=========================== Testing Django applications =========================== **New in Django development version**. Automated testing is an extremely useful weapon in the bug-killing arsenal of the modern developer. When initially writing code, a test suite can be used to validate that code behaves as expected. When refactoring or modifying code, tests serve as a guide to ensure that behavior hasn't changed unexpectedly as a result of the refactor. Testing a web application is a complex task, as there are many components of a web application that must be validated and tested. To help you test your application, Django provides a test execution framework, and range of utilities that can be used to simulate and inspect various facets of a web application. This testing framework is currently under development, and may change slightly before the next official Django release. (That's *no* excuse not to write tests, though!) Writing tests ============= Tests in Django come in two forms: doctests and unit tests. Writing doctests ---------------- Doctests use Python's standard doctest_ module, which searches for tests in your docstrings. Django's test runner looks for doctests in your ``models.py`` file, and executes any that it finds. Django will also search for a file called ``tests.py`` in the application directory (i.e., the directory that holds ``models.py``). If a ``tests.py`` is found, it will also be searched for doctests. .. admonition:: What's a **docstring**? A good explanation of docstrings (and some guidlines for using them effectively) can be found in :PEP:`257`: A docstring is a string literal that occurs as the first statement in a module, function, class, or method definition. Such a docstring becomes the ``__doc__`` special attribute of that object. Since tests often make great documentation, doctest lets you put your tests directly in your docstrings. You can put doctest strings on any object in your ``models.py``, but it's common practice to put application-level doctests in the module docstring, and model-level doctests in the docstring for each model. For example:: from django.db import model class Animal(models.Model): """ An animal that knows how to make noise # Create some animals >>> lion = Animal.objects.create(name="lion", sound="roar") >>> cat = Animal.objects.create(name="cat", sound="meow") # Make 'em speak >>> lion.speak() 'The lion says "roar"' >>> cat.speak() 'The cat says "meow"' """ name = models.CharField(maxlength=20) sound = models.CharField(maxlength=20) def speak(self): return 'The %s says "%s"' % (self.name, self.sound) When you `run your tests`_, the test utility will find this docstring, notice that portions of it look like an interactive Python session, and execute those lines while checking that the results match. For more details about how doctest works, see the `standard library documentation for doctest`_ .. _doctest: http://docs.python.org/lib/module-doctest.html .. _standard library documentation for doctest: doctest_ Writing unittests ----------------- Like doctests, Django's unit tests use a standard library module: unittest_. As with doctests, Django's test runner looks for any unit test cases defined in ``models.py``, or in a ``tests.py`` file stored in the application directory. An equivalent unittest test case for the above example would look like:: import unittest from myapp.models import Animal class AnimalTestCase(unittest.TestCase): def setUp(self): self.lion = Animal.objects.create(name="lion", sound="roar") self.cat = Animal.objects.create(name="cat", sound="meow") def testSpeaking(self): self.assertEquals(self.lion.speak(), 'The lion says "roar"') self.assertEquals(self.cat.speak(), 'The cat says "meow"') When you `run your tests`_, the test utility will find all the test cases (that is, subclasses of ``unittest.TestCase``) in ``models.py`` and ``tests.py``, automatically build a test suite out of those test cases, and run that suite. For more details about ``unittest``, see the `standard library unittest documentation`_. .. _unittest: http://docs.python.org/lib/module-unittest.html .. _standard library unittest documentation: unittest_ .. _run your tests: `Running tests`_ Which should I use? ------------------- Choosing a test framework is often contentious, so Django simply supports both of the standard Python test frameworks. Choosing one is up to each developer's personal tastes; each is supported equally. Since each test system has different benefits, the best approach is probably to use both together, picking the test system to match the type of tests you need to write. For developers new to testing, however, this choice can seem confusing, so here are a few key differences to help you decide whether doctests or unit tests are right for you. If you've been using Python for a while, ``doctest`` will probably feel more "pythonic". It's designed to make writing tests as easy as possible, so there's no overhead of writing classes or methods; you simply put tests in docstrings. This gives the added advantage of given your modules automatic documentation -- well-written doctests can kill both the documentation and the testing bird with a single stone. For developers just getting started with testing, using doctests will probably get you started faster. The ``unittest`` framework will probably feel very familiar to developers coming from Java. Since ``unittest`` is inspired by Java's JUnit, if you've used testing frameworks in other languages that similarly were inspired by JUnit, ``unittest`` should also feel pretty familiar. Since ``unittest`` is organized around classes and methods, if you need to write a bunch of tests that all share similar code, you can easily use subclass to abstract common tasks; this makes test code shorter and cleaner. There's also support for explicit setup and/or cleanup routines, which give you a high level of control over the environment your test cases run in. Again, remember that you can use both systems side-by-side (even in the same app). In the end, most projects will eventually end up using both; each shines in different circumstances. Testing Tools ============= To assist in testing various features of your application, Django provides tools that can be used to establish tests and test conditions. * `Test Client`_ * Fixtures_ Test Client ----------- The Test Client is a simple dummy browser. It allows you to simulate GET and POST requests on a URL, and observe the response that is received. This allows you to test that the correct view is executed for a given URL, and that the view constructs the correct response. As the response is generated, the Test Client gathers details on the Template and Context objects that were used to generate the response. These Templates and Contexts are then provided as part of the response, and can be used as test conditions. .. admonition:: Test Client vs Browser Automation? The Test Client is not intended as a replacement for Twill_, Selenium_, or other browser automation frameworks - it is intended to allow testing of the contexts and templates produced by a view, rather than the HTML rendered to the end-user. A comprehensive test suite should use a combination of both: Test Client tests to establish that the correct view is being called and that the view is collecting the correct context data, and Browser Automation tests to check that user interface behaves as expected. .. _Twill: http://twill.idyll.org/ .. _Selenium: http://www.openqa.org/selenium/ Making requests ~~~~~~~~~~~~~~~ Creating an instance of ``Client`` (``django.test.client.Client``) requires no arguments at time of construction. Once constructed, the following methods can be invoked on the ``Client`` instance. ``get(path, data={})`` Make a GET request on the provided ``path``. The key-value pairs in the data dictionary will be used to create a GET data payload. For example:: c = Client() c.get('/customers/details/', {'name':'fred', 'age':7}) will result in the evaluation of a GET request equivalent to:: http://yoursite.com/customers/details/?name=fred&age=7 ``post(path, data={}, content_type=MULTIPART_CONTENT)`` Make a POST request on the provided ``path``. If you provide a content type (e.g., ``text/xml`` for an XML payload), the contents of ``data`` will be sent as-is in the POST request, using the content type in the HTTP ``Content-Type`` header. If you do not provide a value for ``content_type``, the values in ``data`` will be transmitted with a content type of ``multipart/form-data``. The key-value pairs in the data dictionary will be encoded as a multipart message and used to create the POST data payload. Submitting files is a special case. To POST a file, you need only provide the file field name as a key, and a file handle to the file you wish to upload as a value. The Test Client will populate the two POST fields (i.e., ``field`` and ``field_file``) required by Django's FileField. For example:: c = Client() f = open('wishlist.doc') c.post('/customers/wishes/', {'name':'fred', 'attachment':f}) f.close() will result in the evaluation of a POST request on ``/customers/wishes/``, with a POST dictionary that contains `name`, `attachment` (containing the file name), and `attachment_file` (containing the file data). Note that you need to manually close the file after it has been provided to the POST. ``login(path, username, password)`` In a production site, it is likely that some views will be protected with the @login_required decorator provided by ``django.contrib.auth``. Interacting with a URL that has been login protected is a slightly complex operation, so the Test Client provides a simple method to automate the login process. A call to ``login()`` stimulates the series of GET and POST calls required to log a user into a @login_required protected view. If login is possible, the final return value of ``login()`` is the response that is generated by issuing a GET request on the protected URL. If login is not possible, ``login()`` returns False. Note that since the test suite will be executed using the test database, which contains no users by default. As a result, logins for your production site will not work. You will need to create users as part of the test suite to be able to test logins to your application. Testing Responses ~~~~~~~~~~~~~~~~~ The ``get()``, ``post()`` and ``login()`` methods all return a Response object. This Response object has the following properties that can be used for testing purposes: =============== ========================================================== Property Description =============== ========================================================== ``status_code`` The HTTP status of the response. See RFC2616_ for a full list of HTTP status codes. ``content`` The body of the response. The is the final page content as rendered by the view, or any error message (such as the URL for a 302 redirect). ``template`` The Template instance that was used to render the final content. Testing ``template.name`` can be particularly useful; if the template was loaded from a file, ``template.name`` will be the file name that was loaded. If multiple templates were rendered, (e.g., if one template includes another template),``template`` will be a list of Template objects, in the order in which they were rendered. ``context`` The Context that was used to render the template that produced the response content. As with ``template``, if multiple templates were rendered ``context`` will be a list of Context objects, stored in the order in which they were rendered. =============== ========================================================== .. _RFC2616: http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html Exceptions ~~~~~~~~~~ If you point the Test Client at a view that raises an exception, that exception will be visible in the test case. You can then use a standard ``try...catch`` block, or ``unittest.TestCase.assertRaises()`` to test for exceptions. The only exceptions that are not visible in a Test Case are ``Http404``, ``PermissionDenied`` and ``SystemExit``. Django catches these exceptions internally and converts them into the appropriate HTTP responses codes. Persistent state ~~~~~~~~~~~~~~~~ The Test Client is stateful; if a cookie is returned as part of a response, that cookie is provided as part of the next request issued by that Client instance. Expiry policies for these cookies are not followed; if you want a cookie to expire, either delete it manually or create a new Client instance (which will effectively delete all cookies). There are two properties of the Test Client which are used to store persistent state information. If necessary, these properties can be interrogated as part of a test condition. =============== ========================================================== Property Description =============== ========================================================== ``cookies`` A Python ``SimpleCookie`` object, containing the current values of all the client cookies. ``session`` A dictionary-like object containing session information. See the `session documentation`_ for full details. =============== ========================================================== .. _`session documentation`: ../sessions/ Example ~~~~~~~ The following is a simple unit test using the Test Client:: import unittest from django.test.client import Client class SimpleTest(unittest.TestCase): def setUp(self): # Every test needs a client self.client = Client() def test_details(self): # Issue a GET request response = self.client.get('/customer/details/') # Check that the respose is 200 OK self.failUnlessEqual(response.status_code, 200) # Check that the rendered context contains 5 customers self.failUnlessEqual(len(response.context['customers']), 5) Fixtures -------- A test case for a database-backed website isn't much use if there isn't any data in the database. To make it easy to put test data into the database, Django provides a fixtures framework. A *Fixture* is a collection of files that contain the serialized contents of the database. Each fixture has a unique name; however, the files that comprise the fixture can be distributed over multiple directories, in multiple applications. .. note:: If you have synchronized a Django project, you have already experienced the use of one fixture -- the ``initial_data`` fixture. Every time you synchronize the database, Django installs the ``initial_data`` fixture. This provides a mechanism to populate a new database with any initial data (such as a default set of categories). Fixtures with other names can be installed manually using ``django-admin.py loaddata``. However, for the purposes of unit testing, each test must be able to guarantee the contents of the database at the start of each and every test. To do this, Django provides a TestCase baseclass that can integrate with fixtures. Moving from a normal unittest TestCase to a Django TestCase is easy - just change the base class of your test, and define a list of fixtures to be used. For example, the test case from `Writing unittests`_ would look like:: from django.test import TestCase from myapp.models import Animal class AnimalTestCase(TestCase): fixtures = ['mammals.json', 'birds'] def setUp(self): # test definitions as before At the start of each test vase, before ``setUp()`` is run, Django will flush the database, returning the database the state it was in directly after ``syncdb`` was called. Then, all the named fixtures are installed. In this example, any JSON fixture called ``mammals``, and any fixture named ``birds`` will be installed. See the documentation on `loading fixtures`_ for more details on defining and installing fixtures. .. _`loading fixtures`: ../django-admin/#loaddata-fixture-fixture This flush/load procedure is repeated for each test in the test case, so you can be certain that the outcome of a test will not be affected by another test, or the order of test execution. Running tests ============= Run your tests using your project's ``manage.py`` utility:: $ ./manage.py test If you only want to run tests for a particular application, add the application name to the command line. For example, if your ``INSTALLED_APPS`` contains ``myproject.polls`` and ``myproject.animals``, but you only want to run the animals unit tests, run:: $ ./manage.py test animals When you run your tests, you'll see a bunch of text flow by as the test database is created and models are initialized. This test database is created from scratch every time you run your tests. By default, the test database gets its name by prepending ``test_`` to the database name specified by the ``DATABASE_NAME`` setting; all other database settings will the same as they would be for the project normally. If you wish to use a name other than the default for the test database, you can use the ``TEST_DATABASE_NAME`` setting to provide a name. Once the test database has been established, Django will run your tests. If everything goes well, at the end you'll see:: ---------------------------------------------------------------------- Ran 22 tests in 0.221s OK If there are test failures, however, you'll see full details about what tests failed:: ====================================================================== FAIL: Doctest: ellington.core.throttle.models ---------------------------------------------------------------------- Traceback (most recent call last): File "/dev/django/test/doctest.py", line 2153, in runTest raise self.failureException(self.format_failure(new.getvalue())) AssertionError: Failed doctest test for myapp.models File "/dev/myapp/models.py", line 0, in models ---------------------------------------------------------------------- File "/dev/myapp/models.py", line 14, in myapp.models Failed example: throttle.check("actor A", "action one", limit=2, hours=1) Expected: True Got: False ---------------------------------------------------------------------- Ran 2 tests in 0.048s FAILED (failures=1) The return code for the script will indicate the number of tests that failed. Regardless of whether the tests pass or fail, the test database is destroyed when all the tests have been executed. Using a different testing framework =================================== Doctest and Unittest are not the only Python testing frameworks. While Django doesn't provide explicit support these alternative frameworks, it does provide a mechanism to allow you to invoke tests constructed for an alternative framework as if they were normal Django tests. When you run ``./manage.py test``, Django looks at the ``TEST_RUNNER`` setting to determine what to do. By default, ``TEST_RUNNER`` points to ``django.test.simple.run_tests``. This method defines the default Django testing behavior. This behavior involves: #. Performing global pre-test setup #. Creating the test database #. Running ``syncdb`` to install models and initial data into the test database #. Looking for Unit Tests and Doctests in ``models.py`` and ``tests.py`` file for each installed application #. Running the Unit Tests and Doctests that are found #. Destroying the test database #. Performing global post-test teardown If you define your own test runner method and point ``TEST_RUNNER`` at that method, Django will execute your test runner whenever you run ``./manage.py test``. In this way, it is possible to use any test framework that can be executed from Python code. Defining a test runner ---------------------- By convention, a test runner should be called ``run_tests``; however, you can call it anything you want. The only requirement is that it accept two arguments: ``run_tests(module_list, verbosity=1)`` The module list is the list of Python modules that contain the models to be tested. This is the same format returned by ``django.db.models.get_apps()`` Verbosity determines the amount of notification and debug information that will be printed to the console; `0` is no output, `1` is normal output, and `2` is verbose output. This method should return the number of tests that failed. Testing utilities ----------------- To assist in the creation of your own test runner, Django provides a number of utility methods in the ``django.test.utils`` module. ``setup_test_environment()`` Performs any global pre-test setup, such as the installing the instrumentation of the template rendering system. ``teardown_test_environment()`` Performs any global post-test teardown, such as removing the instrumentation of the template rendering system. ``create_test_db(verbosity=1, autoclobber=False)`` Creates a new test database, and run ``syncdb`` against it. ``verbosity`` has the same behavior as in the test runner. ``Autoclobber`` describes the behavior that will occur if a database with the same name as the test database is discovered. If ``autoclobber`` is False, the user will be asked to approve destroying the existing database. ``sys.exit`` is called if the user does not approve. If autoclobber is ``True``, the database will be destroyed without consulting the user. ``create_test_db()`` has the side effect of modifying ``settings.DATABASE_NAME`` to match the name of the test database. ``destroy_test_db(old_database_name, verbosity=1)`` Destroys the database with the name ``settings.DATABASE_NAME`` matching, and restores the value of ``settings.DATABASE_NAME`` to the provided name. ``verbosity`` has the same behavior as in the test runner.