Spring Data Cassandra Turkish Locale Error [DATACASS-805] #971
Comments
Mark Paluch commented Thanks for the details. After having a look at the CQL generator, the translation into a Turkish i without dot happens within the Cassandra driver. Spring Data specifically calls |
Mark Paluch commented I filed a bug report for the Cassandra driver and attached the link to the ticket. Looking at the Spring Data code, we use Once the driver has addressed the issue, we're going to upgrade to the new driver version anyway so closing the ticket here |
Siddik Acil opened DATACASS-805 and commented
Gist: Spring Data Cassandra fails when using TR locale (-Duser.language=tr) if you have typenames that contains 'I' or 'i'. (UUID, String, int)
I got the error at attachment file spring-cassandra-dump.txt when I was working on a simple demo with my work pc which has Turkish locale.
Having had similar experiences with Turkish locale, I passed -Duser.language=en and everything starts working expectedly.
This is most probably due to notoriously famous Turkey test. Turkish locale has a dotless ı which has a "weird" conversion rule that only makes sense to native Turkish speakers.
For more info on Turkish I and İ refer to this link
The query builder has an uppercase typename lowercased with default locale but I could not find where.
I have attached a sample spring project with docker-compose.yml that spins up a YugabyteDB instance.
Steps to reproduce.
Get the GitHub repo.
Go into compose/master-slave execute "docker-compose up"
Get spring project up and running via VM option -Duser.language=tr
Reference URL: https://github.com/Akaame/SpringCassandraYugabyte
Attachments:
The text was updated successfully, but these errors were encountered: