CREATE SCHEMA schema_identifier 
[EXTENDS parent_schema_name1 [, parent_schema_name2 [, ...]]]

Substitutable Field


A unique name for the schema.


The identifier of a previously-defined named schema. Imports all of the named schema's fields. No parentheses are required, and no other fields are permitted. Named schemas must be defined in dependency order. If a schema is used before it is defined, an error results.


A schema definition, delimited by parentheses, in the following format:

(field_definition [, ...])

A field definition takes the form:

field_identifier field_type


A unique name for a field in the schema associated with the stream. For table indexes, if the field references a named schema, the entire schema is used as the key.


One of the supported StreamBase data types.


A schema_identifier for a schema whose definition is incorporated into the schema being defined.


The following statements create named schemas. The first example presents a schema definition and names it Schema1. The second example defines its schema by referencing the first schema by name.

CREATE SCHEMA Schema1 (ID int, Symbol string, Price double);
CREATE SCHEMA Schema2 Schema1;

In the following sequence, we define a named schema, point, then a second schema, point3d, that extends point and adds the third dimension:

CREATE SCHEMA point (x double, y double);
CREATE SCHEMA point3d (z double) EXTENDS point;


Use CREATE SCHEMA to create a named schema. Unlike private (anonymous) schemas defined for a particular stream or table, a named schema is defined at the module level, and can be referenced by any component in the module that takes a schema, or in any module that imports the defining module.

For example:

CREATE SCHEMA point (x double, y double);                    [1]
CREATE SCHEMA nested (tag string, nums (a int, y double));   [2] 
CREATE INPUT STREAM in (comment string, p point);            [3]
CREATE OUTPUT STREAM out;                    
SELECT point(1,2) AS mixed FROM in INTO out;                 [4]

The following notes refer to the line numbers in brackets in the example above.

  1. This named schema contains two double fields.

  2. This named schema contains a string field (tag) and a nested tuple field (nums).

  3. This input stream includes the named schema, point, by reference. It is equivalent to:

    CREATE INPUT STREAM instr (comment string, (x double, y double) );
  4. This statement demonstrates another way to use a named schema: its identifier can be used to invoke the automatically generated named schema constructor function to create a tuple with that schema. Here, the point() function creates a tuple with the values 1 and 2. In this case, the int arguments are coerced to doubles, conforming to the named schema. See named schema constructor function for further details.

    You can also use the name of a named schema as a function with zero arguments, which creates a null tuple with the specified schema (not a tuple with all null values).

See Using Parent Schemas for more on that subject.