Case-insensitive \p{} issue

[michaeltang4829]

Hello,

We're using PCRE2 10.43, and we found what looks to be a bug with unicode properties while using character case properties. When the pattern has the case-insensitive modifier, regardless of the character case being only upper-case, should also find lower case characters.

Text: a
Pattern: (?i:\p{Lu})
Result: No matches
Expected: a

There's also documentation in section 17 in the pcre2compat page hinting that this was the original behavior in Perl, but later Perl corrected this behavior.

Thanks!

[PhilipHazel]

Yes, I think that's right. At the time Perl changed I thought they were wrong - I thought if you had said \p{Lu} you should get an upper case letter, even in /i mode. However, now that I think about it, I have changed my mind. After all, when you specify A (for example) you expect it to match the lower case in /i mode. So I now agree; PCRE2 should change. I will look into it.

[PhilipHazel]

Commit 6d82f0c makes PCRE2 behave like Perl. Lu, Ll, and Lt now all behave as Lc when /i is in force.