriction you may already know about if you’ve used Microsoft’s search technology: the “iFilter.”

“iFilters” have been around for a few years for performing file content searches in Windows Vista and Server 2008. They are basically program modules that have an understanding of how data is stored internally in a given file type. So the short answer is that if you can search a file for content, you can probably also classify that file type using FCI, and the converse is also true: if a content search fails, you don’t have an iFilter that can handle classifying a file of that type. That doesn’t mean you should give up hope, however; check http://ifilter.org to see if you can find the iFilter you need. But do some testing first, because Microsoft advises that just because an iFilter exists doesn’t guarantee that it will work with FCI.

• Microsoft
• fci
• file classification infrastructure
• ifilter
• server 2008 r2