Skip to main content

Full-Text index in SQL Server

Full-Text index
Full-Text search supports to full-text queries against to character-based data. These types of queries can include words and phrases as well as multiple forms of a word or phrase. To support full-text queries, full-text indexes must be implemented on the columns referenced in the query. The columns can be configured with character data types (such as char and varchar) or with binary data types (such as varbinary and image).
The full-text index breaks the column into tokens and these tokens make up the index data. Before creating a full-text index you must create the full-text catalog which stored the full-text index data. A full-text catalog can contain many indexes but a full-text index can only be part of one catalog.  Also, only one full-text index can be created on each table. The full-text index is not updated like as it is regular indexes. Populating full-text indexes can be resource intensive so there are more options that let you control when they are updated. 

The 3 different options are:
1.     Full population
2.     Automatic or manual population based on change tracking
3.     Incremental population based on a timestamp

Benefits of Full-Text Index
The clustered and non-clustered indexes are not supported for any of the Large Object (LOB) data-types. Full-text indexes can be created on LOB data-type columns like TEXT, VARCHAR (MAX), IMAGE, VARBINARY (MAX) (it can also index CHAR and VARCHAR column types).
How to create a full-text index?
Before creating a full-text index we need some required setup. First, we create the full-text catalog for their storage. Catalog name must be unique across all databases on the server. Using the following script we can create the full-text catalog

CREATE FULLTEXT CATALOG ft_catalog AS DEFAULT;


After creating the full-text catalog we can the full-text index. There are two options we can add when creating a full-text index
1-        Key Index
2-        Stoplist
For creating full-text index a table should have a unique key and KEY INDEX index_name is the name of the unique key index on table_name. The KEY INDEX must be a unique, single-key, non-nullable column.

Note: Select the smallest unique key index for the full-text unique key. For the best performance, we recommend an integer data type for the full-text key.
Use the following script to create full-text index.

CREATE FULLTEXT INDEX ON Inventory.Item(ItemSummary)
KEY INDEX PK_Item_ItemsNode
WITH STOPLIST = SYSTEM;


Popular posts from this blog

Check for changes to an SQL Server table?

Problem: Suppose your team is working on the under-development project so it might be possible continuous work on the database and perform changes in Table, Stored procedure as per requirement, and daily you have to update the testing server database as per changes are done in developing server database then how it is possible to trace those changes. There are a lot of solutions for this problem which is listed below Solution 1: For SQL Server 2000, 2005 and above use the CHECKSUM command SELECTCHECKSUM_AGG(BINARY_CHECKSUM(*))FROMYour_Table_NameWITH (NOLOCK); That will return the same number each time its run as long as the table contents haven't changed. Unfortunately CHECKSUM does not work always properly to detect changes. It is only a primitive checksum and no CRC calculation. Therefore you can't use it to detect all changes, e. g. symmetrical changes result in the same CHECKSUM! Solution 2: 1.Run the following query. Before executing query replace DB_Name with your database name…

Merge and Merge join transformation in SSIS

How to drop multiple tables with common prefix in one query?

Problem: Suppose we have a situation where we need to drop those tables that have some prefixes string, so it is possible to drop those tables with a common prefix in a single query.
Solution: yes it is possible to drop all those tables that have the common prefix in a single query. Using following query you can delete all those tables that begin with a certain prefix string. In where condition just put the common prefix string in where condition (Like ‘prefix%’)
DECLARE@queryNVARCHAR(MAX)=N'';
SELECT@query+=' DROP TABLE ' +QUOTENAME(s.name) +'.'+QUOTENAME(t.name)+';' FROMsys.tablesASt INNERJOINsys.schemasASs ONt.[schema_id]=s.[schema_id] WHEREt.nameLIKE'MX_100%';
EXECsp_executesql@query;
This query may create an issue, if a table has a foreign key relationship, you'll either need to drop them first or arrange the output to drop the tables in a certain order. If you want to monitor exactly what goes on when the query is running then use the following que…