Andreas

@seferov Currently the only thing I get is the stack trace I posted, I have not found a hint in there what part of the codebase Psalm/Psalm-Plugin was analyzing at...

But is there a way to identify which file Psalm is looking at when the error occurs? I have 57 commands in this project, so it would be important to...

Maybe the baseline functionality in Psalm and PHPStan can be an inspiration on how to implement it in PHPCS, and having a similar behavior as those two libraries could also...

After testing with 6.1.3 and 6.1.4 these are the orders of event listener executions in 6.1.3: - My own CSP response listener with priority 10 - `ErrorListener::removeCspHeader` - `ErrorListener::onKernelException` -...

After some more investigation I found out where the problem occurs: in `RequestDataCollector` on line 113: ```php 'stateless_check' => $this->requestStack?->getMainRequest()->attributes->get('_stateless') ?? false, ``` `RequestDataCollector` is called by `ProfilerListener`, and `getMainRequest`...

I was thinking the same thing, as even after an exception occured `getMainRequest` should still return a request as far as I would understand it, just like it did before...

I found the last piece of information explaining this bug and have created a new symfony project from scratch showcasing the bug: [exceptionpagebug.zip](https://github.com/symfony/symfony/files/9482383/exceptionpagebug.zip) In my case, a special exception was...

I am using quite a "vanilla" Webpack + TailwindCSS configuration, so it would be preferable to not add SASS as a dependency, because I am not using it otherwise. This...

If PostCSS can help with this it would definitely be a good option, as PostCSS is always part of TailwindCSS projects and most other projects using Webpack. Tailwind also supports...

Sure, I did just that: [Symfony Issue 32900 with Validator.zip](https://github.com/symfony/symfony/files/3464790/Symfony.Issue.32900.with.Validator.zip) Just run `composer install` and then `composer phpunit`. The one test fails, which is the behavior I think should be...