Python - Specifics of Usage

• Duplicate function names are not allowed within a TestComplete script unit.

• In Python, code execution can start outside a function. In TestComplete, a test run must start from a function. You can start your tests from a function that does not have executable code. This function can contain an ellipsis and (optionally) comments:

  Python

  Python

  Log.Message("Test execution will start from the top.")
  
  def Test():
    # The ellipsis in the following line is not executed
    # However, the test will start
    ...

We do not guarantee stable work of your script tests that use third-party Python packages. When you run such tests, you may experience unexpected errors or TestComplete can stop responding.

# Set the new path to search for modules
import sys
sys.path.insert(0, "C:\\MyFolder\\MyPackage")

# Import the module
import MyModule

Notes:

• If you want to call TestComplete global scripting objects (for example, Sys, Mobile, Log, NameMapping and others) in your custom Python package, you need to add your package file to the Script collection of your TestComplete project by using the Add > Existing Item command. See Adding and Removing Project Items and Their Child Elements.

• If your custom Python packages have tags assigned to routines, the tags will not be available in TestComplete, and you will not be able to use them to run routines.

• In Python, you can import units to each other. At that, the Python units of your test project can import units that are not included in the project. TestComplete objects (like Log or Sys) are not available in these units. To use TestComplete objects in these units, include the units in your project, or import them as described in the Importing Tests topic.

• Python supports circular references between script units.

Though the TestComplete Python library includes the threading module, TestComplete does not support multithreaded Python scripts. Code in the threads that you create will not work. Calls to TestComplete objects, methods and properties from these threads will cause an exception.

In Python applications, the print statement writes an expression to standard application output. In TestComplete, there is no standard output, so print does nothing. To post log messages, use the methods of the Log object. See Posting Messages to the Log.

• Set Next Statement command is not supported for Python scripts.

• In the Inspect dialog, native Python members whose name starts with an underscore ( _ ) are displayed only if the Show hidden members option is enabled.

• The Inspect dialog is available for Python tuples and lists in the Watch List and Locals panels during debugging.

  Note:

  If the tuples and lists contain many items, the debugger panels and dialogs group them by 100 for convenient presentation.

Python is a strongly typed language. COM objects can have properties that return a collection object. Most scripting languages allow accessing collection items directly, for example, obj.Cells(i, j). TestComplete Python scripts do not support this to avoid ambiguity. To access items of a collection object, you should specify the appropriate property (Item, Items) explicitly:

To get or set the value of a parameterized property using the default parameter values, use the __getprop__ and __setprop__ methods instead of the property[] syntax:

In TestComplete, the __name__ variable always refers to the name of the unit where the function resides, for example, Unit1. TestComplete never assigns the value "__main__" to this variable. So, the condition

is always false.

Using Tkinter to create GUI elements for your TestComplete tests may cause errors. Use TestComplete User Forms instead of Tkinter.

If you have TestComplete version 14.73 or later, you can use the assignment expressions in your tests. The := operator allows assigning values as part of a larger expression. For example:

• Asynchronous coroutines - TestComplete does not support asynchronous coroutines declared with the async and await statements.

• Raw byte literals - literals with the rb or br prefixes, such as rb"test". As a workaround, you can create raw bytes objects by converting strings:

  Python

  Python

  a = str.encode(r'Test')
  # or
  a = bytes(r'Test', 'utf-8')