When you provide values for the test step properties in an API test, the default way to provide these values is to manually enter them in the Properties pane.
However, using this method does not necessarily provide a realistic test of your application. In many applications, internal API processes receive their information - often dynamically- from data sources, other test step outputs, or a variety of both.
As a result, UFT provides a number of different ways to provide (parameterize) step values:
When you manually enter the step values, you select each step and type or select the appropriate values for the input and checkpoint property values.
This method does not provide for easy test maintenance, as each time that your application's properties change, you must update each step and each property in the test.
|Linking to a data source||
When you link the step properties to a data source, UFT takes the values from the data source during a test run and uses the value provided in the data source. If your data source has multiple sets of data, you can run a test with multiple iterations to provide different values for the input and checkpoints to see how your application performs with different data input.
This method provides easier test maintenance, as you only need to update the data source values instead of each test step.
|Linking to the output of a previous step||When you link the step properties to the output of a previous step, UFT takes the values from the output of the step and uses these values during the test run. This enables you to mimic real application behavior, in which the output of an application's API process often passes a value to another process as input.|
|Linking to multiple sources||If your application's input and checkpoints are provided from multiple sources - a static string, data, and output of other steps/processes, you can create custom expressions to perform this in your test. UFT then provides the values using the custom expressions and uses the values during the test run.|