Scroll down to the script below, click on any sentence (including terminal blocks!) to jump to that spot in the video!Cool, got it! Show me the script!
We've already helped push forward a pull request, solved an issue and even reported a bug. Hello! We deserve cake!
And we deserve to move up one more level of difficulty: it's time to contribute new code with a pull request. Let's look at an issue I found: #27835.
This comes from the Security component. Let me give you some background: if you try to access a protected page as an anonymous user - like
/admin - Symfony stores that URL to a special key in the session. Then, after you login, Symfony reads this key and redirects the user back to that URL.
Occasionally, it's useful to manually set that session key to control where the user goes after logging in. To help with that, Symfony has
TargetPathTrait. The problem is that, to use its
saveTargetPath() method, you need something called the "provider key"... which is actually just the "name" of your firewall. You could hardcode it, but, that really shouldn't be necessary.
In a recent version of Symfony, a feature was added so that you can read the firewall name by getting a
FirewallMap object, calling
getFirewallConfig() and then calling
But, here's the problem: that
FirewallMap service is not an autowireable service. That makes it inconvenient to use. And, one of the core contributors gives a reason behind why that service is not autowireable.
You can work around this. But, I had an idea: I'm not even sure if it's a good idea, but let's try it. What if we created a new
TargetPathHelper class that allowed you to set this "target path", without needing the provider key. Internally, it would use the
FirewallMap to figure it out automatically.
The end user could use this new class without needing to worry about the firewall name at all.
If you don't completely understand, that's ok. The important thing is the process we're going to use to bring this new idea to life!
Go back to PhpStorm: let's change our project to look only at the
symfony/ directory. Then, find the terminal that's in this directory. We're still on the feature branch from
colinodell. Start by making sure your copy of Symfony is up to date by running:
git fetch origin
And then create a new branch for our feature:
git checkout -b target-path-helper origin/master
This is important: we just created a new branch based off of Symfony's
master branch. Why? Why not base the branch off of Symfony's 4.1 branch - that's the currently-released version?
Let's talk about Symfony's branching system. It's... kinda simple. If you're adding a feature, it should always be made to the
master branch. Then, it will be included in the next Symfony minor release: in this case Symfony 4.2. But if you're fixing a bug, you should fix that in the oldest, supported branch where the bug exists. For example,: if you found a bug that was introduced in Symfony 3.4, create your branch based off of
But, if a bug was first found in version 3.2, you actually would not base your branch off of Symfony's 3.2 branch. Why? Because Symfony 3.2 is no longer supported. To help understand this, go to https://symfony.com/roadmap. At this moment, only three versions of Symfony are supported: 2.8, 3.4 and 4.1.
So, if you found a bug in Symfony 3.2, you would fix it in 3.4, because that's the oldest, supported version of Symfony that contains the bug. If you found a bug in Symfony 2.7, you would fix it on the 2.8 branch.
But... if I fix a bug in 2.8... won't that bug still exist in 3.4 and 4.1? Ah, a very good question. But... no! The core team routinely merges all old branches - like 2.8 - up into the newer branches, like 3.4 and 4.1. If you fix a bug in 2.8, it will also be merged & included in all newer versions. Booya!
Anyways, because this will be a new feature, our branch is based off of
origin/master. And now, we're ready to code. Let's do that next!