Scheduled pipelines now available in Bitbucket Pipelines

BitBucket blog announces the support for scheduled Bitbucket Pipelines.  This is super cool and has been on the wishlist for a while now.  Here are a few examples of how this feature is useful:

  • Nightly builds that take longer to run
  • Daily or weekly deployments to a test environment
  • Data validation and backups
  • Load tests and tracking performance over time
  • Jobs and tasks that aren’t coupled to code changes

Making “Push on Green” a Reality

Making “Push on Green” a Reality is an insider look at how Google handles continuous deployment.  Very few teams and companies need to deal with such level of complexity, but the overall principals still probably apply.

Updating production software is a process that may require dozens, if not hundreds, of steps. These include creating and testing new code, building new binaries and packages, associating the packages with a versioned release, updating the jobs in production datacenters, possibly modifying database schemata, and testing and verifying the results. There are boxes to check and approvals to seek, and the more automated the process, the easier it becomes. When releases can be made faster, it is possible to release more often, and, organizationally, one becomes less afraid to “release early, release often”. And that’s what we describe in this article—making rollouts as easy and as automated as possible. When a “green” condition is detected, we can more quickly perform a new rollout. Humans are still needed somewhere in the loop, but we strive to reduce the purely mechanical toil they need to perform.

PHP Smart Analyzer

PHP Smart Analyzer (or PHPSA for short) is yet another item in a growing list of tools for PHP code static analysis.  It’s in an early alpha state, but looking at the list of goals, it’s quite promising.

If that’s up your valley, have a look also at PHPQA and PHPStan, which I wrote about earlier.

BitBucket Pipelines improved support for Docker

Here are some exciting news from the BitBucket Pipelines blog: Bitbucket Pipelines now supports building Docker images, and service containers for database testing.

We developed Pipelines to enable teams to test and deploy software faster, using Docker containers to manage their build environment. Now we’re adding advanced Docker support – building Docker images, and Service containers for database testing.

PHPUnit Snapshot Assertions – a way to test without writing actual test cases

phpunit-snapshot-assertions – is an interesting addition to the PHPUnit assertions which allows testing against previously created snapshots.  This is particularly useful for testing the outputs of API end-points, format conversion functions, and the like.  Instead of testing the actual functionality, these assertions allow to compare the output of the current test run with the known good output of a previously created snapshot.

This works well for generic text, but even better for widely used formats like JSON and XML, where, in case of a failed assertion, a meaningful difference can be provided.

Here is a blog post providing some more details on philosophy and methodology.

PHPQA all-in-one Analyzer CLI tool

PHPQA all-in-one Analyzer CLI tool.  This project bundles together all the usual PHP quality control tools, and then some.  It simplifies the installation and configuration of the tools and helps developers to push up the quality control bar on their projects.

The tools currently included are:

Preparing for the PHPUnit 6 and PHP 7

If you woke up today and found that most of your PHP projects’ and libraries’ tests break and fail, I have news for you:  you are doing something wrong.  How do I know?  Because I was doing something wrong too…

First of all, let me save you all the extra Googling.  Your tests are failing, because a new major version of PHPUnit has been released – version 6.0.0.  This version drops support for PHP 5 and, using the opportunity of the major version bump, gets rid of a bunch of stuff that was marked obsolete earlier.

But why does it fail, you ask.  Well, because PHPUnit is included in pretty much every composer.json file out there.  And the way it’s included is almost always is this:

"require-dev": {
"phpunit/phpunit": "*",
}

PHPUnit being a part of pretty much every composer.json file, is probably the reason why people want to be much more relaxed with the used version, than with any other component of the system.  That’s usually good.  Until it breaks, much like today with the release of the PHPUnit 6.

How can you fix the problem? Well, the quickest and the easiest solution is to update the composer.json with “^5.0” instead of “*”.  This will prevent PHPUnit from upgrading until you are ready.

While you are doing it, check the other dependencies and make sure that none of them are using the asterisk either.  Because, chances are, the exact same problem will happen later with those too.

The only difficult bit about this whole situation is the correlated drop for the PHP 5 support.  Yes, sure, it has reached its end of life, but there are still a lot of projects and environments that require it, and will require it for a lonweg time.

As you are the master of your code and dependencies, other people are of their own.  So you can’t really control when each of your dependencies will update the requirement for the PHPUnit 6, or any other tool that requires PHP 7.

On the bright side, major releases of PHP don’t happen that often, so this shouldn’t be the frequent problem.

Quick and easy introduction into PHP Mess Detector (PHPMD)

PHP Mess Detector is yet another one of those tools that help to keep the code base manageable and clean.  Here’s the description straight from the site:

What PHPMD does is: It takes a given PHP source code base and look for several potential problems within that source. These problems can be things like:

  • Possible bugs
  • Suboptimal code
  • Overcomplicated expressions
  • Unused parameters, methods, properties

Here is how you can jump right in.  It’s super easy.  It only takes 6 steps.

Step 1: Pick a project to try it on.

You can use any of your own PHP projects, or grab one from GitHub.  It doesn’t matter.  You’ll know better where to apply it once you get comfortable with the tool.  For sake of this quick guide, I’ll use one of our Open Source repositories – cakephp-groups plugin.

cd /tmp
git clone git@github.com:QoboLtd/cakephp-groups.git
cd cakephp-groups

Step 2: Install PHPMD with composer.

composer require phpmd/phpmd

Step 3: Run PHPMD.

If you run “./vendor/bin/phpmd“, you’ll see a help screen. But what’s the purpose of this blog post if you have to read the manual, right? So, let me simplify it for you. PHPMD needs three parameters:

  1. Path to the PHP source code that it will be examining.  We’ll use “src/“.
  2. Report format – one of: xml, text, or html.  We’ll use “html“.
  3. A choice of mess detection rules that you want it to apply.  You can create your own, or you can pick one from: cleancode, codesize, controversial, design, naming, unusedcode.  We’ll use “unusedcode“.

Also, we’ll give it an extra one: “–reportfile“, because by default PHPMD will spit everything to the standard output.  So, let’s put it together and see what we’ve got.

phpmd src/ html unusedcode --reportfile phpmd.html

Step 4: Examine the report.

After running PHPMD command above, you’ll find a phpmd.html file in the same folder. Here’s how it looked for me, when open in the browser.

PHP mess detector

So, PHPMD found one problem in the “src/Shell/Task/ImportTask.php” file on line 93.  Here’s the relevant piece of code:

    protected function _getImportErrors($entity)
    {
        $result = []; 
        if (!empty($entity->errors())) {
            foreach ($entity->errors() as $field => $error) {
                if (is_array($error)) {
                    $msg = implode(', ', $error);
                } else {
                    $msg = $errors;
                }
                $result[] = $msg . ' [' . $field . ']';
            }
        }

        return $result;
    }

As you can see (line 09 above is line 93 in the report), the issue reported by the PHPMD is a typo in the variable name. It should be $error, not $errors.

Step 5: Fix the problem.

  • Rename the $errors variable to $error.
  • Rerun the PHPMD report as per Step 3.
  • Examine report as per Step 4 to make sure that the problem is fixed and no new issues were introduced.
  • Create a new branch.
  • Commit the code.
  • Push the branch to GitHub.
  • Create the Pull Request.

All of the above mini steps took about 7 seconds.

Step 6: Pour yourself a drink.

You’ve just learned how to use a new tool, found a bug, and submitted a patch to the Open Source project.  At least I hope you did.

Not bad at all.

If you are wondering what to do next, here are a few suggestions:

  • Try running PHPMD for other types of issues.  As I said, it supports cleancode, codesize, controversial, design, naming, unusedcode, and we’ve only ran it for the “unusedcode”.  See what else is there.
  • Integrate PHPMD into your projects, to run automatically, together with your unit tests.  You do have automated unit tests, right?
  • Customize the ruleset that PHPMD is using to find more/less issues, which are maybe more specific to your project.
  • Use your newly acquired knowledge to fix issues with more Open Source projects.  You’ll make a name for yourself and you’ll make a world a better place.

Let me know how it goes.