-
-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix psalm build #6322
Fix psalm build #6322
Conversation
08480b9
to
717e2ad
Compare
@@ -74,6 +74,7 @@ public function __construct( | |||
|
|||
public function reverseTransform($value) | |||
{ | |||
/** @var ArrayCollection<array-key, object> $collection */ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
IMHO, inlining hints for 3rd party dependencies doesn't seem a good choice.
If that's the solution, I'd prefer to allow the failure indicating the cause. That way, when the dependency fixes the issue, we only need to remove the entry for that failure.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe it's considered as the right way to do. I don't know. I'll open an issue on psalm.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What happens if you use this plugin: https://github.com/weirdan/doctrine-psalm-plugin ?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@phansys I found an existing issue and this seems to be the recommended way: vimeo/psalm#3436 (comment)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Surely declaring locally a type that is missing in a dependency will fix the issues caused by that dependency. My point is, we should not create such a compatibility layer for these cases, and instead, let them to fail, declaring these failures as allowed in our own list (psalm-baseline.xml
).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure to understand.
If you ignore the error in psalm-baseline,
- either you'll have to add a new entry every time a
add()
is called on a new createdArrayCollection
. - either you will ignore error like
$collection->add(1)
It's not a type that is missing in a dependency
and Psalm does not consider this situation as a bug:
When you write new ArrayCollection()
you're setting no elements, so the inferred type is ArrayCollection<array-key, empty>`.
IMHO it's not compatibility layer
, it's declaring the right information.
If $collection
was a class property, you would see no issue writting
/**
* @var Collection<array-key, object>
*/
$collection
If it was a param property, you would see no issue writting
/**
* @param Collection<array-key, object>
*/
public function foo($collection)
Here it's the same to me.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
On the previous code, relying on the modelManager we would have written return type like that, so I am okay with the solution, we should probably end up using templates on this function because the object depends on the class string passed on the construct
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's fine. I don't like the idea to provide specific information for a tool in such intrinsic way. I'd expect the tool to leverage the types it knows and let the other cases without any warning.
If the tool doesn't consider this as a problem solvable with a definition change in the package providing the offending method, I think it's our responsibility to fix the issue, since the decision to add the tool, being aware about this situation, was ours.
717e2ad
to
b791dfd
Compare
@jordisala1991 you added the patch label but it is pedantic to me. Do you agree ? |
Yes, Agree |
Thank you @VincentLanglet |
Subject
I am targeting this branch, because BC.
Pedantic