I'm using boost::test with Visual Studio 2010, in a Win32 console project. boost::test appears to take care of the main() function, by providing its own. It looks like I don't have to provide a main().
Thing is, I want my console output not to disappear immediately, but this means that I'd have to define the main() function somehow, to get it to stay on screen. Thing is any attempts I make to redefine main() appear to get ignored.
I know that Cppunit can output test results during compilation - does boost::test have anything like this?
I use this configuration: In one solution I have a main project and a test project; the main project depends on the test project; the test project has a post-build event with $(TargetPath) which runs the test project and redirects the output to the output-window. Test errors can be located by double-clicking on the corresponding line in the output window.
Ctrl+F5 - start without debugging. This will keep the output window open.
You might try out Gallio in combination with the Boost test adapter for Gallio from Alexander Tsvyashchenko. The interesting part is that you will benefit all the goodies of Gallio (test output, filtering, reports, and integration with many 3th party tools such as CC.Net, NAnt, etc.)
Here is described how to setup your project to get what you want.
Related
We've just switched to using google test for unit testing due to deficiencies in our previous framework.
In Visual Studio, I have a solution with multiple projects. Each project contains unit tests using google test for a specific class (or collection of linked classes) within our application. We only use VS for managing and running google tests and reference the test classes found elsewhere.
When I want to run all tests, I set startup projects to multiple, then press ctrl-f5. This pops ups 1 console window per project with the standard google test output. This is great! I have to click a key to close each of the output windows, though. It sure would be nice to integrate into "Test Explorer" within VS2013 express.
I've found this script runner:
https://visualstudiogallery.msdn.microsoft.com/9dd47c21-97a6-4369-b326-c562678066f0.
It appears to be exactly what I want, but apparently express editions don't allow extensions? (I'm new the land of visual studio.) Anyway, we're using express instead of a different version because we're commercial. My fallback solution is spending the money on a full version of VS. However, I'd much rather find a free solution so I don't have to justify purchase, then deal with licensing visual studio for every new person that joins the group.
Has anyone figured out a solution to this?
Our team also tried using "Google Test Runner" extension. The
integration with VS Test Explorer was great but we did not like
the fact that we could not see gmock warnings. This means that
some tests can pass with warnings and you will simply let that
slip by and create problems in the future. The same problem
exists when trying to use XML output generated by gmock.
Since viewing output in console is a poor option, we've decided to
parse the output of executable that runs the tests and display them
in a browser. To this end, we've modified the project so that it directs
its output to a file ( "> test_output.txt" in Debugging -> Command
Arguments) and modified the main function to parse the output,
insert it into a html file template, save this html file and then open
it in default browser using:
ShellExecute(NULL, NULL, "test_output.html", "", "", SW_SHOW);
This also enabled us to make some customizations to the way results
are displayed using javascript.
You can use this approach in you case by creating a simple project which
will run last in the list of multiple startup projects, parse the output
of other executables and display it in a browser. I know this solution has
its drawbacks, but it is (in my opinion) better than ignoring warnings.
I created a test framework to test a sample application. This basically required creating a Solution for my sample application and then adding a Native Test Project, which can test the application.
The test project generates an .exe, but if I were to execute the .exe on it's own, it does nothing.
Is there a way to get information from the test programme so it can be used for other purposes outside of the VS environment?
Yes, you can run unit tests without VS2013.
Use Run automated tests from the command line using MSTest source to do so.
In general it says:
Open a Visual Studio command prompt.
Either change directory to your solution folder or, when you run the MSTest.exe program in step 3, specify a full or relative path to the metadata file or to the test container.
Run the MSTest.exe program.
I hope it will solve your issue.
It took a couple of days, but I've found that DeJaVo's answer is incorrect. Because I am running a Native Test Project, MSTest will not work with it. To run a Native Test from the command line, use VSTest.Console.
I've fixed a null reference problem. Now when I run the test related to that class, it fails. I'd like to know whether is possible to run a test in debug mode so I'll know why I am getting the exception?
The test was written by some else, it's calling a method, as part of the setup, with a lot of logic in it. I'd like to step in to see how values are being returned.
Thank for helping.
It is indeed possible, using the NUnit Test Adapter for Visual Studio 2012.
Also, just so you're aware: failed assertions will throw exceptions, causing the debugger to break execution. You can adjust your configuration to work around this, as described in this thread:
Debugging in VS2012 with NUnit without throwing exceptions?
You can also use ReSharper, which makes it pretty easy. Or, you can just go into the properties for your DLL project that has the test in it and make the test runner you are using the exe that runs when you run the project.
Right click the project in the solution explorer, select "Properties" from the menu, go to the "Debug" tab, click the "Start external program" radio button then set the project as the active project and run in debug mode.
There is already a thread here that partially answers my question .
On Eclipse 3.7.2 I followed the approach provided there and I could successfully accomplish the steps creating and setting up a new error parser and adding it to my current project. After executing my Boost.Test (boost rel. 1.48.0) Unit Test, on the Eclipse console I get the same output as the output I get when no parsing is done (e.g. when executing the Unit Test outside Eclipse (e.g. on a Linux terminal)). I searched for a new Eclipse console where the parsed Unit Test output could be displayed (similar to the consoles by e.g. gcov, gprof or cppcheck in Eclipse) but found nothing alike.
Where should the parsed unit test output be displayed? In case the parsed output shall be displayed in the Eclipse Console view, are there any suggestions what might have gone wrong with the parsing in my case?
Thanks in advance.
P.S.: Thanks to casperOne and kleopatra for teaching me manners.
If you're using Eclipse there is also a really nice plugin called TestRunner for running CDT unit tests much like Java or Python unit tests. It handles the unit-testing parsing for you. You'll get a separate console window for your unit tests and it arranges them with level filters. Clicking on an error/warning takes you to the unit-test line number in your editor. You also can set the verbosity level as well as a few other settings from inside Eclipse.
You can directly install the plugin by using the following link in Eclipse updates https://raw.github.com/xgsa/cdt-tests-runner/tests_runner_demo/testsrunner/org.eclipse.cdt.testsrunner-updatesite/site.xml
In order to eclipse parse errors, the error must be print in the console view at compile time. If you are emitting errors in runtime, you must add your program to be called by the Makefile.
That's how I do in embedded systems.
I've just started using resharper and I found a very annoying thing about it. When debugging a unit test, I try to step-into (F11) a method, but visual studio complains that source code is not available. The problem is that resharper is wrapping method calls with its own classes. Of course I can put a braeakpoint further in my source, but this is very annoying. Do you know if there´s a solution for this? By the way, I'm using NUnit to write my tests
Thanks
Federico
After some playing around I managed to recreate this on my setup (Xunit.net in resharper, visual studio 2008)... The steps I took to recreate this are:
Set the dll that contains the tests as the startup project (I know you don't need to do this, was just trying to get it to fail)
Put a breakpoint in the unit test
Push F11 (the shortcut key for 'step into' in my configuration)
This complains 'Cannot start test project 'RowanBeach.Crm.Domain.Test' because the project does not contain any tests.'
Run the 'Debug Unit Tests' command from the Resharper menu (I have this bound to a key combination on my machine)
This displays the 'No source code available...' message
Of course that's not how you should start unit tests from resharper! :) If that's what you were doing (or something similar), try this instead:
Don't bother setting the dll as the startup project - it doesn't need to be
build or rebuild the dll that contains the tests
set the breakpoint
Put the cursor somewhere in the source code for the unit test you want to debug to set it as the 'current context' unit test
Run the 'Debug Unit Tests' command from the Resharper menu (You might want to bind this to a key combination if you haven't already)
Hopefully, that should work
What type of tests are you running - NUnit? MS Test? Something else?
I've never found it difficult to debug tests using R# - just put a breakpoint in your test and go. What method are you trying to step into? I've generally not tried to step into the NUnit methods themselves (assertions etc) but stepping into your own code should be okay.
If you can come up with a quick example, I'd be happy to try it myself.