Qt used to be at the forefront of Unicode handling, but there are several features that are not supported. In particular, final sigma handling is missing in toLower(), and in the process of implementing it, I found that we don't have all the necessary Unicode properties in our tables. In particular this is true of PropLists.txt's Other_Lowercase.

This epic exists to collect other such missing features.

In general, there are three ways to proceed:

• ditch our own implementation and make ICU mandatory
  • the problem with this approach is mainly that ICU continues to be 32-bit only (https://unicode-org.atlassian.net/browse/ICU-22089), and chunking is nearly impossible without encoding knowledge about Unicode, anyway. But if ICU ever becomes 64-bit-clean, this is by far the most attractive option
• wait for C++ to land Unicode support and rebase on top
  • the problem here is that this is, realistically speaking, a decade or more out. Not just on C++, but until Qt can use such a new C++ version
• improve our own implementation