Monday 31 July 2017

Yeah, you do want 100% test coverage

I've read a number of blog articles recently that have been eshewing 100% test coverage. The rationale being that there's little difference between 90% and 100%, and that that last 10% can be quite tricky to attain as - we being human 'n' all - sometimes leave the worst to last. Obviously we should all be doing TDD so that situation would not arise, but - again - we're human so don't always do TDD for one reason or another ("can't be arsed" is the usual one).

Another rationale is that it's not a race to 100%. IE: the object of the exercise is delivering working software which adds business value; it's not "getting 100% test coverage".

The latter is true. But we still aim for 100% test coverage for basically one good reason: visibility.

First, look at this:

Not bad I guess. Not great.

Now look at this:

We've accidentally forgotten to cover something new. Can you see it (it's one of the provider methods).

On the other hand look at these two:

Somewhat easier to spot, yeah?

This is bloody handy. In this case I just commented-out one of my tests to get the report off 100%, but in the real world I'd see that and go "huh, WTF?", and perhaps discover I had a logic branch untested, or a test I thought was covering something wasn't actually doing so. Or I just forgot to put the @covers annotation on the test. Whatever the situation: it's handy for the shortfall to be so obvious.

What it also means is part of our pull-request process is running a Bamboo build which runs all our tests (plus some other code quality tools):

 Everything's green here, so it's OK for code review. If it was failing, I'd get a nudge from my reviewers going "erm... you sure this is ready to go?" And even if I thought it was, we can't merge anything whilst the build is failing anyhow. The issue needs to be addressed.

Now the key thing here is this:

We don't test everything, but we cover everything. Where "cover" might be "we don't need to test this". In this case a single variable assignment does not need testing. We trust PHP to follow simple instructions. The thing is: this still shows up as green in the test coverage report.

We do this in a few situations:

  • the method has not branching logic and is overwise pretty simple. Ignore it.
  • we're backfilling coverage on old code we need to maintain, but it's such a mess we simply don't have time to solve all the testing shortfalls right now, so we test what is urgent (the bit we're changing), but just mark as ignore the rest of it. Now that file is green, and we can see if it stops being green.
  • similar to the above but we know the code the code in question is  not long for this world. So we ignore it and stick a TODO cross referencing to the ticket that will be dealing with it. And the ticket must already be lined-up to be worked on "soon".

There might be other situations where we need to be pragmatic and skip some stuff, but this needs to be explained to the code reviewers, and get their agreement.

But being able to run a report that says "100%", and shows up anything that isn't 100% is gold.

Get yer coverage report green, but don't be afraid to be pragmatic and skip stuff, if there's an informed reason to do so.

Now this isn't to say one can just pepper ignore annotations around the place just to get the report green. That's daft. The object of the exercise is not to make some report green / say 100%. It's to get good code quality. This is just a tool to help identify the quality: it's still up to the devs to provide the quality. Don't forget that. It needs to be an informed, honest decision for one to ignore code.



Sunday 23 July 2017

PHP: A function that returns two different things

I'm gonna plagiarise one of me own answers to a Stack Overflow question here, as I think it's good generic advice.

The question was "PHP 7: Multiple function return types". The subject line sums it up really. The person here wants to return a value or false from a function. I see this a lot in PHP code. Indeed... PHP itself does it a lot.

My answer was as follows:


From a design perspective, having a function that potentially returns different types of result indicates a potential design flaw:

  • If you're returning your result, or otherwise false if something didn't go to plan; you should probably be throwing an exception instead. If the function processing didn't go according to plan; that's an exceptional situation: leverage the fact. I know PHP itself has a habit of returning false if things didn't work, but that's just indicative of poor design - for the same reason - in PHP.
  • If your function returns potentially different things, then it's quite possible it's doing more than one thing, which is bad design in your function. Functions should do one thing. If your function has an if/else with the true/false block handling different chunks of processing (as opposed to just handling exit situations), then this probably indicates you ought to have two functions, not one. Leave it to the calling code to decide which is the one to use.
  • If your function returns two different object types which then can be used in a similar fashion in the calling code (ie: there's no if this: do that; else do this other thing in the calling code), this could indicate you ought to be returning an interface, not a concrete implementation.
There will possibly be legit situations where returning different types is actually the best thing to do. If so: all good. But that's where the benefit of using a loosely typed language comes in: just don't specify the return type of the function. But… that said… the situation probably isn't legit, so work through the preceding design considerations first to determine if you really do have one of these real edge cases where returning different types is warranted.

I'll add two more considerations here.

There's another option which is kinda legit. In PHP 7.1 one can declare the return type to be nullable.

Consider a function which potentially returns a boolean:

function f($x) {
    if ($x) {
        return true;

We can't declare that with a return type of bool as if our logic strays done that falsey branch, we get a fatal error:

function f($x) : bool {
    if ($x) {
        return true;

f(1); // OK

f(0); // PHP Fatal error:  Uncaught TypeError: Return value of f() must be of the type boolean, none returned

But in 7.1, one can declare the return type as nullable:

function g($x) : ?bool {
    if ($x) {
        echo "returning true";
        return true;

    echo "returning null";
    return null;

g(1); // returning true
g(0); // returning null

Now sometimes this is legit. One might be looking for something from some other resource (eg: a DB record) which quite legitimately might not exist. And there might not be a legit "null implementation" of the given class one is wanting to return: for example returning the object with some/all of its properties not set or something. In this case a null is OK. But remember in this case one is forcing more logic back on the calling code: checking for the null. So not necessarily a good approach. I'd consider whether this situation is better handled with an exception.

The last consideration is that one can still specify multiple return types in a type-hint annotation, eg:

 * @param bool $b
 * @return A|B
function f($b){
    if ($b) {
        return new A();
    }else {
        return new B();

This doesn't actually enforce anything, so it's mostly a lie, just like all comments are, but at least one's IDE might be able to make sense of it, and give code assistance and autocomplete suggestions when writing one's code:

See how it offers both the method from an A and a B there.

Still: I really dislike these annotations as they're really just comments, and especially in this case they're lying: f can actually return anything it likes.

Right... must dash. I am supposed to be down @ my ex's place hanging out with my son in 10min. I'll press "send" and proof-read this later ;-)



Saturday 8 July 2017

@dac_dev Twitter account

Yes, this thing is still alive.

This morning I indicated on Twitter I was discontinuing my usage of the @dac_dev Twitter account. I'll just be using my @adam_cameron one from now on.

@NXRK made this observation just now:

Just to be clear: my twitter account is the @adam_cameron one. The @dac_dev one was created specifically for promoting this blog, and the "brand" that was my existence in the dev world, such as it is. Specifically the CFML dev world.

I don't want to promote that brand any more, or at least not such that I want to keep it separate from my "real" self. Or more that there's simply not enough usage of that account to warrant me having to keep it separate from my own account.

I won't be deleting the @dac_dev one, and I have set it to send me emails if anyone hits me up there, but I am not paying attention to it any more, nor do I intend to post anything to it.

As with anyone's account on Twitter, you are free to keep an eye on my mumblings on the @adam_cameron one. I'm less polite there (if that can be believed...), and it's definitely more "general" commentary. Rugby, cricket, specious social commentary, naive political observations and photos of my meals. Well: not the latter. I'm not a complete twat(*).

In a closely-related moved, I've chucked-in my participation on the CFML Slack channel too. I was not getting anything positive out of my membership there, and she shitness of some elements of the CFML community was doing my head in (if yer reading this, I'm almost certainly not including you in that shit demographic).



(*) based on just that metric, I mean.