I've added some integration tests to verify that everything works. Some new results:
In Postgres 17, the "builtin" locale provider seems to be the best option with locale "C.UTF-8" as it provides the same sorting as Solr and supports uppercase/lowercase transformations for Unicode characters while being independent of the OS, see https://postgresql.verite.pro/blog/2024/07/01/pg17-utf8-collation.html for some context. In Postgres 16 and earlier, the "builtin" provider doesn't exist. The locale "C.utf8" seems to be the next best option whose behavior is quite similar but which depends on the system.
The Oracle database seems to have the same problem. Further, Oracle doesn't even correctly handle lower/upercase conversion with non-ASCII character. I haven't yet found out how to configure the Oracle database to fix this.
This message was sent by Atlassian Jira (v9.3.0#930000-sha1:287aeb6)
If image attachments aren't displayed, see this article.