What Is the Landmark Text Search Feature?
The text search feature of the Microsoft SQL Server database can be enabled for Lawson binary document (BLOB) fields, large text (CLOB) fields, and standard alpha fields. When a binary document (such as word processing, RTF, PDF, HTML, or XML documents) is stored, the text search feature parses the document into words that are then included in a text index.
Microsoft SQL Server text search allows you to configure searching features such as a 'stop list' to filter unwanted words, phrase or context matching in order to locate words that occur close to each other within the text, or thesaurus entries to return better results using synonyms. For example, a coffee shop owner could associate java and coffee as synonyms, while a software development vendor could associate java and program language as synonyms.
Text Search Change Management Considerations
All text index administration is done manually by the database
administrator, since the Landmark database interface layer manages
text search tables and indexes (BLOB or CLOB fields) differently than
other data. The database has parameters that control when and how
changes are implemented. You configure these database settings by
setting parameter values in the %LARUNDIR%\productline\db.cfg
file. Detailed descriptions
of text search management parameters can be found in What Are the Parameters in the Lawson Database Interface Configuration File?.
After setting the db.cfg
parameters to control
text search management, you create the indexes using the
Text Index Job Population utility. Implementing Text Indexing and Text Searches
Any time you change the text search parameters in the db.cfg
file, or change the text index design, you must recreate
all text indexes using dbmaint
. This command generates
and executes DDL to implement the parameter values set in the db.cfg
file, and drops and creates only the text indexes in
the database. Recreating Text Indexes in the Database
If the CHANGE_TRACKING parameter
in the db.cfg
file is set to AUTO, you do not need
to run dbmaint
. The index will be repopulated when
a row is changed.
If you are working with critical data, do not use the
update (-U
) mode without first examining the DDL.
Instead, the preferred method is to direct dbmaint
output to a file that you can examine, edit if necessary, and issue
the command to the database through the database SQL utility.
About the Text Index Population Job Utility
The textindexjob
utility generates and executes
the SQL commands that drop or create jobs in Microsoft SQL Server that populate
text indexes. The job defines a schedule for populating text indexes
based on settings the db.cfg
file.
Names for the text index population job, job step and schedule are based on the data area's database, schema, and table names (where TS stands for Text Search):
-
TS_DatabaseName_SchemaName_TableName
-
TS_STEP_DatabaseName_SchemaName_TableName
-
TS_SCHEDULE_DatabaseName_SchemaName_TableName