Tag Archives: catch

Deterministic Testing of Concurrent Behavior in RxCpp

A Retrospective

After getting inspired by The Reactive Manifesto, it is hard not to get excited about Reactive Extensions. Such excitement has lead to a series of hello-world articles and some code examples. While Reactive Extensions take over the programming world in C#, Java and JavaScript, it seems, the world of C++ is slow to adopt RxCpp.

The new ReactiveX Tutorial link list is a great place to start learning and grokking. This article is an attempt to bring RxCpp closer to C++ developers who might not see yet, how a reactive programming model might help writing better, more robust code.

Testing concurrency with RxCpp

A previous article showed how to test ViewModels in C# by parameterizing the ViewModels with a scheduler. In a UI setting, the scheduler usually involves some kind of synchronization with the GUI thread. Testing keystrokes arriving at certain speed would require some effort to simulate events, probably leading to brittle tests. With the scheduler abstraction, the concurrent behavior of a component is decoupled from physical time, and thus can be tested repeatedly and very fast. This was the C# test:

(new TestScheduler()).With(scheduler =>
    var ticker = new BackgroundTicker(scheduler);

    int count = 0;
    ticker.Ticker.Subscribe(_ => count++);

    // full control of the time without waiting for 1 second

Show Me The Code

Without further ado, the C++ version is not very far from the C# version. In a simple test, we can parameterize a sequence of integer values arriving at specified intervals (a ticker) with a coordination (why coordination and not scheduler, read in the RxCpp developer manual:

auto seq = rxcpp::observable<>::interval(

The deterministic test scheduler API is currently available through a worker created on the test scheduler:

auto sc = rxcpp::schedulers::make_test();
auto worker = sc.create_worker();
auto test = rxcpp::identity_same_worker(worker);

The rest should read like English:

int count = 0;

WHEN("one subscribes to an observable sequence on the scheduler") {
  auto seq = rxcpp::observable<>::interval(
              test // on the test scheduler
             ).filter([](int i) { return i % 2; });


  THEN("the sequence is not run at first") {
    worker.sleep(2 /* ms */);

    CHECK(count == 0);

    AND_WHEN("the test scheduler is advanced manually") {

      THEN("the sequence is run as expected") {
        worker.advance_by(8 /* ms */);
        CHECK(count == 5);

The full test can be seen @github, and is built on Travis CI

RxCpp 2

RxCpp 2 and API

The last article on rxcpp was based on a now obsolete version of RxCpp. The key contributor to the library, Kirk Shoop, has kindly provided a rewrite based on the newer, 2.0 API of the library: see the pull request, upon which this article is based.

Since the first article, the project has been enriched with somewhat more readable GIVEN/WHEN/THEN-style tests using Catch 1.

Still Ticking: Scheduler and Coordination in RxCpp 2

The previous articles give examples of managing periodic events, such as ticker ticks and measurements in c++. The following example creates an event loop that will be used for coordinated output of various events to the console:

auto scheduler = rxcpp::schedulers::make_same_worker(

auto coordination = rxcpp::identity_one_worker(scheduler);

One such sequence of events is some kind of measurement 2

auto measure = rxcpp::observable<>::interval(
        // when to start
        scheduler.now() + std::chrono::milliseconds(250),
        // measurement frequency
    // take Hz values instead of a counter
    .map([&FM](int) { return FM.Hz(); });

auto measure_subscription = measure
    .subscribe([](int val) {
        std::cout << val << std::endl;

Why didn’t it tick?

If this code were the end of the main program, there wouldn’t be any observable ticks, as all the objects would be destroyed before the first scheduled event. To see the code in action, we shall wait for some condition that will change when we’re done. This step is not necessary if there’s a GUI toolkit event loop that keeps objects alive, but it has to be simulated for a console example.

To demonstrate the subscription change and wait for some time, we’ll wait twice for an atomic variable to become zero:

std::atomic<long> pending(2);


// after all subscriptions defined
while (pending) {
    sleep(1000); // wait for ticker and measure to finish

Tick and Stop

The other ticker will have another period, will only tick 10 times, and then decrement the pending counter:

auto ticker = rxcpp::observable<>::interval(
    scheduler.now() + std::chrono::milliseconds(500),

    .subscribe([](int val) {
        std::cout << "tick " << val << std::endl;
        --pending; // take completed the ticker

Now, we can schedule the termination of the measurement (decrement pending) subscription halfway through the 10-tick run. This scheduling is done on the same scheduler that is running all the subscriptions:

scheduler.create_worker().schedule(scheduler.now() + std::chrono::seconds(2), 
    [&](const rxcpp::schedulers::schedulable&) {
        std::cout << "Canceling measurement ..." << std::endl;
        measure_subscription.unsubscribe(); // cancel measurement
        --pending; // signal measurement canceled

The result:

tick 1
tick 2
tick 3
Canceling measurement ...
tick 4
tick 5
tick 6
tick 7
tick 8
tick 9
tick 10

Thanks, Kirk & other library contributors!

Code @ github

Next: deterministic testing of concurrent behavior

  1. i.e. create.cpp
  2. Observe the convergence of the API towards the C# version.

Presenting at TU-Munich: testing on c++ projects, Thursday, March 26, 2015 7:00 PM

Expecting Thank you to all for a superb heated debate! next week

“no excuses for not testing on c++ projects”

Thursday, March 26, 2015
7:00 PM

details: http://www.meetup.com/MUCplusplus/events/220628575/

If only all test were comprehensible…

SCENARIO("acquiring wisdom") {

  GIVEN("an oracle") { 
    oracle gus;
    WHEN("I ask it to speak") {
      auto answer = gus.speak();

      THEN("wisdom is apparent") {
        CHECK( answer != "bla" );


→ The code can be found @github, including the presentation slides.

Setting travis-ci with github for a c++ project for the first time


Here, I’m trying to use travis-ci, c++, github, CATCH, premake together with my undoredo-cpp library to reduce entropy, try out continuous integration and behavior-style tests.

As a “one-man show” programmer at least at home, I’m trying to keep the discipline of writing tests first. “Growing Object-Oriented Software Guided by Tests” is perhaps a good, although a comparatively dry book for those who are not yet convinced. The blog post by Phil Nash about his latest version of the c++ single-header testing framework CATCH, moved me to finally get my hands on the free continuous integration service travis-ci, along with CATCH with a goal to rewrite the tests for my undo-redo c++ adventure in a more behavior-driven-style.

The undo-redo library is already there, and the tests as well – in gtest (see the master branch). I’d label them “explorative” at the moment since there are just too many assertions per test case, which means I’m repeating myself.

Starting continuous integration at travis-ci for c++

To start my CATCH-“BDD” exploration I’ve setup the branch first: https://github.com/d-led/undoredo-cpp/tree/catchmoci. At my landing page the project is switched on for catching the commit hooks:


The following configuration file .travis.yml is placed for travis-ci to know what to do with my non-conforming repo:

language: cpp

   - catchmoci

before_script: ls

  - make -C Build

As in all TDD practice, the build fails due to the reason that the build doesn’t work yet at all. Adding a status image to my README shines:


Fixing the build

The makefiles are created using premake4, which is a single-file makefile generator based on lua. Unfortunately, I couldn’t force the CI-virtual machine execute my binary premake4, so I had to add the generated makefiles. Now that the make process works, the tests still don’t compile:


Fixing the tests

Once the bulk of the assertions have been rewritten for CATCH, the build still failed due to an ambiguity in serializing std::nullptr_t. Fortunately, Phil has thought of (or rather tested) that, and has a macro which can be defined for the build, fixing it: CATCH_CONFIG_CPP11_NULLPTR.

Voila, travis-ci vm is happy:


Just checking locally if test reporting is fine by adding a spurious test temporarily:


The test-rewrite has been successful and all pass, the badge is green and I can go to bed


But! It’s not the end of the story! BDD! Mock-objects!