IDBSchemaRowset

IDBSchemaRowset is an optional interface on sessions. It is used to provide advanced schema information.

Consumers can get information about a data store without knowing its structure by using the IDBSchemaRowset methods. For example, the data store might be one of the following types:

• A Microsoft? SQL Server database that organizes each database into a set of schemas that contain the tables and queries for each schema.

• A Microsoft Access 2.0 database that has a container of tables and a container of queries.

• A Microsoft Access for Windows? 95 database that enables users to define folders to group tables into an arbitrary hierarchy.

The schemas of each of the preceding data stores can be exposed to the consumer by using the methods of IDBSchemaRowset.

Schema rowsets in OLE DB are identified by GUIDs and have predefined restriction columns. The restrictions allow the consumer to limit the set of rows returned by the schema rowset but does not define the format for those restrictions.

IDBSchemaRowset::GetSchemas returns to the consumer a list of schema rowsets, identified by their GUIDs, that are accessible by IDBSchemaRowset::GetRowset. Also returned is a pointer to descriptions of the restriction columns supported for that schema rowset.

IDBSchemaRowset::GetRowset allows the consumer to retrieve the specified schema rowset by passing in the GUID identifying the rowset and the count of restriction columns or an array of restriction values. Each schema rowset will return all of the columns in the order indicated by that rowset's respective table in Appendix B: Schema Rowsets.

Specifying restriction columns allows the consumer to limit the returned rowset to only those rows in the data store whose columns and values match the values specified in rgRestrictions. These values are applied in the same order as the restriction columns.

Programming considerations

At a minimum, providers that implement IDBSchemaRowset must support the passing of the restrictions according to the VARIANT types indicated in the following table. Providers are allowed to support restrictions passed as other types but must convert as dictated by the OLE DB specification.

All providers must support the TABLES, COLUMNS, and PROVIDER_TYPES schema rowsets, identified by the following GUIDs: DBSCHEMA_TABLES, DBSCHEMA_COLUMNS, and DBSCHEMA_PROVIDER_TYPES.

Providers can return provider-specific columns after the last column defined by OLE DB. Other OLE DB models, such as OLE DB for OLAP, can define their own schema rowsets to meet specific needs.

Provider writers should pay particular attention to rowset construction and data retrieval performance, because these methods will be used frequently by consumers.

Although the OLE DB schema rowsets define a minimum content of the system schema, consumers can also look at the schema tables in the database and thus discover other schema tables and attributes beyond the minimum set.

It is suggested that consumers precompile and store the specialized queries used for processes such as command planning and execution-time plan validation. Storage engines may have specialized optimizations associated with those plans.

Providers must return all of the schema rowset columns. If they cannot return the information in a column, they must return an appropriate value, usually NULL. Providers can return provider-specific columns after the last column defined by OLE DB.

For more information about schema rowsets, see Appendix B: Schema Rowsets.

This topic is a part of:

• Microsoft OLE DB