Liquibase 4 UTF-8 not recognized

After upgrade from last Liquibase 3 to version 4, now UTF-8 files containing special characters ( like é ) does not process those charasters correctly and some strange chars are inserted into Oracle instead. If the file is encoded into UTF8 BOM, then it is inserted correctly. I cannot find any change description about this. Is this bug or feature?

@ludek ,

Hmm good question, let me include one of our Product folks @mariochampion who should know the answer.