Test Project Overview¶
Contents
Setup¶
Initialization¶
To generate files for a new test suite, change into the desired directory and run:
wtt init
This will generate a new test package with template files and project directories.
(Optional) Setup Virtual Environment¶
If using virtualenv, initialize the virtual environment before installing the test package:
virtualenv venv
source ./venv/bin/activate
Test Package Installation¶
After initializing the test project, run the following command from the project root directory:
pip install -e .
Installing with the -e
flag will update the package automatically when
changes are made to the source code.
Configuration¶
After initializing a project, the URL of the site to be tested will need to be
configured. In <test_package>/config/site.py
, set the SITE_URL
and
BASE_URL
of the SiteConfig
class. You can add any other URLs you’ll need
as class variables as well.
Basic Command Line Usage¶
Usage:
<test_package> [-h] <command>
If no <command>
is specified, the run
command will be executed by
default.
Note: If the test package was not installed with pip
, run test packages
commands using python -m <test_package> <command>
.
For info on command line arguments, use the --help
(or -h
) argument:
<test_package> --help
Creating New Project Files¶
New tests and page objects can be generated using the new
command:
<test_package> new [<type>] [<module_name>] [<ClassName>] [-d <description>] [-f]
Where:
<type>
: The type of file to create (test
orpage
)<module_name>
: Filename to use for the new python module<ClassName>
: Name to use for the initial class<description>
: (Optional) Description for the initial class-f
: (Optional) Force overwrite if a file with the same name already exists
If no arguments are provided, a prompt will walk you through generating the new file. Alternatively, you can skip the prompts by using the arguments shown in the following sections.
Creating New Tests¶
New test modules can be generated using the new test
command:
<test_package> new test <module_name> <TestCaseClass>
Where <module_name>
is the filename for the new test and <TestCaseClass>
is the class name for the test case.
The --description
(or -d
) argument can be used to add a description for
the initial test case class:
<test_package> new test <module_name> <TestCaseClass> -d "Test case description"
If a test module with the same <module_name>
already exists, new test
will not overwrite it by default. The --force
(or -f
) argument can be
used to force overwrite existing files:
<test_package> new test <module_name> <TestCaseClass> --force
Creating New Page Objects¶
New page object modules can be generated using the new page
command:
<test_package> new page <module_name> <PageObjectClass>
Where <module_name>
is the filename for the new module and
<PageObjectClass>
is the class name for the page object.
The --description
(or -d
) argument can be used to add a description for
the initial page object class:
<test_package> new page <module_name> <PageObjectClass> -d "Page object description"
By default, the new class will be a generic BasePage
subclass. The
--prototype
(or -p
) argument can be used to specify a page object
prototype class to use as a parent class for the new page object:
<test_package> new page <module_name> <PageObjectClass> -p <prototype>
For a list of valid <prototype>
options, run <test_package> new page
--help
.
Page object prototypes support YAML file parsing to simplify the syntax of
representing elements. Whether YAML files or Python-only files are generated by
default is configured in <test_package>/config/projectfiles.py
by setting
the ENABLE_PAGE_OBJECT_YAML
variable of the ProjectFilesConfig
class.
The default setting can be overridden by using command line arguments. If
ENABLE_PAGE_OBJECT_YAML
is True
, the --no-yaml
(or -Y
) argument
can be used to only generate .py
files:
<test_package> new page <args> --no-yaml
If ENABLE_PAGE_OBJECT_YAML
is False
, the --yaml
(or -y
) argument
can be used to generate .py
and .yml
files for supported prototypes:
<test_package> new page <args> --yaml
If a page module with the same <module_name>
already exists, new page
will not overwrite it by default. The --force
(or -f
) argument can be
used to force overwrite existing files:
<test_package> new page <module_name> <PageObjectClass> --force
Running Tests¶
Running Specific Tests¶
To run all test cases in one or more modules, use the --module
(or -m
)
argument:
<test_package> --module <test_module> [<test_module> ...]
To skip all test cases in one or more modules, use the --skip-module
(or
-S
) argument:
<test_package> --skip-module <test_module> [<test_module> ...]
To run specific test case classes or methods, use the --test
(or -t
)
argument:
<test_package> --test <TestClass>[.<test_method>] [<TestClass>[.<test_method>] ...]
To skip certain test cases or methods, use the --skip
(or -s
) argument:
<test_package> --skip <TestClass>[.<test_method>] [<TestClass>[.<test_method>] ...]
The --test
and --skip
arguments both support wildcards (*
) in class
and method names.
These arguments can be used together. When combined, they are processed in the following order:
--skip-module
removes the specified modules from the set of tests
--module
reduces the set of tests to those in the specified modules
--test
reduces the set of tests to the specified classes and methods
--skip
removes the specified classes and methods from the set of tests
Using Specific Browsers¶
To do any of the above in specific browsers rather than running in all available
browsers, use the --browser
(or -b
) argument:
<test_package> <args> --browser <browser> [<browser ...]
For a list of options you can specify with --browser
, run <test_package>
--help
.
Using Headless Browsers¶
By default, tests run using the browser’s GUI. While it can be helpful to see what’s going on during test execution, loading and rendering the browser window can be resource-intensive and slows down performance during test execution.
To improve performance, tests can be run in headless browsers using the
--headless
(or -H
) argument:
<test_package> <args> --headless
Note: When using the --headless
argument, tests will only be run with
the following web drivers that support running in a headless environment:
Using BrowserStack¶
Test projects can be configured to run tests on BrowserStack. Once
BrowserStack support is enabled, tests can be run on BrowserStack using the
--browserstack
(or -B
) argument:
<test_package> <args> --browserstack
See the documentation on BrowserStack Support for more details and setup instructions.
Configuring Output¶
By default, detailed output is displayed when running tests. To reduce or
suppress output, use the --verbosity
(or -v
) argument:
<test_package> <args> --verbosity <level>
Where <level>
is one of the following:
0 - Final results only
1 - Final results and progress indicator
2 - Full output
Note: The default output level can be changed in
<test_package>/config/test.py
by setting the DEFAULT_VERBOSITY
attribute of the TestSuiteConfig
class.
List Available Tests¶
Basic Usage¶
To print a list of available test classes and methods:
<test_package> list
To include docstrings for each test class and method in output:
<test_package> list --verbose
Listing Specific Tests¶
To only list test classes from specific modules:
<test_package> list --module <test_module> [<test_module> ...]
To omit specific modules:
<test_package> list --skip-module <test_module> [<test_module> ...]
To only list specific test classes:
<test_package> list --test <TestClass> [<TestClass> ...]
To skip certain test classes in output:
<test_package> --skip <TestClass> [<TestClass> ...]
See Running Specific Tests for more info on these arguments.
Project Structure¶
wtt init
will create the following files and directories
inside the project directory:
<project-directory>/
├── README.rst
├── setup.py
└── <test_package>/
├── __main__.py
├── __init__.py
├── config/
│ ├── __init__.py
│ ├── browser.py
│ ├── browserstack.py
│ ├── projectfiles.py
│ ├── site.py
│ ├── test.py
│ └── webdriver.py
├── data.py
├── log/
├── pages/
│ └── __init__.py
├── screenshot/
└── tests/
└── __init__.py
This test structure is designed to be used with the Page Object Model. Interaction with the page should be handled by page objects to minimize the need to alter tests whenever the HTML is changed.
Test Project Root Contents¶
setup.py
: Python package setup file that allows the new test suite to be installed as a pip package.
Test Package Root Contents¶
__main__.py
: Required to run tests from the command line.__init__.py
: Empty init file so Python recognizes the directory as a package.data.py
: Module for storing static data for tests that must use specific values (e.g. emails, usernames, etc).
Test Package Directories¶
config/¶
Configurations used by test scripts for site URLs, web driver options, and the python unittest framework.
browser.py
: Configure which browsers to run tests in.browserstack.py
: Enable and configure testing with BrowserStack.projectfiles.py
: Configure defaults for generating project files with thenew
command.site.py
: Configure URLs used for testing.test.py
: Configure theunittest.TestRunner
class.webdriver.py
: Configure WebDrivers and log output directory.
pages/¶
Page object classes for pages and components. These classes should handle locating and interacting with elements on the page. See Creating New Page Objects for info on generating new page object modules.
screenshot/¶
Default output directory for screenshots taken during test execution. This can
be changed in config/webdriver.py
.
tests/¶
Test case modules. These use page objects to interact with elements and assert that the expected behavior occurs. See Creating New Tests for info on generating new test modules.