Version

    Data Types in CTL2

    For basic information about data types used in metadata, see Data Types in Metadata.

    In any program, you can use some variables. Data types in CTL are the following:

    booleannumber (double)
    bytestring
    cbytelist
    datemap
    decimalvariant
    integerrecord
    long 

    boolean

    The boolean data type contains values of logical expressions.

    The default value is false.

    It can be either true or false.

    Its declaration looks like this: boolean identifier;

    Example 67.5. Declaration of boolean variable

    boolean b;        // declaration
    boolean b = true; // declaration with assignment

    byte

    This data type stores binary data of a length that can be up to Integer.MAX_VALUE as a maximum.

    The default value is null.

    Its declaration looks like this: byte identifier;

    Example 67.6. Declaration of byte variable

    byte b;
    // declaration of variable with assignment
    byte b = hex2byte("414243");

    cbyte

    This data type is a compressed representation of byte data type to reduce runtime memory footprint. Compressed size of the data can be up to Integer.MAX_VALUE as a maximum.

    The default value is null.

    Its declaration looks like this: cbyte identifier;

    Example 67.7. Declaration of cbyte variable

    cbyte c1;
    cbyte c2 = hex2byte("61"); // declaration with assignment

    date

    The date data type contains date and time.

    The default value is 1970-01-01 00:00:00 GMT.

    Its declaration looks like this: date identifier;

    Example 67.8. Declaration of date variable

    // declaration of variable
    date d;
    // declaration of variable with assignment from function
    date d = str2date("1600-01-31", "yyyy-MM-dd");

    [Note]Note

    If you work with date, you should be aware of time zone of the data.

    decimal

    The decimal data type serves to store decimal numbers.

    Calculations with the decimal data type are performed in fixed point arithmetic. It makes decimal data type suitable for calculations with money.

    The default value is 0.

    Its declaration looks like this: decimal identifier;

    By default, any decimal may have up to 32 significant digits. If you want to have different Length or Scale, you need to set these properties of decimal field in metadata.

    Example 67.9. Usage of decimal data type in CTL2

    If you assign 100.0 / 3 to a decimal variable, its value might for example be 33.333333333333335701809119200333. As 100.0 is double and 3 is integer, the both operands were firstly converted to double, then the value has been calculated and finally the result value has been converted to decimal. Assigning it to a decimal field (with default Length and Scale, which are 12 and 2, respectively), it will be converted to 33.33D.


    You can cast any float number to the decimal data type by appending the d letter to its end.

    Any numeric data type (integer, long, number/double) can be converted to decimal.

    Example 67.10. Declaration of decimal variable

    decimal d;
    decimal d2 = 4.56D; // declaration of variable with assignment

    integer

    The integer data type can contain integral values.

    CTL2 integer can store values from -2147483648 to 2147483647.

    The integer data type can overflow (i.e. adding 1 to the maximum value returns -2147483648; similarly, subtracting 1 from the minimum value returns 2147483647) which may lead to errors and/or incorrect results.

    The default value is 0.

    Its declaration looks like this: integer identifier;

    [Important]Important

    The value -2147483648 can be stored in CTL2 variable but cannot be stored in an integer field of record metadata (value of the field would be null). If the value -2147483648 is expected to arise, consider usage of data type with wider range of values in metadata; e.g. long.

    If you append the L letter to the end of any integer number, you can cast it to the long data type.

    Integer can be converted to long, double or decimal using automatic conversions.

    Example 67.11. Declaration of integer variable

    integeri1;
    integer i2 = 1241;

    long

    long is an integral data type allowing to store greater values than the integer data type.

    CTL2 long can store values from -9223372036854775808 to 9223372036854775807.

    The long data type can overflow (i.e. adding 1 to the maximum value returns -92233720368547758088; similarly, subtracting 1 from the minimum value returns 9223372036854775807) which may lead to errors and/or incorrect results.

    The default value is 0.

    Its declaration looks like this: long identifier;

    [Important]Important

    The value -9223372036854775808 can be stored in CTL2 variable but the value is used in long field in record metadata for null value. If the value -9223372036854775808 is expected to arise, consider usage of data type with wider range of values in metadata; e.g. decimal.

    Any integer number can be cast to long data type by appending the l letter to its end.

    Long data type can be converted to number/double or decimal without explicit casting.

    Example 67.12. Declaration of long variable

    long myLong;
    long myLong2 = 2141L;

    number (double)

    The number data type is used for floating point number.

    The default value is 0.0.

    Its declaration looks like this: number identifier;

    If you need a data type for money amount, we advise using decimal instead of number (double).

    The integer and long data types can be converted to double using automatic conversions. If long is being converted to number (double), lost of precision may occur.

    Number(double) can be converted to decimal without explicit casting.

    Example 67.13. Declaration of number (double) variable

    double d;
    double d2 = 1.5e2;

    string

    This data type serves to store sequences of characters.

    The default value is empty string.

    The declaration looks like this: string identifier;

    Example 67.14. Declaration of string variable

    string s;
    string s2 = "Hello world!";

    list

    Since CloverDX 5.6, the type of elements of a list may be any other data type, including nested lists or maps.

    The elements of a list are indexed by integers starting from 0.

    Its declaration can look like this: string[] identifier;

    For nested lists or maps, use the following syntax instead: list[<type of elements>] identifier;

    The default list is an empty list.

    Example 67.15. List

    integer[] myIntegerList;
    myIntegerList[5] = 123;
    
    // Customer is metadata record name
    Customer JohnSmith;
    Customer PeterBrown;
    Customer[] CompanyCustomers;
    CompanyCustomers[0] = JohnSmith;
    CompanyCustomers[1] = PeterBrown;
    
    // Nested lists and maps:
    list[list[string]] listOfLists;
    list[map[string, integer]] listOfMaps;

    Assignments:
    • myStringList[3] = "abc";

      The string "abc" is put to the fourth position in the string list. The preceding items are filled with null as follows:

      myStringList is [null,null,null,"abc"]

    • myList1 = myList2;

      Assigns a copy of myList2 to myList1. It means that both lists will contain the same elements.

    • myList1 = myList1 + myList2;

      Adds all elements of myList2 to the end of myList1.

      Both lists must be based on the same primitive data type.

    • myList = [];

      Assigns an empty list to myList.

    • myList = ["a", "b", "c"];

      Assigns a list containing three strings to myList.

    • myList = null;

      Discards the previous value of myList.

    map

    This data type is a container of pairs of a key and a value.

    Its declaration looks like this: map[<type of key>, <type of value>] identifier;

    Since CloverDX 5.6, the Value can be any of the other data types, including records, nested lists or other maps, but the Key can only be a primitive data type: boolean, date, decimal, integer, long, number or string.

    The default map is an empty map.

    Example 67.16. Map

    map[string, boolean] map1;
    map1["abc"] = true;
    
    // Customer is the name of record
    Customer JohnSmith;
    Customer PeterBrown;
    
    map[integer, Customer] CompanyCustomersMap;
    CompanyCustomersMap[JohnSmith.ID] = JohnSmith;
    CompanyCustomersMap[PeterBrown.ID] = PeterBrown;
    
    // Nested maps and lists:
    map[string, map[string, integer]] mapOfMaps;
    map[integer, list[string]] mapOfLists;

    The assignments are similar to those valid for a list:

    • myMap["abc"] = 7;

      Puts the value 7 into myMap under the key "abc".

    • myMap2 = myMap1;

      Assigns a copy of myMap1 to myMap2.

    • myMap = {};

      Assigns an empty map to myMap.

    • myMap = { "a" -> 20, "b" -> 10, "c" -> 30 };

      Assigns a map containing three key-value pairs to myMap.

    • myMap = null;

      Discards the previous value of myMap.

    variant

    Variant is a new data type added in CloverDX 5.6. Variables of this type can be assigned values of any other type - no type checking is performed. In particular, variant can contain nested lists and maps, so it can be used for tree-like data with unknown structure, such as JSON.

    Its declaration looks like this: variant identifier;

    [Important]Important

    Variant data type is only supported in CTL, not in data records. In order to transport a variant data structure from one component to another, it needs to be converted to a string or bytes and then reconstructed in the following component.

    You can use the following conversion functions:

    • writeJson

      • Converts the data to JSON, a commonly used text based format. Some type information may be lost, for example, dates are written as strings.

    • writeBson

      • Produces standard BSON, a binary format preserving data types. Use this function for compatibility with third-party applications supporting BSON. Has some limitations originating from the BSON format - the top-level structure must be a map.

    • writeExtendedBson

      • The recommended way of transporting variant data internally within CloverDX. Produces a proprietary binary format based on BSON. Incompatible with third-party applications.

    Variant can be used like lists and maps, allowing to access inner values using square brackets [ ]. The operation will fail unless the variable contains a list or a map at runtime.

    The default value is null, so the variable must be initialized to an empty list or map before inserting inner values.

    Functions with arguments of type variant can be passed any value. However, they may throw runtime exceptions if the value is not valid for the function. For example, append(variant list, variant element) can be passed any value as the first argument, but it will throw an exception unless the value really is a list.

    The type supports only a few basic operations (== and != comparison, toString, etc.). In order to perform type-specific operations, the values must be explicitly type-cast to a more specific type. See typeof operator and cast and getType functions.

    Example 67.17. Variant

    variant myVariant = {};
    myVariant["one"] = 1;
    myVariant["string"] = "not a number";
    myVariant["two"] = 2;
    
    // working with unknown structures:
    integer sum = 0;
    if ( myVariant typeof list ) {
        for  ( integer index = 0; index < length(myVariant); index++) { // iterate through the list
            variant element = myVariant[index]; // get list element by the index 
            string type = getType(element); // get the type of the element 
            printLog(info, "List element " + index + ": " + element + " " + type);
            if (element typeof integer) { // test the type of the element 
                sum += cast(element, integer); // cast to integer and add to the sum 
            }
        }
    } else if ( myVariant typeof map ) {
        variant keys = getKeys(myVariant); // returns the keys as a list 
        for ( integer i = 0; i < length(keys); i++) { // iterate through the list of keys
            variant key = keys[i]; // get the key by the index 
            variant value = myVariant[key]; // get the value by the key 
            string type = getType(value); // get the type of the value 
            printLog(info, "Map entry " + key + " = " + value + " " + type);
            if (value typeof integer) { // test the type of the value 
                sum += cast(value, integer); // cast to integer and add to the sum 
            }
        }
    }
    printLog(info, "Sum: " + sum);
                    

    The assignments are similar to those valid for a list or a map:

    • variant varMap = {};

      Assigns an empty map to varMap.

    • varMap["abc"] = 7;

      If varMap contains a map, puts the value 7 into varMap under the key "abc". Throws an exception otherwise.

    • variant varList = [];

      Assigns an empty list to varList.

    • varList[5] = "abc"

      If varList contains a list with at least 6 elements, sets the list element at index 5 to "abc". Unlike with list data type, variant is not expanded automatically, so if varList contains fewer than 6 elements, the assignment fails. Use append to expand the list. If varList is actually a map, puts "abc" into the map under the key 5. Otherwise, throws an exception.

    • var2 = var1;

      Assigns a copy of var1 to var2.

    • varMap = {
          "name" -> "John Doe",
          "weight" -> 75.3,
          "valid" -> true
      };

      Assigns a JSON-like map containing three key-value pairs to varMap. Note that the values are of mixed types: string, number and boolean, respectively.

    • varMap = null;

      Discards the previous value of varMap.

    record

    Record is a container that can contain different primitive data types.

    The structure of record is based on metadata. Any metadata item represents a data type.

    Declaration of a record looks like this: <metadata name> identifier;

    Metadata names must be unique in a graph. Different metadata must have different names.

    For more detailed information about possible expressions and records usage, see Accessing Data Records and Fields.

    Record does not have a default value.

    It can be indexed by both integer numbers and strings (field names). If indexed by numbers, fields are indexed starting from 0.