concept count in category solr

This is an excerpt from Manning's book Solr in Action.
Before we dive into the mechanics of implementing facets in Solr, it’s also important to note that facets are calculated in real time for each set of search results (or possibly pulled from a cache for a repeated search). Each facet brings back a list of values (such as Fast Food, Sit-down Chain, Local Diner, and Upscale for the restaurant type facet) along with a count for each of those values indicating how many documents contain that value. The count is not how many times the value exists across all documents (as a value may exist multiple times within a single document); it’s only a count of the number of documents matched. The fact that all facet values are calculated based upon a search result set means that every time a new search is fired, different values and counts will be returned for each facet. This allows users to run a search, see the facets that are returned from the search result, and perform another search that filters on any values for which they want to limit the next set of search results.
This example demonstrates the most basic form of faceting in Solr: field faceting on a single-valued field. In this kind of faceting, each unique value is examined, along with a count of documents in which that value is found. Because you only have one value per document, the sum of all of the counted values (company names in this case) will also equal the total number of documents.