Where you use DataWindow expressions

A DataWindow expression is a combination of data, operators, and functions that, when evaluated, results in a value. An expression can include column names, operators, DataWindow expression functions, and constants such as numbers and text strings.

In painters

DataWindow expressions are associated with DataWindow objects and reports. You specify them in the DataWindow painter. You can also specify expressions in the Database painter, although these expressions have a slightly different format and are used only in validation rules.

For information about DataWindow expression functions that you can use in expressions, see “Using DataWindow expression functions”, or look up the function you want in online Help.

In painters, you use expressions in these ways:

**Table 1-1: Using DataWindow expressions in painters**
In this painter	Expressions are used in
DataWindow painter	Computed fields Conditional expressions for property values Validation rules Filters Sorting Series and values in graphs Columns, rows, and values in crosstabs
Database painter	Validation rules

Other types of expressions you use You also use expressions in Quick Select, SQL Select, and the Query painter to specify selection criteria, and in SQL Select and the Query painter to create computed columns. In these painters you are using SQL operators and DBMS-specific functions, not DataWindow expression operators and functions, to create expressions.

You can access and change the value of DataWindow data and properties in code. The format for expressions you specify in code is different from the same expression specified in the painter. These differences are described in Chapter 4, “Accessing Data in Code” and Chapter 5, “Accessing DataWindow Object Properties in Code.”

Some of the specific places where you use expressions are described here.

In computed fields

Expressions for computed fields can evaluate to any value. The datatype of the expression becomes the datatype of the computed field:

**Table 1-2: Using expressions in computed fields**
Expression	Description
Today ( )	Displays the date using the Today function
Salary/12	Computes the monthly salary
Sum (Salary for group 1)	Computes the salary for the first group using the Sum aggregate function
Price*Quantity	Computes the total cost

Expressions for graphs and crosstabs You can use similar expressions for series and values in graphs and for columns, rows, and values in crosstabs.

In filters

Filter expressions are boolean expressions that must evaluate to true or false:

**Table 1-3: Using expressions with filters**
Expression	Description
Academics = "*****" AND Cost = "$$$"	Displays data only for colleges with both a 5-star academic rating and a $$$ cost rating
Emp_sal < 50000	Displays data for employees with salaries less than $50,000
Salary > 50000 AND Dept_id BETWEEN 400 AND 700	Displays data for employees in departments 400, 500, 600, and 700 with salaries greater than $50,000
Month(Bdate) = 9 OR Month(Bdate) = 2	Displays data for people with birth dates in September or February
Match ( Lname, "[ ^ABC ]" )	Displays data for people whose last name begins with A, B, or C

In validation rules for table columns

Validation rules are boolean expressions that compare column data with values and that use relational and logical operators. When the validation rule evaluates to false, the data in the column is rejected.

In the DataWindow painter When you specify a validation rule in the DataWindow painter, you should validate the newly entered value. To refer to the newly entered value, use the GetText function. Because GetText returns a string, you also need a data conversion function (such as Integer or Real) if you compare the value to other types of data.

If you include the column name in the expression, you get the value that already exists for the column instead of the newly entered value that needs validating.

In the Database painter When you specify the validation rule in the Database painter, you are defining a general rule that can be applied to any column. Use @placeholder to stand for the newly entered value. The name you use for @placeholder is irrelevant. You can assign the rule to any column that has a datatype appropriate for the comparison.

When you define a DataWindow object, a validation rule assigned to a column is brought into the DataWindow object and converted to DataWindow object syntax. @placeholder is converted to GetText and the appropriate datatype conversion function.

Other columns in the rule You can refer to values in other columns for the current row by specifying their names in the validation rule:

**Table 1-4: Using expressions with values from other columns**
Expression in Database painter	Expression in DataWindow painter	Description
@column >= 10000	Integer(GetText())>= 10000	If a user enters a salary below $10,000, an error message displays.
@column IN (100, 200, 300)	Integer(GetText()) IN (100, 200, 300)	If a user does not enter a department ID of 100, 200, or 300, an error message displays.
@salary > 0	Long(GetText()) > 0	If a user does not enter a positive number, an error message displays.
Match(@disc_price, "^[0-9]+$") and @disc_price < Full_Price	Match(GetText( ), "^[0-9]+$") and Real(GetText()) < Full_Price	If a user enters any characters other than digits, or the resulting number is greater than or equal to the value in the Full_Price column, an error message displays.