Understanding Full-Text Searching

Full-text searching is the performing of sophisticated searches on blocks of text. Unlike database indexing, which usually sorts data alphabetically, full-text searching accesses its indexed data in many different wayssubstring searches, proximity searches, relevancy searches, and more.

To facilitate full-text searching, ColdFusion includes with a custom implementation of the Verity K2 Server. This integrated Java based engine allows developers to perform full-text searches on database queries and file libraries. Verity generates a read-optimized set of indexes called a collection, and it provides search tools to retrieve result sets ranked by relevancy. Verity runs as a separate service or daemon that must be installed and running to be used.

NOTE

Verity, Inc., offers a range of enterprise-level search and indexing products. The ColdFusion Verity implementation is a customized version of one of its core products. See http://www.verity.com for more details.

A Verity collection is a group of files and associated metadata that is optimized for searching. Collections include various word indexes, an internal- documents table that contains document field information, and pointers to the actual files for file and path indexes. ColdFusion provides a number of Verity functions with which to create, maintain, and optimize collections. This maintenance can be performed through ColdFusion Administrator or with the <cfcollection> and <cfindex> tags.

Searching a Verity collection is fast and leverages the search-command vocabulary of the Verity engine. Users can perform sophisticated searches by using <cfsearch>, which handles boolean operators, wildcards, and other advanced search options.

NOTE

Verity can index and search most popular office file types, including text, HTML, XML, RTF, Adobe PDF, Microsoft Word, WordPerfect, Excel, and PowerPoint.