You are here:
Transforming Lists with Data Mappers
Data Mappers can perform list transformations. For mapping input to output, Data Mappers require a list input in which every value is paired with a key.
For example, the following lists are valid, and each can be converted to the other.
List 1:
{
"Oldest": "Huey",
"Middle": "Dewey",
"Youngest": "Louie"
}
List 2:
{
"Nephews": [
{
"Name": "Huey"
},
{
"Name": "Dewey"
},
{
"Name": "Louie"
}
]
}List 1 lacks a JSON node for the list as a whole, and each value has a different key. List 2 has its own JSON node, and each list item has the same key.
To convert List 1 to List 2, use the following mappings on the Transforms tab of a Data
Mapper Transform. The |1, |2, and |3 are
list position indexes.
Input JSON Path |
Output JSON Path |
|---|---|
Oldest |
Nephews|1:Name |
Middle |
Nephews|2:Name |
Youngest |
Nephews|3:Name |
To convert List 2 to List 1, use the reverse mappings.
In the following list, values don't have keys.
List 3:
{
"Nephews": [
"Huey",
"Dewey",
"Louie"
]
}
You can convert List 1 to List 3 using the following mappings:
Input JSON Path |
Output JSON Path |
|---|---|
Oldest |
Nephews|1 |
Middle |
Nephews|2 |
Youngest |
Nephews|3 |
However, you can't convert List 3 to either List 1 or List 2.
To convert List 2 to List 3, and to perform such conversions of lists of any length, see Convert a List of Objects to a List of Values.
If each object of a list has a different key, Data Mapper Transform creates a single map with combined key and value pairs from that list, and ignores any empty values. Here’s an example for products for an ecommerce website:
{
"Cleaning products": [
{
"Name": "WonderProduct"
"Property": ""
},
{
"Name": ""
"Property": "Fast-acting"
},
{
"Scent": "Lavender"
}
]
}
Transformed output list:
{
"Cleaning products": [
{
"Name": "WonderProduct"
"Property": "Fast-acting"
"Scent": ":Lavender"
}
]
}
For more that you can do with lists, see:
-
AVG, FILTER, IF, LIST, LISTMERGE, LISTMERGEPRIMARY, LISTSIZE, MAX, MIN, SORTBY, and SUM functions in the Supported Data Mapper and Integration Procedure Functions
-
Work with Data and Lists in Integration Procedures
-
List Action for Integration Procedures in Integration Procedures
-
Integration Procedure Action for Integration Procedures in Integration Procedures (the example transforms a list)
Expected Behavior for Array Inputs
When a list input (array) is mapped to an output that is expected to be a single object (non-array), the Data Mapper Transform automatically flattens the array or merges the objects based on the number of records in the input array.
- If the list input array has two records, the Data Mapper Transform flattens the array
and attempts to merge the records into a single JSON object in the output. The output node
is a single object
{}instead of an array[].Example input:
{ "In": [ { "field1": "data1" }, { "field2": "data2" } ] }Example output:
{ "Out": { "field1": "data1", "field2": "data2" } } - If the input array contains more than two records (multiple objects), the Data Mapper
Transform retains the array structure and does not flatten the records. The array is
returned in the output as-is.
Example input:
{ "In": [ { "field1": "data1" }, { "field2": "data2" }, { "field3": "data3" } ] }Example output:
{ "Out": [ { "field1": "data1" }, { "field2": "data2" }, { "field3": "data3" } ] }
