[Rector] Testing Rector dev-main

[samsonasik]

Description

Hi @kenjis @paulbalandan could you test the rector/rector:dev-main with clean up parallel config below and verify if the performance drawback resolved without setting $jobSize config in the parallel() config below?

If that works, we can remove passing arguments in the parallel() config in next rector release.

Current rector dev-main changed the parallel process by @staabm . I mention you here so you can have fast feedback here in case we found a different use case.

Thank you.

Checklist:

• Securely signed commits
• Component(s) with PHPDoc blocks, only if necessary or adds value
• Unit testing, with >80% coverage
• User guide updated
• Conforms to style guide

[kenjis]

It seems no problem.

(test-rector-dev-main-2 $)$ time vendor/bin/rector 
 821/821 [▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓] 100%

                                                                                                                        
 [OK] Rector is done!                                                                                                   
                                                                                                                        


real    6m1.921s
user    12m38.855s
sys     0m30.725s

[paulbalandan]

I still could not test because of the issue in space in path. Sorry.

[staabm]

btw: I can't make sense of a jobsize of 1. this should be pretty inefficient. nice its gone with this PR :)

[samsonasik]

Thanks all 👍 , I will update the config on next dependabot PR for new rector release 👍

[staabm]

Fyi - on my machine the new release is 20-25% faster in my codeigniter tests in comparison to 0.18.2 (with the suggested changes in parallel parameters)