Contact Us 1-800-596-4880

DataWeave Operators

DataWeave supports several operators, including mathematical operators, equality operators, and operators such as prepend, append and update. Before you begin, note that 2.x versions of DataWeave are used by Mule 4 apps. For DataWeave in Mule 3 apps, refer to DataWeave version 1.2 operators. For other Mule versions, you can use the version selector in the DataWeave table of contents.

Mathematical Operators

DataWeave supports the most common mathematical operators:

Operator Description

+

For addition.

-

For subtraction.

*

For multiplication.

/

For division.

In addition to operating with numbers, the (-) and (+) operators can also operate with complex data structures like arrays, objects, and dates.

The following example uses mathematical operators with different data types:

Source
%dw 2.0
output application/json
---
{ "mathOperators" : [
    { "2 + 2" : (2 + 2) },
    { "2 - 2" : (2 - 2) },
    { "2 * 2" : (2 * 2) },
    { "2 / 2" : (2 / 2) },
    { "[1,2,3] - 1 + 4" : [1,2,3] - 1 + 4},
    { "{a:1, b:2, c:3} - 'a' " : {a:1, b:2, c:3} - "a"},
    { "|2021-03-02T10:39:59| - |P1D| + |PT3H|" : |2021-03-02T10:39:59| - |P1D| + |PT3H|}
  ]
}
Output
{
  "mathOperators": [
    { "2 + 2": 4 },
    { "2 - 2": 0 },
    { "2 * 2": 4 },
    { "2 / 2": 1 },
    { "[1,2,3] - 1 + 4": [2,3,4] },
    { "{a:1, b:2, c:3} - 'a' ": {"b": 2, "c": 3} },
    { "|2021-03-02T10:39:59| - |P1D| + |PT3H|": "2021-03-01T13:39:59" }]
}

Several DataWeave functions operate on numbers, for example: sum, mod (for modulo), and avg (for average).

Equality and Relational Operators

DataWeave supports the following equality and relational operators:

Operator Description

<

For less than.

>

For greater than.

<=

For less than or equal to.

>=

For greater than or equal to.

==

For equal to.

~=

Equality operator that tries to coerce one value to the type of the other when the types are different.

Note that you can negate these operators by using the logical operator, not.

The following example uses relational operators:

Source
%dw 2.0
output application/json
---
{ "relational" : [
    { "1 < 1" : (1 < 1) },
    { "1 > 2" : (1 > 2) },
    { "1 <= 1" : (1 <= 1) },
    { "1 >= 1" : (1 >= 1) }
  ]
}
Output
{ "relational": [
    { "(1 < 1)": false },
    { "(1 > 2)": false },
    { "(1 <= 1)": true },
    { "(1 >= 1)": true }
  ]
}

Note that if the operands of the relational operator belong to different types, DataWeave coerces the right-side operand to the type of the left-side operand. For example, in the expression "123" > 12 DataWeave coerces 12 (a Number type) to "12" (a String type) and compares each String value lexicographically. In the expression 123 > "12", DataWeave coerces the String value "12" to the Number value 12 and compares the numbers.

These examples use equality operators:

Source
%dw 2.0
output application/dw
---
{ "equality" :
  [
    (1 == 1),
    (1 == 2),
    ("true" == true),
    ("true" ~= true),
    (['true'] ~= [true]),
    ('1' ~= 1)
  ]
}
Output
{
  equality: [ true, false, false, true, true, true ]
}

Logical Operators

DataWeave supports the following logical operators:

Operator Description

not

Negates the result of the input. See also, !.

!

Negates the result of the input. See also, not. Introduced in DataWeave 2.2.0. Supported by Mule 4.2 and later.

and

Returns true if the result of all inputs is true, false if not.

or

Returns true if the result of any input is true, false if not.

Though the semantics of not and ! are the same, their precedence differs. not true or true is executed as not (true or true), so it returns false, whereas !true or true returns true because the ! only applies to the first true. !(true or true) returns false.

The following examples use logical operators:

Source
%dw 2.0
var myArray = [1,2,3,4,5]
var myMap = myArray map not (($ mod 2) == 0)
output application/json
---
{
  "not" : [
    "notTrue" : not true,
    "notFalse" : not false,
    "myMapWithNot" : myMap
  ],
  "and" : [
    "andTrueFalse" : true and false,
    "andIsTrue" : (1 + 1 == 2) and (2 + 2 == 4),
    "andIsFalse" : (1 + 1 == 2) and (2 + 2 == 2)
  ],
  "or" : [
    "orTrueFalse" : true or false,
    "orIsTrue" : (1 + 1 == 2) or (2 + 2 == 2),
    "orIsFalse" : (1 + 1 == 1) or (2 + 2 == 2)
  ],
  "!-vs-not" : [
	  "example-!" : (! true or true),
	  "example-not" : (not true or true)
  ]
}

Note that myMap iterates through the items in a list (myArray) and determines whether the modulo (mod) expression does not evaluate to 0 when applied to each given item.

Output
{
  "not": [
    { "notTrue": false },
    { "notFalse": true },
    { "myMapWithNot": [ true, false, true, false, true ] }
  ],
  "and": [
    { "andTrueFalse": false },
    { "andIsTrue": true },
    { "andIsFalse": false }
  ],
  "or": [
    { "orTrueFalse": true },
    { "orIsTrue": true },
    { "orIsFalse": false }
  ],
  "!-vs-not": [
    { "example-!": true },
    { "example-not": false }
  ]
}

Note that not works in expressions such as not (true), but not(true) (without the space) does not work.

You can use logical operators together. The following example uses:

  • or not as defined in the orNot expression.

  • and not in andNot.

  • not and and not in notWithAndNot.

%dw 2.0
var orNot = if (1 + 1 == 4 or not 1 == 2) {"answer": "orNot - Condition met"}
             else {"answer": "nope"}
var andNot = if (1 + 1 == 2 and not 1 == 2) {"answer": "andNot - Condition met"}
             else {"answer": "nope"}
var notWithAndNot = if (not (1 + 1 == 2 and not 1 == 1)) {"answer": "notWithAndNot - Condition met"}
              else {"answer": "nope"}
output application/json
---
{ "answers" :
  [
    orNot,
    andNot,
    notWithAndNot
  ]
}
Output
{
  "answers": [
    { "answer": "orNot - Condition met" },
    { "answer": "andNot - Condition met" },
    { "answer": "notWithAndNot - Condition met" }
  ]
}

DataWeave executes the code inside each if block because all conditional expressions in the example evaluate to true.

Prepend, Append, and Remove Operators for Arrays

DataWeave supports operators for appending and prepending items within an array:

Operator Description

>>

Prepends data on the left-hand side of the operator to items in the array on the right-hand side. For example, 1 >> [2] results in [ 1, 2 ], prepending 1 to 2 in the array.

<<

Appends data on the right-hand side of the operator to items in the array on the left-hand side. For example, [1] << 2 results in [ 1, 2 ], appending 2 to 1 in the array.

+

Appends data on the right-hand side of the operator to items in the array on the left-hand side. For example, [1] + 2 results in [ 1, 2 ], appending 2 to 1 in the array. The array is always on the left-hand side of the operator.

-

Removes a specified element of any supported type from an array.

The following examples show uses of prepend, append, and remove operators on arrays:

%dw 2.0
output application/json
---
{
  "prepend-append" : [
  	 // Array on right side when prepending.
     { "prepend" : 1 >> [2] },
     { "prepend-number" : 1 >> [1] },
     { "prepend-string" : "a" >> [1] },
     { "prepend-object" : { "a" : "b"} >> [1] },
     { "prepend-array" : [1] >> [2, 3] },
     { "prepend-binary" : (1 as Binary) >> [1] },
     { "prepend-date-time" : |23:57:59Z| >> [ |2017-10-01| ] },
  	 // Array is on left side when appending.
     { "append-number" : [1] << 2 },
     { "append-string" : [1] << "a" },
     { "append-object" : [1] << { "a" : "b"} },
     { "append-array" : [1,2] << [1, 2, 3] },
     { "append-binary" : [1] << (1 as Binary) },
     { "append-date-time" : [ |2017-10-01| ] << |23:57:59Z| },
     { "append-object-to-array" : [1,2] << {"a" : "b"} },
     { "append-array-to-array1" : ["a","b"] << ["c","d"] },
     { "append-array-to-array2" : [["a","b"],["c","d"]] << ["e","f"] },
     // + always appends within the array
     { "append-with-+" : [1] + 2 },
     { "append-with-+" : [2] + 1 },
     { "removeNumberFromArray" : ( [1,2,3] - 2 ) },
     { "removeObjectFromArray" : ( [ {a : "b"}, {c : "d"} , { e : "f"} ] - { c : "d"} ) }
  ]
}
Output
{
  "prepend-append": [
    { "prepend": [ 1, 2 ] },
    { "prepend-number": [ 1, 1 ] },
    { "prepend-string": [ "a", 1 ] },
    { "prepend-array": [ [ 1 ], 2, 3 ] },
    { "prepend-object": [ { "a": "b" }, 1 ] },
    { "prepend-binary": [ "\u0001", 1 ] },
    { "prepend-date-time": [ "23:57:59Z", "2017-10-01" ] },
    { "append-number": [ 1, 2 ] },
    { "append-string": [ 1, "a" ] },
    { "append-object": [ 1, { "a": "b" } ] },
    { "append-array": [ 1, 2, [ 1, 2, 3 ] ] },
    { "append-binary": [ 1, "\u0001" ] },
    { "append-date-time": [ "2017-10-01", "23:57:59Z" ] },
    { "append-object-to-array": [ 1, 2, { "a": "b" } ] },
    { "append-array-to-array1": [ "a", "b", ["c","d"] ] },
    { "append-array-to-array2": [ ["a","b"], ["c","d"], ["e","f"] ] },
    { "append-with-+": [ 1, 2] },
    { "append-with-+": [ 2, 1] },
    { "removeNumberFromArray": [ 1, 3 ] },
    { "removeObjectFromArray": [ { "a": "b" }, { "e": "f" } ] }
  ]
}

Scope and Flow Control Operators

DataWeave supports operators that control the flow and scope of expressions:

Table 1. Scope Operators:
Operator Description

do

Creates a scope in which new variables, functions, annotations, or namespaces can be declared and used. The syntax is similar to a mapping in that it is composed of a header and body separated by ---. Its header is where all the declarations are defined, and its body is the result of the expression. See do and Examples: Local DataWeave Variables for examples.

using

Replaced by do. Supported for backwards compatibility only.

Table 2. Flow Control Operators:
Operator Description

if else

An if operator evaluates a conditional expression and returns the value under the if only if the conditional expression is true. Otherwise, it returns the expression under else. Every if expression must have a matching else expression. See if else for an example.

else if

An else operator chains expressions together within an if-else construct by incorporating else if. See else if for an example.

Update Operator

DataWeave supports the update operator, which enables you to update specified fields of a data structure with new values.

Introduced in DataWeave 2.3.0. Supported by Mule 4.3 and later.

Versions prior to Mule 4.3 require the following syntax to increase the age value in myInput by one without recreating the entire object:

%dw 2.0
var myInput = {
    "name": "Ken",
    "lastName":"Shokida",
    "age": 30
}
output application/json
---
myInput mapObject ((value,key) ->
    if(key as String == "age")
       {(key): value as Number + 1}
    else
       {(key): value} )

In Mule 4.3, the update operator simplifies the syntax:

%dw 2.0
var myInput = {
    "name": "Ken",
    "lastName":"Shokida",
    "age": 30
}
output application/json
---
myInput update {
    case age at .age -> age + 1
}

Both DataWeave scripts return the same result:

{
  "name": "Ken",
  "lastName": "Shokida",
  "age": 31
}

With update, casting or iterating through all the key-value pairs is not necessary.

The update syntax is as follows:

<value_to_update> update {
    case <variable_name> at <update_expression>[!]? [if(<conditional_expression>)]? -> <new value>
    ...
}
  • value_to_update represents the original value to update.

  • update_expression is the selector expression that matches a value in value_to_update.

  • variable_name is the name of the variable to bind to the matched update_expression value.

  • new_value is the expression with the new value.

  • [if(<conditional_expression>)]? If conditional_expression returns true, the script updates the value. Use of a conditional expression is optional.

  • [!] marks the selector as an upsert. If the expression doesn’t find a match, the ! makes the update operator create all the required elements and insert the new value.

  • update_expression supports multiple case expressions.

The update operator doesn’t mutate the existing value. Instead, the operator creates a new value with the updated expressions.

Selector Expressions

Use selector expressions with the update operator to specify the fields to update.

Value Selector

This matching selector can check for matches to any value within an object.

The following example updates a field at the root level and a field in a nested level using the value selector:

Source
%dw 2.0
var myInput = {
    "name": "Ken",
    "lastName":"Shokida",
    "age": 30,
    "address": {
        "street": "Amenabar",
        "zipCode": "AB1234"
    }
}
output application/json
---
myInput update {
    case age at .age -> age + 1
    case s at .address.street -> "First Street"
}
Output
{
  "name": "Ken",
  "lastName": "Shokida",
  "age": 31,
  "address": {
    "street": "First Street",
    "zipCode": "AB1234"
  }
}

Index Selector

This matching selector enables you to match any array by its index.

The following example updates an element that is at the first location of the array:

Input
{
    "name": "Ken",
    "lastName":"Shokida",
    "age": 30,
    "addresses": [{
        "street": "Amenabar",
        "zipCode": "AB1234"
    }]
}
Source
%dw 2.0
output application/json
---
payload update {
    case s at .addresses[0] -> {
        "street": "Second Street",
        "zipCode": "ZZ123"
    }
}
Output
{
  "name": "Ken",
  "lastName": "Shokida",
  "age": 30,
  "addresses": [
    {
      "street": "Second Street",
      "zipCode": "ZZ123"
    }
  ]
}

Attribute Selector

This matching selector enables you to match any attribute value.

The following example updates an attribute value and changes it to upper case:

Input
<user name="leandro"/>
Source
%dw 2.0
output application/xml
---
payload update {
    case name at .user.@name -> upper(name)
}
Output
<?xml version='1.0' encoding='UTF-8'?>
<user name="LEANDRO"/>

Dynamic Selector

Any selector can incorporate an expression to make the selection dynamic.

The following example dynamically selects and updates the value of a field that is not hardcoded. Such a field can change at runtime. The example uses the variable theFieldName to obtain the field with key name, and it changes the value of that field to Shoki.

Input
{
    "name": "Ken",
    "lastName":"Shokida"
}
Source
%dw 2.0
var theFieldName = "name"
output application/json
---
payload update {
    case s at ."$(theFieldName)" -> "Shoki"
}
Output
{
  "name": "Shoki",
  "lastName": "Shokida"
}

Conditional Update

The update operator has syntax for writing a conditional update if you want to modify a field under a given condition.

The following example updates the category based on the name of the user:

Input
[{"name": "Ken", "age": 30}, {"name": "Tomo", "age": 70}, {"name": "Kajika", "age": 10}]
Source
%dw 2.0
output application/json
---
payload map ((user) ->
    user update {
        case name at .name if(name == "Ken") -> name ++ " (Leandro)"
        case name at .name if(name == "Tomo") -> name ++ " (Christian)"
    }
)
Output
[
  {
    "name": "Ken (Leandro)",
    "age": 30
  },
  {
    "name": "Tomo (Christian)",
    "age": 70
  },
  {
    "name": "Kajika",
    "age": 10
  }
]

Upserting

The update operator enables you to insert a field that is not present already by using the ! symbol at the end of the selector expression. When the field is not present, the operator binds a null value to the variable.

The following example updates each object in the array myInput. If the field name is present in the object, the example applies the upper function to the value of name. If the field name is not present in the object, the example appends the key-value pair "name": "JOHN" to the object.

Source
%dw 2.0
var myInput = [{"lastName": "Doe"}, {"lastName": "Parker", "name": "Peter" }]
output application/json
---
myInput  map ((value) ->
    value update {
        case name at .name! -> if(name == null) "JOHN" else upper(name)
    }
)
Output
[
  {
    "lastName": "Doe",
    "name": "JOHN"
  },
  {
    "lastName": "Parker",
    "name": "PETER"
  }
]

Sugar Syntax

The update operator accepts a default variable named $ in place of the longer <variable_name> at syntax.

The following example binds the value of the expression .age to the default variable name $, which is used to express the new value. The longer equivalent is shown in comments.

Input
{
    "name": "Ken",
    "lastName":"Shokida",
    "age": 30,
    "address": {
        "street": "Amenabar",
        "zipCode": "AB1234"
    }
}
Source
%dw 2.0
output application/json
---
payload update {
    //equivalent expression:
    //case age at .age -> age + 1
    case .age -> $ + 1
    case .address.street -> "First Street"
}
Output
{
  "name": "Ken",
  "lastName": "Shokida",
  "age": 31,
  "address": {
    "street": "First Street",
    "zipCode": "AB1234"
  }
}

Metadata Assignment Operator

DataWeave supports the metadata assignment operator <~, which enables you to set the metadata of any value without using the as operator. This operator is useful when you want to set the metadata and retain the current value type. If you are coercing the value to a new type, use as.

Introduced in DataWeave 2.5.0. Supported by Mule 4.5 and later.

The operator syntax is as follows:

<valueB> <~ <objectA>

The operator assigns objectA as metadata to valueB. The arrow points to the value to which the metadata is assigned.

Versions prior to Mule 4.5 require the following syntax, for example, to add the class metadata field to an object:

%dw 2.0
---
{name: "Ken"} as Object {class: "Person"}

In Mule 4.5, the metadata assignment operator <~ simplifies the syntax:

%dw 2.0
---
{name: "Ken"} <~ {class: "Person"}

Metadata

Metadata is any additional information attached to a value that doesn’t affect its identity. Given any two values A and B that are equal, if you attach metadata to A and call it A', then A == A' == B. So you can define a value that is formed by the value and the metadata attached to it.