N-grams (i.e., N character sequences) are used to identify documents that
potentially satisfy a search query. The documents that potentially
satisfy the query are then searched (e.g., using a full text search) to
determine which documents actually satisfy the query.