SPLITLIST ER function
The SPLITLIST
function splits the specified list into sublists (or batches), each of which contains the specified number of records. It then returns the result as a new Record list value that consists of the batches.
Syntax 1
SPLITLIST (list, number)
Syntax 2
SPLITLIST (list, number, on-demand reading flag)
Arguments
list
: Record list
The valid path of a data source of the Record list data type.
number
: Integer
The maximum number of records per batch.
on-demand reading flag
: Boolean
A Boolean value that specifies whether elements of sublists should be generated on demand.
Return values
Record list
The resulting list of records.
Usage notes
The list of batches that is returned contains the following elements:
Value: List
The list of records that belong to the current batch.
BatchNumber: Integer
The number of the current batch in the returned list.
When the on-demand reading flag is set to True, sublists are generated upon request which allows for a reduction in memory consumption but may cause performance degradation if elements aren't used sequentially.
Example
In the following illustration, a Lines data source is created as a record list that has three records. This list is divided into batches, each of which contains up to two records.
The following illustration shows the designed format layout. In this format layout, bindings to the Lines data source are created to generate output in XML format. This output presents individual nodes for each batch and the records in it.
The following illustration shows the result when the designed format is run.