Comment 3 for bug 1483500

Branch at
working user/stompro/lp1483500_pgtap_evergreen_lowercase;a=shortlog;h=refs/heads/user/stompro/lp1483500_pgtap_evergreen_lowercase

I think I might have found a bug in evergreen.lowercase when handling Turkish characters. There are two i's in the Turkish character set, İ-> i, I -> ı, Capitol I with a dot, and without a dot, and lowercase versions of each. But select evergreen.lowercase('İI'); returns " i̇i" when it should return "iı". But I would like someone that has more UNICODE experience to confirm it.

This site talks about the double I problems.

I also can now think of an instance where I used the built in Postgresq lower() function that might break with unicode characters. I'll go back and fix that.