Updating testing documentation for plugin

This commit is contained in:
Yoann Roman
2011-04-18 12:32:57 -04:00
parent 21f342047b
commit 4217446484

View File

@@ -17,7 +17,7 @@ Structure
---------
This guide assumes that you have all your tests in a ``tests`` directory. If
you have created a project from the ``base`` project template that Pecan
provides you should already have this directory with a few tests.
provides, you should already have this directory with a few tests.
The template project uses UnitTest-type tests and some of those tests use
WebTest. We will describe how they work in the next section.
@@ -37,34 +37,54 @@ This is how running those tests with ``py.test`` would look like::
Configuration and Testing
-------------------------
When running tests, you would want to avoid as much as possible setting up test
cases by creating a Pecan app on each instance. To avoid this, you need to
create a proper test configuration file and load it at setup time.
When you create a new project using the ``base`` project template, Pecan adds
a reference to its ``py.test`` plugin to your project's ``setup.cfg`` file.
This handles loading your Pecan configuration and setting up your app as
defined by your project's ``app.py`` file.
To do this, you need to know the absolute path for your configuration file and
then call ``set_config`` with it. A typical ``setUp`` method would look like::
If you've created your own project without using Pecan's template, you can
load the plugin yourself by adding this to your ``setup.cfg`` file::
def setUp(self):
config_path = '/path/to/test_config.py'
pecan.set_config(config_path)
[pytest]
addopts = -p pecan.testing --with-config=./config.py
self.app = TestApp(
make_app(
config.app.root
template_path = config.app.template_path
)
)
Alternatively, you can just pass those options to ``py.test`` directly.
As you can see, we are loading the configuration file into Pecan first and then
creating a Pecan application with it. Any interaction after ``setUp`` will be
exactly as if your application was really running via an HTTP server.
By default, Pecan's testing plugin assumes you will be using the ``config.py``
configuration file to run your tests. To change which configuration file gets
used once, run ``py.test`` with the `--with-config` option. To make the change
permanent, modify that option in the `addopts` setting of your ``setup.cfg``
file.
Pecan's ``py.test`` plugin exposes two new variables in the ``py.test``
namespace: ``temp_dir`` and ``wsgi_app``.
``py.test.temp_dir`` is a temporary directory that you can use for your tests.
It's created at startup and deleted after all tests have completed. When using
locally distributed testing with py.test, this is guaranteed to be shared by
each test process. This is useful if you need to create some initial resource
(e.g., a database template) that is later copied by each test. If you're using
remotely distributed testing, the directory won't be shared across nodes.
``py.test.wsgi_app`` is your Pecan app loaded and configured per your project's
``app.py`` file. In your test's ``setUp`` method, you would wrap this with
``TestApp``::
from unittest import TestCase
from webtest import TestApp
import py.test
class TestRootController(TestCase):
def setUp(self):
self.app = TestApp(py.test.wsgi_app)
Using WebTest with a UnitTest
-----------------------------
Once you have a ``setUp`` method with your Pecan configuration loaded you have
a wealth of actions provided within the test class to interact with your Pecan
Once you have a ``setUp`` method with your ``TestApp`` created, you have a
wealth of actions provided within the test class to interact with your Pecan
application::
* POST => self.app.post
@@ -72,19 +92,18 @@ application::
* DELETE => self.app.delete
* PUT => self.app.put
For example, if I wanted to assert that I can get the root of my application,
I would probably do something similar to this::
For example, if you want to assert that you can get to the root of your
application, you could do something similar to this::
response = self.app.get('/')
assert response.status_int == 200
If you are expecting error responses from your application, you should make
sure that you pass the `expect_errors` flag and set it to True::
If you are expecting error responses from your application, make sure to pass
`expect_errors=True`::
response = self.app.get('/url/does/not/exist', expect_errors=True)
assert response.status_int == 404
If you would like to dig in to more examples in how to test and verify more
actions, make sure you take a look at the
actions, take a look at the
`WebTest documentation <http://pythonpaste.org/webtest/>`_