(Reporting on behalf of Venelin @ Ontotext)
The Workbench does not appear to process queries with Cyrillic characters correctly.
Here is the scenario to reproduce :
1/ Insert following triple using Sesame Workbench
2/ Verify that triple is stored correctly : verification passes OK - title is returned with correct encoding in Workbench
3/ Execute a regex query - which returns no results, which is the problem
When doing this same experiment using the commandline console instead of workbench, the query returns the expected result - so this seems to indicate that the problem is in how the Workbench encoded the query.
As an aside: I should point out that I have tried the workaround solution mentioned here: http://openrdf.callimachus.net/sesame/2.7/docs/articles/serverinstall.docbook?view#workbench-utf-8 for POST requests, but it made no difference.