Table types¶

A table type describes tabular data composed of named fields and rows. Each field has a value type. For example, TABLE {id INT64, name STRING} describes a table with an id field of type INT64 and a name field of type STRING.

Table types cannot be used as node or edge property types.

View a table type¶

Use the typeof() function to view the type of a table variable:

'''
TABLE t TYPED TABLE {id INT64, name STRING} =
    {id: 10001, name: 'Alice'},
    {id: 10002, name: 'Bob'}
RETURN typeof(t) AS t_type
'''

Return:

+-------------------------------+
| t_type                        |
+-------------------------------+
| TABLE {id INT64, name STRING} |
+-------------------------------+

The returned type contains the name and value type of each field.

Table types and table variables¶

A table type describes the structure of tabular data, whereas a table variable stores or refers to the data:

TABLE t TYPED TABLE {id INT64, name STRING}

In this declaration:

t is the table variable.
TABLE {id INT64, name STRING} is its table type.
The first TABLE keyword introduces the table variable declaration.

For the complete declaration syntax, see Table variables.

Use a table type as a procedure parameter¶

A user-defined procedure (UDP) can accept a table through a parameter declared with TABLE:

parameter_name TABLE

In this syntax, TABLE accepts a table variable passed by the caller. For more information, see CREATE PROCEDURE.

Usage¶

Table types are mainly used for:

Table variables that store intermediate data or results.
UDP parameters that receive tables.
Data exchanged through IMPORT and EXPORT.

Table variables in NebulaGraph Database Developer Edition¶

In NebulaGraph Database Developer Edition, the field names and types of a table variable can also be inferred from an initializer or query result. Local table variables can be iterated with FOR.

Limitations¶

Table types cannot be used as node or edge property types.
The values written to a table must be compatible with its field types.