I've created a small (test) addition to the Dolphin Smalltalk framework
that I want to submit on GitHub later. (1 method: Integer>>isPrime)
But first, I want add my testing method of this method to the standard regression test set, with ~ 2400 tests now. (IntegerTest>>testIsPrime)
I've found the classes TestCase, DolphinTest, IntegerTest and the SUnit browser.
But I didn't find out how to add my test to the standard test set.
Can someone point me the right direction?
I assume you are working from a Git checkout and have the test classes in your image. From there the easiest thing is to modify an existing class (such as IntegerTest) in the code browser, save the package back to the file system, and then Git should show the files as modified.
The neat thing about SUnit is that by default it will include all methods that start with 'test' in the test suite. So just add the test, run the suite, and see the number of tests increase by one!
Related
Currently, I define the following function in the REPL at the start of a coding session:
(defn rt []
(let [tns 'my.namespace-test]
(use tns :reload-all)
(cojure.test/test-ns tns)))
And everytime I make a change I rerun the tests:
user=>(rt)
That been working moderately well for me. When I remove a test, I have to restart the REPL and redefine the method which is a little annoying. Also I've heard bad rumblings about using the use function like this. So my questions are:
Is using use this way going to cause me a problem down the line?
Is there a more idiomatic workflow than what I'm currently doing?
most people run
lein test
form a different terminal. Which guarantees that what is in the files is what is tested not what is in your memory. Using reload-all can lead to false passes if you have changed a function name and are still calling the old name somewhere.
calling use like that is not a problem in it's self, it just constrains you to not have any name conflicts if you use more namespaces in your tests. So long as you have one, it's ok.
using lein lets you specify unit and integration tests and easily run them in groups using the test-selectors feature.
I also run tests in my REPL. I like doing this because I have more control over the tests and it's faster due to the JVM already running. However, like you said, it's easy to get in trouble. In order to clean things up, I suggest taking a look at tools.namespace.
In particular, you can use clojure.tools.namespace.repl/refresh to reload files that have changed in your live REPL. There's alsorefresh-all to reload all the files on the classpath.
I add tools.namespace to my :dev profile in my ~/.lein/profiles.clj so that I have it there for every project. Then when you run lein repl, it will be included on the classpath, but it wont leak into your project's proper dependencies.
Another thing I'll do when I'm working on a test is to require it into my REPL and run it manually. A test is just a no-argument function, so you can invoke them as such.
I am so far impressed with lein-midje
$ lein midje :autotest
Starts a clojure process watching src and test files, reloads the associated namespaces and runs the tests relevant to the changed file (tracking dependencies). I use it with VimShell to open a split buffer in vim and have both the source and the test file open as well. I write a change to either one and the (relevant) tests are executed in the split pane.
In my Mac Cocoa unit tests, I would like to output some files as part of the testing process, and delete them when the test is done, but only when there are no failures. How can this be done (and/or what's the cleanest way to do so)?
Your question made me curious so I looked into it!
I guess I would override the failWithException: method in the class SenTestCase (the class your tests run in inherits from this), and set a "keep output files" flag or something before calling the super's method.
Here's what SenTestCase.h says about that method:
/*"Failing a test, used by all macros"*/
- (void) failWithException:(NSException *) anException;
So, provided you only use the SenTest macros to test and/or fail (and chances are this is true in your case), that should cover any test failure.
I've never dug into the scripts for this, but it seems like you could customize how you call the script that actually runs your tests to do this. In Xcode 4, look at the last step in the Build Phases tab of your test target. Mine contains this:
# Run the unit tests in this test bundle.
"${SYSTEM_DEVELOPER_DIR}/Tools/RunUnitTests"
I haven't pored through the contents of this script or the many subscripts it pulls in on my machine, but presumably they call otest or some other test rig executable and the test results would be returned to that script. After a little time familiarizing yourself with those scripts you would likely be able to find a straightforward way to conditionally remove the output files based on the test results.
I have been developing a project which contains a TestLauncher class that'll read a given directory and for each file it contains, run it against my tool and yield the results.
So, when coding in Eclipse, it would show up one result for each test (as expected). Today I've been toying with Intellij, and I've decided to try to run and code a bit of this project in Intellij.
When trying to run the tests, though, it seems to be only showing up 2 results instead of the 100+ it should. Although I am sure it is running the full suite, it seems to be folding all the results of a given category in a single result. That means that if I have at least one failing test in each category, it shows up as a "failed test".
I guess this must not be a bug, but rather some configuration that I am not aware about and that is on by default in Intellij but not in Eclipse. Could anyone explain what might be going on?
Edit: I am using the latest Intellij (downloaded one of these days).
Thanks
What you're seeing is simply a difference in the way the Eclipse and IDEA plug-ins are implemented. I implemented the Eclipse plug-in to be pretty clever in its display, so it will show different things depending on various factors such as the presence of a toString() method in your test class or whether your test class implements org.testng.ITest.
I suggest you ask this question on the IDEA forums and if you don't get any response, feel free to email the testng-users list and I can put you in touch with the JetBrains engineer in charge of the TestNG plug-in.
The IntelliJ-IDEA TestNG Plugin has a filter symbol called "Hide Passed" above the output Test Results. You can toggle that to display all tests, including the passed ones.
For our webapp testing environment we're currently using watin with a bunch of unit tests, and we're looking to move to selenium and use more frameworks.
We're currently looking at Selenium2 + Gallio + Xunit.net,
However one of the things we're really looking to get around is compiled testcases. Ideally we want testcases that can be edited in VS with intellisense, but don't require re-compilling the assembly every single time we make a small change,
Are there any frameworks likely to help with this issue?
Are there any nice UI tools to help manage massive ammount of testcases?
Ideally we want the testcase writing process to be simple so that more testers can aid in writing them.
cheers
You can write them in a language like ruby (e.g., IronRuby) or python which doesnt have an explicit compile step of such a manner.
If you're using a compiled a compiled language, it needs to be compiled. Make the assemblies a reasonable size and a quick Shift F6 (I rewire it to shift Ins) will compile your current project. (Shift Ctrl-B will typically do lots of redundant stuff). Then get NUnit to auto-re-run the tests when it detects the assembly change (or go vote on http://xunit.codeplex.com/workitem/8832 and get it into the xunit GUI runner).
You may also find that CR, R# and/or TD.NET have stuff to offer you in speeding up your flow. e.g., I believe CR detects which tests have changed and does stuff around that (at the moment it doesnt support the more advanced xunit.net testing styles so I dont use it day to day)
You wont get around compiling test frameworks if you add new tests..
However there are a few possibilities.
First:
You could develop a native language like i did in xml or similar format. It would look something like this:
[code]
action name="OpenProfile"
parameter name="Username" value="TestUser"
[/code]
After you have this your could simply take an interpreter and serialize this xml into an object. Then with reflection you could call the appropriate function in the corresponding class. After you have a lot of actions implemented of course perfectly moduled and carefully designed structure ( like every page has its own object and a base object that every page inherits from ), you will be able to add xml based tests on your own without the need of rebuilding the framework it self.
You see you have actions like, login, go to profile, go to edit profile, change password, save, check email etcetc. Then you could have tests like: login change password, login edit profile username... and so on and so fort. And you only would be creating new xmls.
You could look for frameworks supporting similar behavior and there are a few out there. The best of them are cucumber and fitnesse. These all support high level test case writing and low level functionality building.
So basically once you have your framework ready all your have to do is writing tests.
Hope that helped.
Gergely.
I have a suite of unit tests that I use before checking in my project. However, very often it's the case that only one of them finds some regression in the code. In these cases I'd like to only run that particular unit test while debugging the failure. I haven't found any way to do this in Xcode. Is it possible?
If you're happy restricting your testing to a single test class, a simple option is to create a second test target (duplicate the existing target, change the product name and remove the contents of the "Compile Sources" build phase, if you wish) and add only the test source file you're trying to fix to it.
Alternatively, you can use the "Other Test Flags" option to pass a -SenTest argument to otest, the test runner:
% /Developer/Tools/otest
2009-08-29 22:28:39.555 otest[70089:10b] Usage: otest [-SenTest Self | All | None |
<TestCaseClassName/testMethodName>] <path of unit to be tested>
More information about using this method is here.
Thanks for that push in the right direction. I ended using the same basic concept, but I added a GUI that lets you select what gets run as well as get a nice red/green status for each test. If anyone is interested, the code is at the URL below. The UI needs to more spit and polish, but it seems to be working.
http://github.com/nall/XcodeUnitTestGUI/tree/master
After I started the project above, I found this project which is really fantastic.
http://github.com/gabriel/gh-unit
For new readers: A much better way now available in Xcode is to edit the scheme for the target to be tested and select "Test" in the left hand column of the scheme pane.
Use the widgets in the Tests column to expand targets and suites.
You can disable/enable tests on a per test target, per suite or per test basis using the check boxes on the right