Markus Staab

hmm thinking again about it.. shouldn't the rector release have a prefixed version of nikic/php-parser? ``` PHP Fatal error: Cannot declare interface PhpParser\Parser, because the name is already in use...

we don't have the `DebugClassLoader` in our vendor/. but you are right, that the description in https://github.com/rectorphp/rector/discussions/8566 sounds like the very same problem is there a reason php-parser is not...

I have a feeling that for some reason the preload.php of rector is included while phpunit is running. and the hardcoded require statements lead to a cannot redeclare error, because...

ok, I got more information: the preload.php is loaded via the following stacktrace  it is triggered because we have test-cases in our phpunit suite which `extends AbstractRectorTestCase` and these...

maybe it would make sense to have something like ``` if (interface_exists('\PhpParser\Node')) { return; } ``` at the very beginning of `preload.php` so the require calls are skipped? ---- different...

continue the discusson from https://github.com/rectorphp/rector-src/pull/5742#discussion_r1530735959 ... a `conflict` in `composer.json` with e.g. `nikic/php-parser:5.*` would lead to a `nikic/php-parser:4.*` at the project level, which still would trigger "cannot re-declare" errors as...

Phpstan has a separate `--debug` switch for enabling/disabling parallel (and an additional `--xdebug` switch to attach xdebug. They do work independently

I don't see a way to improve it. maybe it would help to remove it from the known set-lists so people don't use it accidentally

we would need a full picture of all known classes within `FamilyRelationsAnalyzer` which would mean it would only work if we run `FinalizeClassesWithoutChildrenRector ` after everything else and only when...

we could do inefficient stuff like trigger static reflection of all classes registered in the composer autoloader, so at least the hacks mentioned on https://getrector.com/documentation/static-reflection-and-autoload would take effect. but I...