Our team uses Google Test for automated testing. Most of our tests pass consistently, but a few seem to fail ~5% of the time due to race conditions, network time-outs, etc.
We would like the ability to mark certain tests as "flaky". A flaky test would be automatically re-run if it fails the first time, and will only fail the test suite if it fails both times.
Is this something Google Test offers out-of-the-box? If not, is it something that can be built on top of Google Test?
You have several options:
Use --gtest_repeat for the test executable:
The --gtest_repeat flag allows you to repeat all (or selected) test methods in a program many times. Hopefully, a flaky test will eventually fail and give you a chance to debug.
You can mimic tagging your tests by adding "flaky" somewhere in their names, and then use the gtest_filter option to repeat them. Below are some examples from Google documentation:
$ foo_test --gtest_repeat=1000
Repeat foo_test 1000 times and don't stop at failures.
$ foo_test --gtest_repeat=-1
A negative count means repeating forever.
$ foo_test --gtest_repeat=1000 --gtest_break_on_failure
Repeat foo_test 1000 times, stopping at the first failure. This
is especially useful when running under a debugger: when the test
fails, it will drop into the debugger and you can then inspect
variables and stacks.
$ foo_test --gtest_repeat=1000 --gtest_filter=Flaky.*
Repeat the tests whose name matches the filter 1000 times.
See here for more info.
Use bazel to build and run your tests:
Rather than tagging your tests in the test files, you can tag them in the bazel BUILD files.
You can tag each test individually using cc_test rule.
You can also define a set of tests (using test_suite) in the BUILD file and tag them together (e.g. "small", "large", "flaky", etc). See here for an example.
Once you tag your tests, you can use simple commands like this:
% bazel test --test_tag_filters=performance,stress,-flaky //myproject:all
The above command will test all tests in myproject that are tagged as performance,stress, and are not flaky.
See here for documentation.
Using Bazel is probably cleaner because you don't have to modify your test files, and you can quickly modify your tests tags if things change.
See this repo and this video for examples of running tests using bazel.
Related
I test my code with go test ./... -v -short.
Unfortunately, -v only prints out each test as it happens, but does not leave a summary of the results at the bottom like in Java. This means that if any test failed somewhere at the top, I have to scroll up and look for the word FAIL or search for it in a text editor.
The -failfast flag isn't helping either because some of my tests still get printed after the first test failure for some reason.
I don't really care if tests get run after the initial test failure. I just want to be able to easily tell if any test failed, preferably in just one place (e.g. a summary of how many tests passed or failed, or by seeing a flag if all tests passed or not).
Is there a way to easily tell if there was a test failure because I don't want to accidentally continue coding if I still have test failures.
I'm on Windows 10 64-bit.
UPDATE: Many thanks to #icza for the findstr tip. I later realized that I also wanted to see the error descriptions along with the test failures, but did not want to run go test twice. This is what I came up with for CMD (does not work on Powershell):
go test ./... -v -short > test-results.txt & findstr "FAIL _test" test-results.txt
Now findstr should report test failures as well as error descriptions. And if you want to see the full test results, simply open test-results.txt.
Failing tests are indicated with FAIL in the output. So all you have to do is filter the output for that word.
On Unix systems:
go test ./... |grep FAIL
On Windows:
go test ./... |findstr FAIL
Note that this is purely text processing, it doesn't know anything about go tests and their results. This means you might get "false positives" if a test outputs the word FAIL even if it succeeds. But in practice, this pretty much does the job you want.
A more sophisticated and more accurate way to achieve this would be to pass -json flag to go test, so it generates JSON output, which you can process with a program (e.g. written in Go itself). Failing tests are indicated with a JSON object having an "Action":"fail" field, e.g.
{"Time":"2019-03-01T12:06:21.108544405+01:00","Action":"fail",
"Package":"some/package","Test":"TestSomething","Elapsed":0.01}
And even if you don't want to write a program for this, filtering the JSON output leaves less chance for false positive (filtering for "Action":"fail"):
Unix:
go test ./... -json |grep '"Action":"fail"'
Windows:
go test ./... -json |findstr /C:"\"Action\":\"fail\""
I found it painless to install gotestsum and get the neat summary at the end.
go install gotest.tools/gotestsum#latest
gotestsum --format testname # Or dots
An alternative, if you only care about the count is:
go test |grep FAIL |wc -l
I have ancestral project and my Nightly build fails. I can`t find out how to fix it. Problem is about test cases sequence dependent execution.
Environment:
There is used TFS2015 build definitions. Basically, I have a definition with a Visual Studio Build task and then a Visual Studio Test task. Visual Studio Test task is also overwritten with Powershell file and I see that in TFS Nightly build process my predefined VS Ordered Test statements is ignored.
Important
Test cases are sequence dependent (as I said, this project is ancestral).
Problem
Interesting is that build log files always shows test execution in sequence what I define in VS Ordered Test, but at TFS2015 Detailed report Test results sequence always is different. So I can`t find out what affects test case execution procedures in TFS. Also I am not sure how tests are executed - parallel or sequentially (As I see, both TFS and PowerShell has no indications to run test cases in Parallel).
I have 2 questions:
Powershell brake down all in VS defined conditions?
Which is the best way to define test execution order, so that it takes into account?
Actually, those Test method run in the order that you defined in Ordered Test file during TFS build process. The build log already shows the correct sequence.
Just like you mentioned above, in the test result page, the order is the same as what you defined in the Ordered Test, but you can see that in front of each test method, it has an order number. You could download the test result file to check again and you will find those test methods are run in the correct order.
In the higher version, like TFS 2017, you could click the Column title 'Test' to make it sort by order().
You could also add a Date started column to know which test method run the first.
I want to exclude some tests from my continuous integration build but I haven't found a way to do so.
One of the things I've tried was to set up the priority of those tests to -2 and then on the build I specified Minimum Test Priority = -1 but it still run those tests.
Any help would be greatly appreciated.
Instead of using "Test Lists" that have been described, you should use the "Test Category" method. The test lists & VSMDI functionality have actually been deprecated in Visual Studio 2010 and Microsoft may remove the feature completely in a future version of Visual Studio.
If you'd like some more information about how to use test categories especially with your automated build process, check out this blog post: http://www.edsquared.com/2009/09/25/Test+Categories+And+Running+A+Subset+Of+Tests+In+Team+Foundation+Server+2010.aspx
You can also exclude test categories from running by specifying the ! (exclamation point) character in front of the category name to further define your filter.
If you are using MSTest you can create a Test List for the tests that you need in you continuous integration.
With MSTest, you can simply create two test projects (assemblies) and only specify one in the build config to use for testing. In MSBuild, this was the way to go. For the new WF-Based build definitions, I currently don't have a sample at hand:
<ItemGroup>
<!-- TEST ARGUMENTS
If the RunTest property is set to true then the following test arguments will be used to run
tests. Tests can be run by specifying one or more test lists and/or one or more test containers.
To run tests using test lists, add MetaDataFile items and associated TestLists here. Paths can
be server paths or local paths, but server paths relative to the location of this file are highly
recommended:
<MetaDataFile Include="$(BuildProjectFolderPath)/HelloWorld/HelloWorld.vsmdi">
<TestList>BVT1;BVT2</TestList>
</MetaDataFile>
To run tests using test containers, add TestContainer items here:
<TestContainer Include="$(OutDir)\AutomatedBuildTests.dll" />
<TestContainer Include="$(SolutionRoot)\TestProject\WebTest1.webtest" />
<TestContainer Include="$(SolutionRoot)\TestProject\LoadTest1.loadtest" />
Use %2a instead of * and %3f instead of ? to prevent expansion before test assemblies are built
-->
</ItemGroup>
<PropertyGroup>
<RunConfigFile>$(SolutionRoot)\LocalTestRun.testrunconfig</RunConfigFile>
</PropertyGroup>
Tip: To use a generic build definition, we name all our Test projects "AutomatedBuildTests", i.e. there is no solution difference. So the build definition can be included in any existing build definition (or even be a common one) that always executes the right set of tests. It would be an easy task to prepend an "if exists" check in order to allow a build definition to only run tests when a Test assembly is present. We do not use this in order to get build errors when no test assembly is found as we absolutely want test with all those builds that use this definition.
My preference would be as above using a Test List, but some people have issued merging/editing the vsmdi files... We end up with separate solutions and use a pattern match to execute all tests in the appropriate DLL.
In Visual Studio 2012 and later you can configure your build definition using the Test case filter setting.
This setting is part of your build definition.
Open the build definition and navigate to the Process tab. In the section 3. Test you can define mutiple test sources. For each test source your can specify a Test case filter.
You can find the details in this MSDN article: Running selective unit tests in VS 2012 RC using TestCaseFilter
I have copied the supported operators and some examples from this article:
Operators supported in RC are:
1.= (equals)
2.!= (not equals)
3.~ (contains or substring only for string values)
4.& (and)
5.| (or)
6.( ) (paranthesis for grouping)
Expresssion can be created using these operators as any valid logical condition. & (and) has higher
precedence over | (or) while evaluating expression.
E.g.
"TestCategory=NAR|Priority=1"
"Owner=vikram&TestCategory!=UI"
"FullyQualifiedName~NameSpace.Class"
"(TestCategory!=UI&(Priority=1|Priority=2))|(TestCategory=UI&Priority=1)"
Another possibility would be to have some test sources in one build definition in some (i.e. more or fewer) test sources in other build definitions.
I've written several XMLUnit tests (that fit in to the JUnit framework) in groovy and can execute them easily on the command line as per the groovy doco but I don't quite understand what else I've got to do for it to produce the xml output that is needed by Jenkins/Hudson (or other) to display the pass/fail results (like this) and detailed report of the errors etc (like this). (apologies to image owners)
Currently, my kickoff script is this:
def allSuite = new TestSuite('The XSL Tests')
//looking in package xsltests.rail.*
allSuite.addTest(AllTestSuite.suite("xsltests/rail", "*Tests.groovy"))
junit.textui.TestRunner.run(allSuite)
and this produces something like this:
Running all XSL Tests...
....
Time: 4.141
OK (4 tests)
How can I make this create a JUnit test report xml file suitable to be read by Jenkins/Hudson?
Do I need to kick off the tests with a different JUnit runner?
I have seen this answer but would like to avoid having to write my own test report output.
After a little hackage I have taken Eric Wendelin's suggestion and gone with Gradle.
To do this I have moved my groovy unit tests into the requisite directory structure src/test/groovy/, with the supporting resources (input and expected output XML files) going into the /src/test/resources/ directory.
All required libraries have been configured in the build.gradle file, as described (in its entirety) here:
apply plugin: 'groovy'
repositories {
mavenCentral()
}
dependencies {
testCompile group: 'junit', name: 'junit', version: '4.+'
groovy module('org.codehaus.groovy:groovy:1.8.2') {
dependency('asm:asm:3.3.1')
dependency('antlr:antlr:2.7.7')
dependency('xmlunit:xmlunit:1.3')
dependency('xalan:serializer:2.7.1')
dependency('xalan:xalan:2.7.1')
dependency('org.bluestemsoftware.open.maven.tparty:xerces-impl:2.9.0')
dependency('xml-apis:xml-apis:2.0.2')
}
}
test {
jvmArgs '-Xms64m', '-Xmx512m', '-XX:MaxPermSize=128m'
testLogging.showStandardStreams = true //not sure about this one, was in official user guide
outputs.upToDateWhen { false } //makes it run every time even when Gradle thinks it is "Up-To-Date"
}
This applies the Groovy plugin, sets up to use maven to grab the specified dependencies and then adds some extra values to the built-in "test" task.
One extra thing in there is the last line which makes Gradle run all of my tests every time and not just the ones it thinks are new/changed, this makes Jenkins play nicely.
I also created a gradle.properties file to get through the corporate proxy/firewall etc:
systemProp.http.proxyHost=10.xxx.xxx.xxx
systemProp.http.proxyPort=8080
systemProp.http.proxyUser=username
systemProp.http.proxyPassword=passwd
With this, I've created a 'free-style' project in Jenkins that polls our Mercurial repo periodically and whenever anyone commits an updated XSL to the repo all the tests will be run.
One of my original goals was being able to produce the standard Jenkins/Hudson pass/fail graphics and the JUnit reports, which is a success: Pass/Fail with JUnit Reports.
I hope this helps someone else with similar requirements.
I find the fastest way to bootstrap this stuff is with Gradle:
# build.gradle
apply plugin: 'groovy'
task initProjectStructure () << {
project.sourceSets.all*.allSource.sourceTrees.srcDirs.flatten().each { dir ->
dir.mkdirs()
}
}
Then run gradle initProjectStructure and move your source into src/main/groovy and tests to test/main/groovy.
It seems like a lot (really it's <5 minutes of work), but you get lots of stuff for free. Now you can run gradle test and it'll run your tests and produce JUnit XML you can use in build/test-reports in your project directory.
Since you're asking for the purposes of exposing the report to Jenkins/Hudson, I'm assuming you have a Maven/Ant/etc build that you're able to run. If that's true, the solution is simple.
First of all, there's practically no difference between Groovy and Java JUnit tests. So, all you need to do is add the Ant/Maven junit task/plugin to your build and have it execute your Groovy junit tests (just as you'd do if they were written in Java). That execution will create test reports. From there, you can simply configure your Hudson/Jenkins build to look at the directory where the test reports get created during the build process.
You can write your own custom RunListener (or SuiteRunListener). It still requires you to write some code, but it's much cleaner than the script you've provided a link to. If you'd like, I can send you the code for a JUnit reporter I've written in JavaScript for Jasmine and you can 'translate' it into Groovy.
In CPP unit we run unit test as part of build as part of post build setup. We will be running multiple tests as part of this. In case if any test case fails post build should not stop, it should go ahead and run all the test cases and should report summary how many test cases passed and failed. how can we achieve this.
Thanks!
His question is specific enough. You need a test runner. Encapsulate each test in its own behavior and class. The test project is contained separately from the tested code. Afterwards just configure your XMLOutputter. You can find an excellent example of how to do this in the linux website. http://www.yolinux.com/TUTORIALS/CppUnit.html
We use this way to compile our test projects for our main projects and observe if everything is ok. Now it all becomes the work of maintaining your test code.
Your question is too vague for a precise answer. Usually, a unit test engine return a code to tell it has failed (like a non zero return code in the shell on linux) or generate some output file with results. The calling system handle this. If you have written it (some home made scripts) you have to give the option to go on tests execution even if an error occurred. If you are using some tools like continuous integration server, then you have to go through the doc and find the option that allows you to go on when tests fails.
A workaround is to write a script that return a "OK" result even if the unit test fails, but there you lose some automatic verification ...
Be more specific if you want more clues.
my2c
I would just write your tests this way. Instead of using the CPPUNIT_ASSERT macros or whatever you would write them in regular C++ with some way of logging errors.
You could use a macro for this too of course. Something like:
LOGASSERT( some_expression )
could be defined to execute some_expression and to log the expression together with FILE and LINE if it fails, and you can also log exceptions of course, as well as ones that are not thrown, simply by writing them in your tests (with macros if you want to log the expression that caused them with FILE and LINE).
If you are writing macros I would advise you to limit the content of your macro to calling an inline function with extra parameters.