Table functions¶

The following functions operate on tables.

table_split()¶

Splits a table into multiple sub-tables and returns a list containing the sub-tables. This is useful for parallel processing or batch operations on large datasets.

Null behavior: default

Function signature¶

table_split(t: TABLE) -> LIST

or

table_split(t: TABLE, size: UINT64) -> LIST

t: The variable of the table to be split.
size: The value expression that evaluates to the maximum number of rows per sub-table (the last sub-table may contain fewer rows). If not specified, the system automatically determines split boundaries based on memory distribution. The value of size must be a positive integer.

Examples¶

Split a table and count total rows¶

'''
TABLE result_table TYPED TABLE {i INT, j INT}
FOR i IN range(1, 100)
FOR j IN range(1, 100)
EXPORT i, j INTO result_table

FOR t IN table_split(result_table)
RETURN sum(size(t))
'''

Return:

+--------------+
| sum(size(t)) |
+--------------+
| 10000        |
+--------------+

Split a table with a specified size¶

'''
TABLE result_table TYPED TABLE {i INT, j INT}
FOR i IN range(1, 5)
FOR j IN range(1, 5)
EXPORT i, j INTO result_table

FOR t IN table_split(result_table, cast(10 as UINT))
RETURN size(t)
'''

Return:

+---------+
| size(t) |
+---------+
| 10      |
| 10      |
| 5       |
+---------+

The table contains 25 rows (5 * 5) and is split into three sub-tables: two with 10 rows each, and one with 5 rows.

Iterate over rows in sub-tables¶

'''
TABLE result_table TYPED TABLE {i INT, j INT}
FOR i IN range(1, 3)
FOR j IN range(1, 3)
EXPORT i, j INTO result_table

FOR t IN table_split(result_table)
FOR r IN t
RETURN r.i as i, r.j as j
'''

Return:

+---+---+
| i | j |
+---+---+
| 1 | 1 |
| 1 | 2 |
| 1 | 3 |
| 2 | 1 |
| 2 | 2 |
| 2 | 3 |
| 3 | 1 |
| 3 | 2 |
| 3 | 3 |
+---+---+