I don't like the SimplifyRegexPatternRector rule. Because it may change the behavior.
In PHP, [0-9] and \d (also [a-zA-Z] and \w) are not exactly the same when using u option.
https://3v4l.org/N8sQ5

Ok, it seems due to called on $rectorConfig->rule(), I removed it and revert its change on tests files 48f36ef

btw, the u usage already skipped on rector :) https://getrector.com/demo/00372e82-e4bc-451b-a7a2-495637f29d88

Thanks!

@kenjis do you want to re-enable SimplifyRegexPatternRector ? as it already skipped on u usage, see https://getrector.com/demo/00372e82-e4bc-451b-a7a2-495637f29d88

I don't want. Because these patterns have a bit different meanings.

An answer by ChatGPT:
In PHP's PCRE, \d and [0-9] both match digits but are not exactly the same.

1. Meaning:

   • \d matches any digit character, including Unicode digits in some cases.
   • [0-9] matches only ASCII digits (0-9).

2. Compatibility:

   • \d can match non-ASCII digits if Unicode is enabled.
   • [0-9] always matches ASCII digits only.

3. Performance:

   • [0-9] can be slightly faster since it's a simple character class.

Example

$pattern_d = '/\d/';
$pattern_class = '/[0-9]/';
$string = "123abc";

preg_match($pattern_d, $string); // Matches
preg_match($pattern_class, $string); // Matches

Conclusion

Use [0-9] for consistent matching of ASCII digits. Use \d if you need to match broader digit sets, considering encoding and Unicode support.

By the way, where does this code https://github.com/rectorphp/rector/blob/main/rules/CodeQuality/Rector/FuncCall/SimplifyRegexPatternRector.php check for the u option?

It only check delimiter, which limited to ~, #, / via RegexPatternDetector, so u in the last will always be skipped.