1 [[Back To NumPy | Python9-NumPy]] - [[Forward To Home | Home]]
5 **Presented By Tommy Guy**
7 **Based on materials by Katy Huff and Rachel Slaybaugh**
11 Software testing is a process by which one or more expected behaviors and results from a piece of software are exercised and confirmed. Well chosen tests will confirm expected code behavior for the extreme boundaries of the input domains, output ranges, parametric combinations, and other behavioral edge cases.
16 Unless you write flawless, bug-free, perfectly accurate, fully precise, and predictable code every time, you must test your code in order to trust it enough to answer in the affirmative to at least a few of the following questions:
22 Does it do what you think it does?
24 Does it continue to work after changes are made?
26 Does it continue to work after system configurations or libraries are upgraded?
28 Does it respond properly for a full range of input parameters?
30 What about edge or corner cases?
32 What’s the limit on that input parameter?
37 Verification is the process of asking, “Have we built the software correctly?” That is, is the code bug free, precise, accurate, and repeatable?
41 Validation is the process of asking, “Have we built the right software?” That is, is the code designed in such a way as to produce the answers we’re interested in, data we want, etc.
44 Say we have an averaging function:
53 The test could be runtime exceptions in the function.
62 print "The number list was not a list of numbers."
64 print "There was a problem evaluating the number list."
68 Sometimes they’re alongside the function definitions they’re testing.
77 print "The number list was not a list of numbers."
79 print "There was a problem evaluating the number list."
84 assert(mean([0,0,0,0])==0)
85 assert(mean([0,200])==100)
86 assert(mean([0,-200]) == -100)
87 assert(mean([0]) == 0)
88 def test_floating_mean(self):
89 assert(mean([1,2])==1.5)
91 Sometimes they’re in an executable independent of the main executable.
101 print "The number list was not a list of numbers."
103 print "There was a problem evaluating the number list."
106 Where, in a different file exists a test module:
114 assert(mean([0,0,0,0])==0)
115 assert(mean([0,200])==100)
116 assert(mean([0,-200]) == -100)
117 assert(mean([0]) == 0)
118 def test_floating_mean(self):
119 assert(mean([1,2])==1.5)
122 **When should we test?**
124 The short answer is all the time. The long answer is that testing either before or after your software is written will improve your code, but testing after your program is used for something important is too late.
126 If we have a robust set of tests, we can run them before adding something new and after adding something new. If the tests give the same results (as appropriate), we can have some assurance that we didn’t break anything. The same idea applies to making changes in your system configuration, updating support codes, etc.
128 Another important feature of testing is that it helps you remember what all the parts of your code do. If you are working on a large project over three years and you end up with 200 classes, it may be hard to remember what the widget class does in detail. If you have a test that checks all of the widget’s functionality, you can look at the test to remember what it’s supposed to do.
131 In a collaborative coding environment, where many developers contribute to the same code, developers should be responsible individually for testing the functions they create and collectively for testing the code as a whole.
133 Professionals invariably test their code, and take pride in test coverage, the percent of their functions that they feel confident are comprehensively tested.
135 **How does one test?**
137 The type of tests you’ll write is determined by the testing framework you adopt.
141 Exceptions can be thought of as type of runttime test. They alert the user to exceptional behavior in the code. Often, exceptions are related to functions that depend on input that is unknown at compile time. Checks that occur within the code to handle exceptional behavior that results from this type of input are called Exceptions.
145 Unit tests are a type of test which test the fundametal units a program’s functionality. Often, this is on the class or function level of detail.
147 To test functions and classes, we want to test the interfaces, rather than the implmentation. Treating the implementation as a ‘black box’, we can probe the expected behavior with boundary cases for the inputs.
149 In the case of our fix_missing function, we need to test the expected behavior by providing lines and files that do and do not have missing entries. We should also test the behavior for empty lines and files as well. These are boundary cases.
153 System level tests are intended to test the code as a whole. As opposed to unit tests, system tests ask for the behavior as a whole. This sort of testing involves comparison with other validated codes, analytical solutions, etc.
157 A regression test ensures that new code does change anything. If you change the default answer, for example, or add a new question, you’ll need to make sure that missing entries are still found and fixed.
161 Integration tests query the ability of the code to integrate well with the system configuration and third party libraries and modules. This type of test is essential for codes that depend on libraries which might be updated independently of your code or when your code might be used by a number of users who may have various versions of libraries.
164 Putting a series of unit tests into a suite creates, as you might imagine, a test suite.
166 **Elements of a Test**
170 The behavior you want to test. For example, you might want to test the fun() function.
174 This might be a single number, a range of numbers, a new, fully defined object, a system state, an exception, etc.
176 When we run the fun function, we expect to generate some fun. If we don’t generate any fun, the fun() function should fail its test. Alternatively, if it does create some fun, the fun() function should pass this test.
180 Require that some conditional be true. If the conditional is false, the test fails.
184 Sometimes you have to do some legwork to create the objects that are necessary to run one or many tests. These objects are called fixtures.
186 For example, since fun varies a lot between people, the fun() function is a member function of the Person class. In order to check the fun function, then, we need to create an appropriate Person object on which to run fun().
188 **Setup and teardown**
190 Creating fixtures is often done in a call to a setup function. Deleting them and other cleanup is done in a teardown function.
193 Putting all this together, the testing algorithm is often:
202 But, sometimes it’s the case that your tests change the fixtures. If so, it’s better for the setup and teardown functions to occur on either side of each test. In that case, the testing algorithm should be:
216 ----------------------------------------------------------
218 ----------------------------------------------------------
220 The testing framework we’ll discuss today is called nose, and comes packaged with the enthought python distribution that you’ve installed.
222 **Where is a nose test?**
224 Nose tests are files that begin with Test-, Test_, test-, or test_. Specifically, these satisfy the testMatch regular expression [Tt]est[-_]. (You can also teach nose to find tests by declaring them in the unittest.TestCase subclasses chat you create in your code. You can also create test functions which are not unittest.TestCase subclasses if they are named with the configured testMatch regular expression.)
227 To write a nose test, we make assertions.
231 assert (ShouldBeTrue())
232 assert (not ShouldNotBeTrue())
235 In addition to assertions, in many test frameworks, there are expectations, etc.
237 **Add a test to our work**
239 There are a few tests for the mean function that we listed in this lesson. What are some tests that should fail? Add at least three test cases to this set.
241 *Hint: think about what form your input could take and what you should do to handle it. Also, think about the type of the elements in the list. What should be done if you pass a list of integers? What if you pass a list of strings?*
243 **Test Driven Development**
245 Some people develop code by writing the tests first.
247 If you write your tests comprehensively enough, the expected behaviors that you define in your tests will be the necessary and sufficient set of behaviors your code must perform. Thus, if you write the tests first and program until the tests pass, you will have written exactly enough code to perform the behavior your want and no more. Furthermore, you will have been forced to write your code in a modular enough way to make testing easy now. This will translate into easier testing well into the future.
249 --------------------------------------------------------------------
251 --------------------------------------------------------------------
252 The overlap method takes two rectangles (red and blue) and computes the degree of overlap between them. Save it in overlap.py. A rectangle is defined as a tuple of tuples: ((x_lo,y_lo),(x_hi),(y_hi))
256 def overlap(red, blue):
257 '''Return overlap between two rectangles, or None.'''
259 ((red_lo_x, red_lo_y), (red_hi_x, red_hi_y)) = red
260 ((blue_lo_x, blue_lo_y), (blue_hi_x, blue_hi_y)) = blue
262 if (red_lo_x >= blue_hi_x) or \
263 (red_hi_x <= blue_lo_x) or \
264 (red_lo_y >= blue_hi_x) or \
265 (red_hi_y <= blue_lo_y):
268 lo_x = max(red_lo_x, blue_lo_x)
269 lo_y = max(red_lo_y, blue_lo_y)
270 hi_x = min(red_hi_x, blue_hi_x)
271 hi_y = min(red_hi_y, blue_hi_y)
272 return ((lo_x, lo_y), (hi_x, hi_y))
275 Now let's create a set of tests for this class. Before we do this, let's think about *how* we might test this method. How should it work?
280 from overlap import overlap
282 def test_empty_with_empty():
283 rect = ((0, 0), (0, 0))
284 assert overlap(rect, rect) == None
286 def test_empty_with_unit():
287 empty = ((0, 0), (0, 0))
288 unit = ((0, 0), (1, 1))
289 assert overlap(empty, unit) == None
291 def test_unit_with_unit():
292 unit = ((0, 0), (1, 1))
293 assert overlap(unit, unit) == unit
295 def test_partial_overlap():
296 red = ((0, 3), (2, 5))
297 blue = ((1, 0), (2, 4))
298 assert overlap(red, blue) == ((1, 3), (2, 4))
305 [rguy@infolab-33 ~/TestExample]$ nosetests
307 ======================================================================
308 FAIL: test_overlap.test_partial_overlap
309 ----------------------------------------------------------------------
310 Traceback (most recent call last):
311 File "/usr/lib/python2.6/site-packages/nose/case.py", line 183, in runTest
313 File "/afs/ictp.it/home/r/rguy/TestExample/test_overlap.py", line 19, in test_partial_overlap
314 assert overlap(red, blue) == ((1, 3), (2, 4))
317 ----------------------------------------------------------------------
318 Ran 4 tests in 0.005s
323 Oh no! Something failed. The failure was on line in this test:
327 def test_partial_overlap():
328 red = ((0, 3), (2, 5))
329 blue = ((1, 0), (2, 4))
330 assert overlap(red, blue) == ((1, 3), (2, 4))
333 Can you spot why it failed? Try to fix the method so all tests pass.