Fix PHPStan L2 error "Binary operation "[something]" between Foo and Bar results in an error."

Applied the patch locally
Change phpstan to level 2 and ran

./core/scripts/dev/commit-code-check.sh > phpstan.txt

And found no instances of "Binary operation"

so looks good.

A few comments/questions:

1. +++ b/core/lib/Drupal/Component/Uuid/Php.php
   @@ -32,6 +32,7 @@ public function generate() {
   +    /** @var int $clock_seq_hi_and_reserved */
        $clock_seq_hi_and_reserved = base_convert(substr($hex, 16, 2), 16, 10);
   

   Why do we use @var here but casting in some other places?

2. +++ b/core/lib/Drupal/Core/Datetime/Element/DateElementBase.php
   @@ -40,7 +40,7 @@ protected static function datetimeRangeYears($string, $date = NULL) {
   -        $min_year = $this_year + $matches[0];
   +        $min_year = $this_year . $matches[0];
   

   I don't think this is correct. The comment says "Valid patterns would be -5:+5, 0:+1, 2008:2010" so if someone specifies "-5:+5" then they would expect addition (2023 + -5 = 2018) not concatenation (20235). I assume there is no test coverage for this!

3. +++ b/core/modules/file/file.module
   @@ -200,7 +200,7 @@ function file_validate_size(FileInterface $file, $file_limit = 0, $user_limit =
   +  if ($user_limit && (\Drupal::entityTypeManager()->getStorage('file')->spaceUsed($user->id()) + (int) $file->getSize()) > $user_limit) {
   

   Should we consider fixing this on the interface instead? Why does getSize() return the size as a string?

4. +++ b/core/modules/jsonapi/tests/src/Functional/EntryPointTest.php
   @@ -57,7 +57,9 @@ public function testEntryPoint() {
   +    /** @var string $user_pass_raw */
   +    $user_pass_raw = $user->passRaw;
   
   +++ b/core/modules/system/tests/src/Functional/Menu/LinksetControllerTestBase.php
   @@ -72,7 +72,9 @@ abstract class LinksetControllerTestBase extends BrowserTestBase {
   +      /** @var string $account_pass_raw */
   +      $account_pass_raw = $account->passRaw;
   

   passRaw is a hack that is only used in tests, wonder if we should refactor it to be consistently defined somehow?

Correct me if I'm wrong but #7.1 and #7.4 are still open?

Yes, I did forget to review the comments after updating the MR.

#7.1 and 2: Fixed
7.4: I think if that is to be done it is better in a separate issue.

Opened 📌 Refactor passRaw Active for 7.4

Rest looks good.

One question in the MR

Oh nice catch.