SST - Web Test Framework
Automated Testing with Python
SST (selenium-simple-test) is a web test framework that uses Python to generate functional browser-based tests.
Tests are made up of scripts, created by composing actions that drive a browser and assert conditions. You have the flexibilty of the full Python language, along with a convenient set of functions to simplify web testing.
SST consists of:
- user actions and assertions (API) in Python
- test case loader (generates/compiles scripts to unittest cases)
- console test runner
- data parameterization/injection
- selectable output reports
- selectable browsers
- headless (xvfb) mode
- screenshots on errors
Test output is displayed to the console and optionally saved as JUnit-compatible XML for compatibility with CI systems.
pip install -U sst
For example, on an Ubuntu/Debian system, you could Install SST (system-wide) like this:
$ sudo apt-get install python-pip xvfb $ sudo pip install -U sst
or with a virtualenv:
$ sudo apt-get install python-virtualenv xvfb $ virtualenv ENV $ source ENV/bin/activate (ENV)$ pip install sst
- note: xvfb is only needed if you want to run SST in headless mode
Example SST test script
a sample test case in SST:
from sst.actions import * go_to('http://www.ubuntu.com/') assert_title_contains('Ubuntu')
Running a test with SST
Create a Python script (.py) file, and add your test code.
Then call your test script from the command line, using sst-run:
$ sst-run mytest
- note: you don’t add the .py extension to your test invocation
Actions reference (sst.actions)
Test scripts perform actions in the browser as if they were a user. SST provides a set of “actions” (functions) for use in your tests. These actions are defined in the following API:
Command line options for sst-run
Usage: sst-run <options> [testname]
- Calling sst-run with testname(s) as arguments will just run those tests. The testnames should not include ‘.py’ at the end of the filename.
- You may optionally create a data file for data-driven testing. Create a ‘^’ delimited txt data file with the same name as the test, plus the ‘.csv’ extension. This will run a test using each row in the data file (1st row of data file is variable name mapping)
For logical organization of tests, you can use directories in your filesystem. SST will recursively walk your directory tree and gather all tests for execution.
For example, a simple test setup might look like:
/selenium-simple-test /mytests foo.py
and you would call this from the command line:
$ sst-run -d mytests
A more complex setup might look like:
/selenium-simple-test /mytests /project_foo /feature_foo foo.py /project_bar feature_bar.py feature_baz.py /shared module.py utils.py
and you would still call this from the command like:
$ sst-run -d mytests
SST will find all of the tests in subdirectories (including symlinks) and execute them. SST won’t look in directories starting with an underscore. This allows you to put Python packages/modules directly in your test directories if you want. A better option is to use the shared directory.
Using sst in unittest test suites
sst uses unittest test cases internally to wrap the execution of the script and taking care of starting and stopping the browser. If you prefer to integrate some sst tests into an existing unittest test suite you can use SSTTestCase from runtests.py:
from sst.actions import * from sst import runtests class TestUbuntu(runtests.SSTTestCase): def test_ubuntu_home_page(self): go_to('http://www.ubuntu.com/') assert_title_contains('Ubuntu')
So, with the above in a file name test_ubuntu.py you can run the test with (for example):
python -m unittest test_ubuntu.py
sst-run provides an headless xserver via the -x option. SSTTestCase provides the same feature (sharing the same implementation) via two class attributes.
xserver_headless when set to True will start an headless server for each test (and stop it after the test). If you want to share the same server across several tests, set xvfb. You’re then responsible for starting and stopping this server (see src/sst/xvfbdisplay.py for details or src/sst/tests/test_xvfb.py for examples.
Inside tests you can import the sst.config module to know various things about the current test environment. The sst.config module has the following information:
Development on Ubuntu/Debian
SST is primarily being developed on Linux, specifically Ubuntu. It should work fine on other platforms, but any issues (or even better - patches) should be reported on the Launchpad project.
Get a copy of SST Trunk, create and activate a virtualenv, install requirements, and run examples/self-tests from the dev branch:
$ sudo apt-get install bzr python-virtualenv xvfb $ bzr branch lp:selenium-simple-test $ cd selenium-simple-test $ virtualenv ENV $ source ENV/bin/activate (ENV)$ pip install -r requirements.txt (ENV)$ ./sst-run -d examples
(optional) Install test dependencies and run SST’s internal unit tests:
(ENV)$ pip install mock nose pep8 (ENV)$ nosetests --match ^test_.* --exclude="ENV|testproject|selftests"
(optional) Install django and run SST’s internal test application with acceptance tests
(ENV)$ pip install django (ENV)$ ./sst-run –test -x
To manually setup dependencies, SST uses the following non-stdlib packages:
- django (optional - needed for internal self-tests only)
Running the examples
SST source code repository and package download contain some trivial example scripts.
You can run them from your local sst directory like this:
$ ./sst-run -d examples
Running the self-tests
SST source code repository and package download contain a set of self-tests based on an included test Django project.
You can run the suite of self-tests (and the test Django server) from your local branch like this:
$ ./sst-run --test