Regarding bug 18152 MARC::Record->encoding should not be called directly from our codebase. C4::Charset::SetMarcUnicodeFlag should call MARC::Record->encoding for MARC21 and NORMARC but does the UNIMARC trick for UNIMARC records. It seems that MARC::Record->encoding calls should be replaced with C4::Charset::SetMarcUnicodeFlag everywhere.
Given that bug 18152 was reverted, and calling C4::Charset::SetMarcUnicodeFlag actually calls MARC::Record->encoding even for UNIMARC, should this bug be closed, or should it be depending on another bug that will try again to do what bug 18152 failed to do?