In contrast to certain programming languages, Ballerina does not facilitate implicit conversion among its numerical types. This implies that without explicit conversion, a numerical value cannot be assigned to a variable of a differing numerical type.

Ballerina takes this approach to prevent unintended loss of precision and to safeguard against unexpected program behavior. For instance, when a float value is assigned to an int variable, the fractional portion of the float value will be discarded. Similarly, assigning an int value to a float variable converts the value to a float, introducing a fractional part of zero. While these transformations may not pose a problem when done knowingly, unawareness can lead to unpredictable program behavior. Consequently, Ballerina insists on explicit conversion of numerical values.

In the specification, the conversion rules are defined as follows: (See under NumericConvert)

If required, you can use a type-cast expression to do a numeric conversion. This is useful when converting a mapping value to another with the same field name but with a different numerical type.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22

import ballerina/io;

public function main() {

    int intValue = 10;
    float floatValue = 2.1;
    decimal decimalValue = 5.0;

    int v1 = <int>floatValue;
    int v2 = <int>decimalValue;
    float v3 = <float>intValue;
    float v4 = <float>decimalValue;
    decimal v5 = <decimal>intValue;
    decimal v6 = <decimal>floatValue;

    io:println("v1:", v1);
    io:println("v2:", v2);
    io:println("v3:", v3);
    io:println("v4:", v4);
    io:println("v5:", v5);
    io:println("v6:", v6);
}

v1:2
v2:5
v3:10.0
v4:5.0
v5:10.0
v6:2.1

Ballerina 2201.6.0 (Swan Lake Update 6)

// Partial Code. Change View  to see full code.
public function main() {
    int intValue = 10;
    float floatValue = 2.1;
    decimal decimalValue = 5.0;

    int v1 = <int>floatValue;
    int v2 = <int>decimalValue;
    float v3 = <float>intValue;
    float v4 = <float>decimalValue;
    decimal v5 = <decimal>intValue;
    decimal v6 = <decimal>floatValue;
}

// Highlights Only. Change View  to see full code.
int intValue = 10;
float floatValue = 2.1;
decimal decimalValue = 5.0;

int v1 = <int>floatValue;
int v2 = <int>decimalValue;
float v3 = <float>intValue;
float v4 = <float>decimalValue;
decimal v5 = <decimal>intValue;
decimal v6 = <decimal>floatValue;

Code Breakdown

• Line 9: This is the same as <int>f.round(0). This uses Ballerina default round-to-nearest rounding mode. It is the same as IEEE roundToIntegralTiesToEven operation.
• Line 10: This is the same as <int>d.round(0). This uses Ballerina default round-to-nearest rounding mode. It is the same as IEEE roundToIntegralTiesToEven operation with a minor exception to handling positive exponent.

Another way to convert a numerical value is to use the value:cloneWithType() function.

🚧 More Details Coming Soon!