ADD_MDATA_COLUMN

Use this procedure to map the FILTER BY column named in column_name to the MDATA section named in section_name.

Syntax

The syntax is as follows:

CTX_DDL.ADD_MDATA_COLUMN(
     group_name         IN VARCHAR2,
     section_name       IN VARCHAR2,
     column_name        IN VARCHAR2,
);

group_name

Name of the group that contains the section.

section_name

Name of the MDATA section.

column_name

Name of the FILTER BY column to add to the MDATA section.

Note: The column_name must not be prefixed by the owner, schema or table name.

Restrictions

MDATA sections that are created with CTX_DDL.ADD_MDATA_COLUMN cannot have their values changed using CTX_DDL.ADD_MDATA or CTX_DDL.REMOVE_MDATA. Doing so will result in errors being returned. The section values must be updated using SQL.

Notes

• The stored datatype for MDATA sections is text. Therefore, the value of the FILTER BY column is converted to text during indexing. For non-text datatypes, the FILTER BY columns are normalized to an internal format during indexing. If the section is queried with an MDATA operator, then the MDATA query string will also be normalized to the internal format before processing.

• When a FILTER BY column is mapped as MDATA, the cost-based optimizer in Oracle Text tries to avoid using the Oracle Text composite domain index to process range predicate(s) on that FILTER BY column. This is because range predicates on MDATA FILTER BY columns are processed less efficiently than if they were declared as SDATA. For this reason, you should not add a FILTER BY column as MDATA if you plan to do range searches on the column.

Related Topics

“MDATA”

“ADD_MDATA_SECTION”

“REMOVE_MDATA”

“ADD_SDATA_COLUMN”

See Also: Chapter 8, “Searching Document Sections in Oracle Text” in Oracle Text Application Developer’s Guide