[php-src] Issue #11866: PHP preg_match wrong matching behaviour for lookahead containing < character

Issue: https://github.com/php/php-src/issues/11866 Author: jild ### Description The following code: ```php preg_match( '/<text:p[^>\/]*>(?:.(?!<text:p|<\/text:p))*?\${container.überschriftabw.versumf}(?:(?<!<\/text:p>).)*?<\/text:p>/is', $xml, $matches1, PREG_OFFSET_CAPTURE ); <?php ``` Resulted in this output: ``` > $matches1 = [ [ "<text:p text:style-name="P99"></text:p></table:table-cell><SOME MORE XML>${/container.überschriftabw.versumf}</text:h><text:p text:style-name="P71">${container.abw.versumf}</text:p>", 112845, ], ] ``` But I expected this output instead: ``` This part of the regex: (?:.(?!<text:p|<\/text:p))*? should not match the first passage: <text:p text:style-name="P99"></text:p> ``` To make it work, I remove the < from the mentioned part to get the following regex: /<text:p[^>\/]*>(?:.(?!<text:p|\/text:p))*?\${container.überschriftabw.versumf}(?:(?<!<\/text:p>).)*?<\/text:p>/is Then test it against the matched string. Now the negative lookahead is working again. The output is [] for $matches1 I tested it with php101.com and didn't experience such weird behaviour. ### PHP Version 8.0.27 ### Operating System _No response_